diff --git a/.claude/skills/open-h-rf-convert b/.claude/skills/open-h-rf-convert new file mode 120000 index 0000000..cfa4ff6 --- /dev/null +++ b/.claude/skills/open-h-rf-convert @@ -0,0 +1 @@ +../../skills/open-h-rf-convert \ No newline at end of file diff --git a/.claude/skills/open-h-rf-submission-eval b/.claude/skills/open-h-rf-submission-eval new file mode 120000 index 0000000..77b25b8 --- /dev/null +++ b/.claude/skills/open-h-rf-submission-eval @@ -0,0 +1 @@ +../../skills/open-h-rf-submission-eval \ No newline at end of file diff --git a/.claude/skills/openh-rf-convert b/.claude/skills/openh-rf-convert deleted file mode 120000 index d8fc2bb..0000000 --- a/.claude/skills/openh-rf-convert +++ /dev/null @@ -1 +0,0 @@ -../../skills/openh-rf-convert \ No newline at end of file diff --git a/.claude/skills/openh-rf-submission-eval b/.claude/skills/openh-rf-submission-eval deleted file mode 120000 index 10fbcf1..0000000 --- a/.claude/skills/openh-rf-submission-eval +++ /dev/null @@ -1 +0,0 @@ -../../skills/openh-rf-submission-eval \ No newline at end of file diff --git a/.github/workflows/pages.yml b/.github/workflows/pages.yml index 45a59dc..9cd5710 100644 --- a/.github/workflows/pages.yml +++ b/.github/workflows/pages.yml @@ -5,7 +5,7 @@ on: branches: [main] paths: - "site/**" - - "plots/openh_rf_datasets.py" + - "plots/open_h_rf_datasets.py" - "scripts/authors.csv" - "scripts/compile_authors.py" - ".github/workflows/pages.yml" @@ -13,7 +13,7 @@ on: branches: [main] paths: - "site/**" - - "plots/openh_rf_datasets.py" + - "plots/open_h_rf_datasets.py" - "scripts/authors.csv" - "scripts/compile_authors.py" - ".github/workflows/pages.yml" diff --git a/.gitignore b/.gitignore index eb5ef0a..2753c1c 100644 --- a/.gitignore +++ b/.gitignore @@ -20,7 +20,7 @@ examples/tracked-cirs-phantom/3d_views/ # The Hub repo's documents, cloned by huggingface/setup.sh. Not a submodule: a checkout # that setup.sh does not guard would pull ~39 TB of LFS payload. -huggingface/OpenH-RF/ +huggingface/Open-H-RF/ # The website build (site/build.py); site/figures/ holds a card figure that is not on the Hub. site/_site/ diff --git a/AGENTS.md b/AGENTS.md index 813ee8c..aacb1c0 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -21,14 +21,14 @@ Available skills: | Skill | Purpose | |---|---| -| [`skills/openh-rf-submission-eval/`](./skills/openh-rf-submission-eval/) | Evaluate a contributor submission to the OpenH-RF initiative against the RFP and submission guide. Produces a structured acceptance report. | -| [`skills/openh-rf-convert/`](./skills/openh-rf-convert/) | Source-agnostic dataset conversion adapter — turn an arbitrary source ultrasound dataset into the OpenH-RF (zea) file format, with a confirmed source→zea field mapping and a verified conversion. | +| [`skills/open-h-rf-submission-eval/`](./skills/open-h-rf-submission-eval/) | Evaluate a contributor submission to the Open-H-RF initiative against the RFP and submission guide. Produces a structured acceptance report. | +| [`skills/open-h-rf-convert/`](./skills/open-h-rf-convert/) | Source-agnostic dataset conversion adapter — turn an arbitrary source ultrasound dataset into the Open-H-RF (zea) file format, with a confirmed source→zea field mapping and a verified conversion. | -[`skills/openh-rf-shared/`](./skills/openh-rf-shared/) is a shared, +[`skills/open-h-rf-shared/`](./skills/open-h-rf-shared/) is a shared, non-invocable support directory (no `SKILL.md`, so no discovery symlink) holding files read by both skills above — the zea-format notes, the data-card template, the B-mode pipeline template, and the authoritative `validate_zea_spec.py` -checker. Skills reference it by relative path (`../openh-rf-shared/…`). +checker. Skills reference it by relative path (`../open-h-rf-shared/…`). ### Invocation per tool diff --git a/README.md b/README.md index 8698efa..7828758 100644 --- a/README.md +++ b/README.md @@ -1,13 +1,13 @@

- OpenH-RF + Open-H-RF

An open, large-scale, community-driven dataset of medical ultrasound raw channel data

-[![Website](assets/badge-website.svg)](https://open-h.github.io/OpenH-RF/) -[![Hugging Face Dataset](https://img.shields.io/badge/%F0%9F%A4%97%20Dataset-OpenH--RF-FFD21E?style=for-the-badge)](https://huggingface.co/datasets/nvidia/OpenH-RF) +[![Website](assets/badge-website.svg)](https://open-h.github.io/Open-H-RF/) +[![Hugging Face Dataset](https://img.shields.io/badge/%F0%9F%A4%97%20Dataset-Open--H--RF-FFD21E?style=for-the-badge)](https://huggingface.co/datasets/nvidia/Open-H-RF) [![arXiv](https://img.shields.io/badge/arXiv-2610.13616-B31B1B?style=for-the-badge&logo=arxiv&logoColor=white)](https://arxiv.org/abs/2610.13616) [![Discord](https://img.shields.io/badge/Discord-Join%20Community-7289DA?style=for-the-badge&logo=discord&logoColor=white)](https://discord.gg/gNTJeUsH2B) [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg?style=for-the-badge)](LICENSE) @@ -16,24 +16,24 @@ ## About -OpenH-RF is a collaborative initiative led by [Stanford University](https://med.stanford.edu/ultrasound.html), [Eindhoven University of Technology (TU/e)](https://www.tue.nl/en/research/research-groups/signal-processing-systems/biomedical-diagnostics) and [NVIDIA](https://www.nvidia.com/) to build a large-scale, openly licensed dataset of pre-beamformed (channel capture) medical ultrasound measurements. *The goal:* train general-purpose foundation models capable of multi-task *raw-to-insight* inference across echocardiography, general, fetal and transcranial imaging, blood flow measurement and ultrasound inverse problems. +Open-H-RF is a collaborative initiative led by [Stanford University](https://med.stanford.edu/ultrasound.html), [Eindhoven University of Technology (TU/e)](https://www.tue.nl/en/research/research-groups/signal-processing-systems/biomedical-diagnostics) and [NVIDIA](https://www.nvidia.com/) to build a large-scale, openly licensed dataset of pre-beamformed (channel capture) medical ultrasound measurements. *The goal:* train general-purpose foundation models capable of multi-task *raw-to-insight* inference across echocardiography, general, fetal and transcranial imaging, blood flow measurement and ultrasound inverse problems. The initial (v1.0) release of the dataset contains **19,336** real and synthetic channel capture measurements, spanning a wide variety of acquisition types, targets and downstream tasks, including reconstruction, flow, quantitative imaging, motion estimation and interpretation tasks, contributed by **38 institutions** across academia and industry. Released under [**CC BY 4.0**](https://creativecommons.org/licenses/by/4.0/). -The OpenH-RF format is implemented using the [`zea`](https://zea.readthedocs.io/en/stable/getting-started.html) ultrasound toolbox. Each file holds the raw per-element channel data together with the full transmit and receive geometry needed to beamform it. Where available, files also include derived images, maps, labels and auxiliary signals that help interpret the data. +The Open-H-RF format is implemented using the [`zea`](https://zea.readthedocs.io/en/stable/getting-started.html) ultrasound toolbox. Each file holds the raw per-element channel data together with the full transmit and receive geometry needed to beamform it. Where available, files also include derived images, maps, labels and auxiliary signals that help interpret the data. > [!TIP] -> Explore the dataset on our [catalog page](https://open-h.github.io/OpenH-RF/) and filter by imaging target, acquisition type, downstream task, or contributor. +> Explore the dataset on our [catalog page](https://open-h.github.io/Open-H-RF/) and filter by imaging target, acquisition type, downstream task, or contributor. ## 📣 Updates -- 📄 **October 2026:** [The OpenH-RF preprint is out on arXiv!](https://arxiv.org/abs/2610.13616) -- 🤗 **October 2026:** [OpenH-RF v1.0 is publicly released on Hugging Face!](https://huggingface.co/datasets/nvidia/OpenH-RF) -- 🚀 **September 2026:** [OpenH-RF v1.0 early-access pre-release is live on Hugging Face!](https://huggingface.co/datasets/nvidia/OpenH-RF) +- 📄 **October 2026:** [The Open-H-RF preprint is out on arXiv!](https://arxiv.org/abs/2610.13616) +- 🤗 **October 2026:** [Open-H-RF v1.0 is publicly released on Hugging Face!](https://huggingface.co/datasets/nvidia/Open-H-RF) +- 🚀 **September 2026:** [Open-H-RF v1.0 early-access pre-release is live on Hugging Face!](https://huggingface.co/datasets/nvidia/Open-H-RF) ## Get the Data -The dataset is on Hugging Face at [nvidia/OpenH-RF](https://huggingface.co/datasets/nvidia/OpenH-RF). The files can be streamed using [`zea`](https://zea.readthedocs.io/en/stable/getting-started.html). +The dataset is on Hugging Face at [nvidia/Open-H-RF](https://huggingface.co/datasets/nvidia/Open-H-RF). The files can be streamed using [`zea`](https://zea.readthedocs.io/en/stable/getting-started.html). You can use `zea app` to inspect and preview the data in each dataset. For GPU-enabled installation instructions, [see here](https://zea.readthedocs.io/en/stable/installation.html). For a minimal install: @@ -42,12 +42,12 @@ pip install "zea[jax,app]" zea app ``` -Taking the [`tue-carotid`](https://open-h.github.io/OpenH-RF/#dataset/tue-carotid) dataset as an example, paste `hf://nvidia/OpenH-RF/tue-carotid` as the dataset and `hf://nvidia/OpenH-RF/tue-carotid/pipeline.yaml` as the config, pick a file to see its contents, and press Run to beamform it. You can also stream the data directly in Python: +Taking the [`tue-carotid`](https://open-h.github.io/Open-H-RF/#dataset/tue-carotid) dataset as an example, paste `hf://nvidia/Open-H-RF/tue-carotid` as the dataset and `hf://nvidia/Open-H-RF/tue-carotid/pipeline.yaml` as the config, pick a file to see its contents, and press Run to beamform it. You can also stream the data directly in Python: ```python import zea -with zea.Dataset("hf://nvidia/OpenH-RF/tue-carotid") as dataset: +with zea.Dataset("hf://nvidia/Open-H-RF/tue-carotid") as dataset: print(dataset.file_paths[:3]) # print the first 3 filenames file = dataset[0] # streamed (only the bytes you read are downloaded) @@ -58,24 +58,24 @@ with zea.Dataset("hf://nvidia/OpenH-RF/tue-carotid") as dataset: Alternatively, copy the whole `tue-carotid` subset to disk: ```bash -zea data copy hf://nvidia/OpenH-RF/tue-carotid ./OpenH-RF/tue-carotid --key all +zea data copy hf://nvidia/Open-H-RF/tue-carotid ./Open-H-RF/tue-carotid --key all ``` -To get only the data you need, use the [dataset explorer](https://open-h.github.io/OpenH-RF/#explorer) on the website. Filter the datasets by acquisition type, imaging target, included metadata keys and more, then click **Generate download script** to get a script that fetches just the matching files. +To get only the data you need, use the [dataset explorer](https://open-h.github.io/Open-H-RF/#explorer) on the website. Filter the datasets by acquisition type, imaging target, included metadata keys and more, then click **Generate download script** to get a script that fetches just the matching files. ## How to Participate > [!NOTE] -> Contributions to OpenH-RF v1.0 are now closed. We are planning a v2.0 release in the future. If you are interested, please [get in touch](#contact) and [join our Discord](https://discord.gg/gNTJeUsH2B) to stay updated. +> Contributions to Open-H-RF v1.0 are now closed. We are planning a v2.0 release in the future. If you are interested, please [get in touch](#contact) and [join our Discord](https://discord.gg/gNTJeUsH2B) to stay updated.
-OpenH-RF v1.0 participation instructions +Open-H-RF v1.0 participation instructions
-1. **Review the RFP:** Read the [Request for Proposals](assets/OpenH-RF%20Request%20for%20Proposals%20(RFP).pdf) for technical scope, eligibility and evaluation criteria. +1. **Review the RFP:** Read the [Request for Proposals](assets/Open-H-RF%20Request%20for%20Proposals%20(RFP).pdf) for technical scope, eligibility and evaluation criteria. 2. **Submit a Proposal:** Prepare a concise proposal (≤ 5 pages) describing your dataset, collection methodology and target tasks. Submit to [this Google Form](https://forms.gle/tqiqYSnnar1AekB19). -3. **Contribute Data:** Once approved, prepare your dataset in the OpenH-RF format, implemented in [`zea`](https://github.com/tue-bmd/zea) as documented [here](https://zea.readthedocs.io/en/stable/data-acquisition.html), along with a data card specifying the CC BY 4.0 license. Approved contributors are given a dedicated shared storage location (S3 or Google Drive) for delivery and a Discord channel for coordination. +3. **Contribute Data:** Once approved, prepare your dataset in the Open-H-RF format, implemented in [`zea`](https://github.com/tue-bmd/zea) as documented [here](https://zea.readthedocs.io/en/stable/data-acquisition.html), along with a data card specifying the CC BY 4.0 license. Approved contributors are given a dedicated shared storage location (S3 or Google Drive) for delivery and a Discord channel for coordination. 4. **Co-author the Release:** Approved contributions are included in the public dataset and foundation model release. Contributors are named co-authors in related publications upon project completion.
@@ -102,7 +102,7 @@ To get only the data you need, use the [dataset explorer](https://open-h.github. ## Dataset Format -OpenH-RF uses the file format of the [`zea`](https://github.com/tue-bmd/zea) ultrasound toolbox, described in its [data specification](https://zea.readthedocs.io/en/stable/data-acquisition.html). Every dataset on the [OpenH-RF Hugging Face page](https://huggingface.co/datasets/nvidia/OpenH-RF) follows the layout below, so any of them is a good example to copy. +Open-H-RF uses the file format of the [`zea`](https://github.com/tue-bmd/zea) ultrasound toolbox, described in its [data specification](https://zea.readthedocs.io/en/stable/data-acquisition.html). Every dataset on the [Open-H-RF Hugging Face page](https://huggingface.co/datasets/nvidia/Open-H-RF) follows the layout below, so any of them is a good example to copy. Each submission contains: @@ -121,17 +121,17 @@ If a submission bundles several sub-datasets with different acquisition types or
-The [`examples/templates/`](examples/templates/) directories are starting points for a submission. Each has a `convert.py` (source data to OpenH-RF `.hdf5`), a `reconstruct.py` and a `pipeline.yaml`. +The [`examples/templates/`](examples/templates/) directories are starting points for a submission. Each has a `convert.py` (source data to Open-H-RF `.hdf5`), a `reconstruct.py` and a `pipeline.yaml`. | Example | Modality | |---------|----------| | [`echocardiography template`](examples/templates/echocardiography/convert.py) | Cardiac (phased array): focused transmits, ECG + clinical metadata, annotations | | [`color-doppler template`](examples/templates/color-doppler/convert.py) | Linear array (plane-wave): B-mode + color Doppler velocity map, ECG, annotations | | [`segmentation template`](examples/templates/segmentation/convert.py) | Raw RF data with per-frame segmentation masks and view labels | -| [`verasonics template`](examples/templates/verasonics/convert.py) | Converting a real Verasonics `.mat` workspace to OpenH-RF | -| [`nv-raw2insights-us/`](examples/nv-raw2insights-us/) | Worked example on a real public dataset: stream a sample from [NV-Raw2Insights-US](https://huggingface.co/datasets/nvidia/NV-Raw2Insights-US), convert to OpenH-RF, and beamform raw channel data back to a B-mode | -| [`pala-ulm-ratbrain/`](examples/pala-ulm-ratbrain/) | Converting [PALA](https://zenodo.org/records/7883227) rat-brain plane-wave channel data from Zenodo to OpenH-RF format and beamforming to a B-mode | -| [`tracked-cirs-phantom/`](examples/tracked-cirs-phantom/) | Converting a tracked CIRS phantom acquisition to OpenH-RF, reconstructing a B-mode, and visualizing the tracked probe trajectory | +| [`verasonics template`](examples/templates/verasonics/convert.py) | Converting a real Verasonics `.mat` workspace to Open-H-RF | +| [`nv-raw2insights-us/`](examples/nv-raw2insights-us/) | Worked example on a real public dataset: stream a sample from [NV-Raw2Insights-US](https://huggingface.co/datasets/nvidia/NV-Raw2Insights-US), convert to Open-H-RF, and beamform raw channel data back to a B-mode | +| [`pala-ulm-ratbrain/`](examples/pala-ulm-ratbrain/) | Converting [PALA](https://zenodo.org/records/7883227) rat-brain plane-wave channel data from Zenodo to Open-H-RF format and beamforming to a B-mode | +| [`tracked-cirs-phantom/`](examples/tracked-cirs-phantom/) | Converting a tracked CIRS phantom acquisition to Open-H-RF, reconstructing a B-mode, and visualizing the tracked probe trajectory | | [`examples/reconstruct_example.py`](examples/reconstruct_example.py) | Reconstructing a B-mode image from raw RF channel data using the default `zea` pipeline | | [`examples/save_pipeline_example.py`](examples/save_pipeline_example.py) | Saving a `zea` processing pipeline as a reusable YAML config | @@ -142,8 +142,8 @@ The [`examples/templates/`](examples/templates/) directories are starting points This repo uses [`uv`](https://docs.astral.sh/uv/) for environment + dependency management. Install it once: `curl -LsSf https://astral.sh/uv/install.sh | sh`. ```bash -git clone https://github.com/open-h/OpenH-RF -cd OpenH-RF +git clone https://github.com/open-h/Open-H-RF +cd Open-H-RF uv sync export KERAS_BACKEND=jax uv run python examples/save_pipeline_example.py @@ -170,11 +170,11 @@ Set the matching `KERAS_BACKEND` (`jax`, `torch`, or `tensorflow`) before runnin ## Citation -If you use OpenH-RF in your research, please cite the [OpenH-RF paper](https://arxiv.org/abs/2610.13616): +If you use Open-H-RF in your research, please cite the [Open-H-RF paper](https://arxiv.org/abs/2610.13616): ```bibtex @misc{stevens2026openhrf, - title = {{OpenH-RF}: A Large-Scale Ultrasound Raw Channel Dataset Initiative}, + title = {{Open-H-RF}: A Large-Scale Ultrasound Raw Channel Dataset Initiative}, author = {T. S. W. Stevens and S. W. Penninga and W. L. van Nierop and O. Nolan and F. Duelmer and B. N. Frey and B. Federici and V. van Hal and L. D. van Harten and H. de Hoop and R. Lopata and B. Luijten and M. van Sambeek and V. van der Schaft and H.-M. Schwab and @@ -238,14 +238,14 @@ If you use [`zea`](https://github.com/tue-bmd/zea) to access or process the data } ``` -If you use a specific sub-dataset, please also check whether it has its own citation. Open the [dataset explorer](https://open-h.github.io/OpenH-RF/#explorer), click the dataset's card and see its Citation section. +If you use a specific sub-dataset, please also check whether it has its own citation. Open the [dataset explorer](https://open-h.github.io/Open-H-RF/#explorer), click the dataset's card and see its Citation section. ## Contact -- **Technical questions:** [openh.data+rf@gmail.com](mailto:openh.data+rf@gmail.com), the [Hugging Face dataset community page](https://huggingface.co/datasets/nvidia/OpenH-RF/discussions), or the [`zea` community page](https://github.com/tue-bmd/zea/discussions) +- **Technical questions:** [openh.data+rf@gmail.com](mailto:openh.data+rf@gmail.com), the [Hugging Face dataset community page](https://huggingface.co/datasets/nvidia/Open-H-RF/discussions), or the [`zea` community page](https://github.com/tue-bmd/zea/discussions) - **Administrative questions:** [wsimson@nvidia.com](mailto:wsimson@nvidia.com) - **Community:** [Join our Discord](https://discord.gg/gNTJeUsH2B) ## License -Code in this repository is licensed under the [Apache License 2.0](LICENSE). The OpenH-RF v1.0 dataset is licensed under [CC BY 4.0](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/v1.0/LICENSE). +Code in this repository is licensed under the [Apache License 2.0](LICENSE). The Open-H-RF v1.0 dataset is licensed under [CC BY 4.0](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/v1.0/LICENSE). diff --git a/assets/OpenH-RF Request for Proposals (RFP).pdf b/assets/Open-H-RF Request for Proposals (RFP).pdf similarity index 100% rename from assets/OpenH-RF Request for Proposals (RFP).pdf rename to assets/Open-H-RF Request for Proposals (RFP).pdf diff --git a/assets/badge-website.svg b/assets/badge-website.svg index a0a5748..c0e278a 100644 --- a/assets/badge-website.svg +++ b/assets/badge-website.svg @@ -1 +1 @@ -WEBSITE: OPENH-RFWEBSITEOPENH-RF \ No newline at end of file +WEBSITE: OPEN-H-RFWEBSITEOPEN-H-RF \ No newline at end of file diff --git a/assets/openh-rf-header.jpg b/assets/openh-rf-header.jpg deleted file mode 100644 index 1e7ec40..0000000 Binary files a/assets/openh-rf-header.jpg and /dev/null differ diff --git a/datasets/README.md b/datasets/README.md index 7105346..50ba203 100644 --- a/datasets/README.md +++ b/datasets/README.md @@ -1,6 +1,6 @@ -# OpenH-RF — reconstructions +# Open-H-RF — reconstructions -The code counterpart to [**OpenH-RF**](https://huggingface.co/datasets/nvidia/OpenH-RF) on the Hugging Face Hub. The data lives there; the reconstructions scripts are here. +The code counterpart to [**Open-H-RF**](https://huggingface.co/datasets/nvidia/Open-H-RF) on the Hugging Face Hub. The data lives there; the reconstructions scripts are here. Each directory consists of: diff --git a/datasets/colorado-boulder/README.md b/datasets/colorado-boulder/README.md index 42b5aeb..09ae573 100644 --- a/datasets/colorado-boulder/README.md +++ b/datasets/colorado-boulder/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - 3d language: - en @@ -20,7 +20,7 @@ size_categories: | 2D ATS 539 phantom | 3D phantom | In-vivo quadriceps | |:---:|:---:|:---:| | ![Tracked SSA sweep through the 2D ATS 539 phantom](assets/ssa-sweep-phantom2d.gif) | ![Tracked SSA sweep through the 3D phantom](assets/ssa-sweep-phantom3d.gif) | ![Tracked SSA sweep along in-vivo quadriceps muscle](assets/ssa-sweep-invivo.gif) | -| [`Sub-dataset-1`](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/colorado-boulder/Sub-dataset-1) | [`Sub-dataset-2`](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/colorado-boulder/Sub-dataset-2) | [`Sub-dataset-3`](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/colorado-boulder/Sub-dataset-3) | +| [`Sub-dataset-1`](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/colorado-boulder/Sub-dataset-1) | [`Sub-dataset-2`](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/colorado-boulder/Sub-dataset-2) | [`Sub-dataset-3`](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/colorado-boulder/Sub-dataset-3) | *Motion-compensated tracked SSA reconstructions, one per sub-dataset. Each frame beamforms the raw RF channel data with its own tracked probe pose; a 40 mm window of frames is then coherently summed to synthesise a larger effective aperture, and the window slides along the freehand sweep.* @@ -30,7 +30,7 @@ This dataset contains tracked swept synthetic aperture (SSA) ultrasound acquisit The data were acquired using a Verasonics Vantage research ultrasound system with a P4-2 phased array transducer. The dataset includes raw RF channel data, acquisition parameters, probe geometry, transmit information, and frame-wise tracked probe pose metadata. -The purpose of this dataset is to provide reproducible examples for generalized ultrasound reconstruction, with a particular focus on motion-compensated tracked SSA reconstruction using the zea/OpenH-RF data format. +The purpose of this dataset is to provide reproducible examples for generalized ultrasound reconstruction, with a particular focus on motion-compensated tracked SSA reconstruction using the zea/Open-H-RF data format. The phantom acquisitions do not contain human subject data, animal data, or protected health information (PHI). The in-vivo acquisitions were collected from healthy volunteer participants and de-identified prior to release. @@ -88,7 +88,7 @@ None (N/A). ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/colorado-boulder/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/colorado-boulder/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. The script selects tracked frames at roughly 1 mm lateral spacing and writes the reconstruction to `ssa_bmode.png`. Point `ZEA_FILE` at any acquisition in the corpus to reconstruct it. [`assets/main_bmode.png`](./assets/main_bmode.png) was produced this way, compounding the full sweep into one image; the loops at the top of this card slide a shorter aperture window along the sweep instead. @@ -98,7 +98,7 @@ The script selects tracked frames at roughly 1 mm lateral spacing and writes the [zea v0.1.7](https://github.com/tue-bmd/zea) -The dataset is distributed in the zea/OpenH-RF HDF5 format. +The dataset is distributed in the zea/Open-H-RF HDF5 format. Each file includes: diff --git a/datasets/colorado-boulder/reconstruct.py b/datasets/colorado-boulder/reconstruct.py index f2e693e..34f027e 100644 --- a/datasets/colorado-boulder/reconstruct.py +++ b/datasets/colorado-boulder/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the colorado-boulder dataset of OpenH-RF. +"""Example reconstruction script for the colorado-boulder dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/colorado-boulder +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/colorado-boulder B-mode reconstruction of tracked swept synthetic aperture phantom channel data. @@ -51,10 +51,10 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/colorado-boulder/Sub-dataset-1/sub-phantom2d_01_tracked-ssa.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/colorado-boulder/Sub-dataset-1/sub-phantom2d_01_tracked-ssa.hdf5" CONFIG = HERE / "pipeline.yaml" # written by build_pipeline() below; this is what the run loads OUT = HERE / "ssa_bmode.png" -HF_CONFIG = "hf://nvidia/OpenH-RF/colorado-boulder/pipeline.yaml" # where CONFIG is published +HF_CONFIG = "hf://nvidia/Open-H-RF/colorado-boulder/pipeline.yaml" # where CONFIG is published FRAME_SPACING = 1e-3 # select tracked frames every 1 mm along the sweep diff --git a/datasets/concordia/README.md b/datasets/concordia/README.md index e53b6e2..ec7919c 100644 --- a/datasets/concordia/README.md +++ b/datasets/concordia/README.md @@ -8,7 +8,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - synthetic - field-ii language: @@ -21,11 +21,11 @@ size_categories: ![B-mode reconstructions of the five SynthUS-FSA phantom classes](assets/classes.png) -*The five phantom classes (anechoic, hypoechoic, hyperechoic, diverse, point-target), one capture per column: [`image_0012`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/concordia/data/image_0012.hdf5), [`image_0410`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/concordia/data/image_0410.hdf5), [`image_0733`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/concordia/data/image_0733.hdf5), [`image_0753`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/concordia/data/image_0753.hdf5), [`image_1808`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/concordia/data/image_1808.hdf5). Rows show the reference B-mode (`data/image`), the simulated scatterer cloud (`data/scatterers`) and the label map (`data/segmentation` for the lesion classes, `data/diverse_source_image` for diverse; point-target captures store none).* +*The five phantom classes (anechoic, hypoechoic, hyperechoic, diverse, point-target), one capture per column: [`image_0012`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/concordia/data/image_0012.hdf5), [`image_0410`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/concordia/data/image_0410.hdf5), [`image_0733`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/concordia/data/image_0733.hdf5), [`image_0753`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/concordia/data/image_0753.hdf5), [`image_1808`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/concordia/data/image_1808.hdf5). Rows show the reference B-mode (`data/image`), the simulated scatterer cloud (`data/scatterers`) and the label map (`data/segmentation` for the lesion classes, `data/diverse_source_image` for diverse; point-target captures store none).* ## Dataset Description -SynthUS-FSA is a fully synthetic corpus of pre-beamformed full-synthetic-aperture (FSA) ultrasound channel-data captures, generated with the Field II simulator to model a 128-element L11-5v linear array. Each capture fires one transmit event per element (128 transmits, receive on all 128 elements), so every element-to-element combination is retained and any receive/transmit beamforming scheme (focused B-mode, multi-angle plane-wave compounding, diverging-wave, adaptive/compressive schemes) can be retrospectively synthesized from the same channel data. The dataset is packaged in [zea](https://zea.readthedocs.io) format — the OpenH-RF reference Python library for ultrasound file I/O and beamforming — which every `.hdf5` file and `reconstruct.py` depend on. The dataset targets three OpenH-RF task categories: (1) generalized reconstruction (FSA channel data paired with a synthesized full synthetic-aperture B-mode as ground truth), (2) compressed sensing / adaptive transmit (retrospective sub-selection of the 128 transmit events — derived from the same captures, no additional files), and (3) segmentation (binary echogenicity masks on a subset of phantoms). All data is simulated — no clinical, phantom-hardware, or animal acquisition is involved. +SynthUS-FSA is a fully synthetic corpus of pre-beamformed full-synthetic-aperture (FSA) ultrasound channel-data captures, generated with the Field II simulator to model a 128-element L11-5v linear array. Each capture fires one transmit event per element (128 transmits, receive on all 128 elements), so every element-to-element combination is retained and any receive/transmit beamforming scheme (focused B-mode, multi-angle plane-wave compounding, diverging-wave, adaptive/compressive schemes) can be retrospectively synthesized from the same channel data. The dataset is packaged in [zea](https://zea.readthedocs.io) format — the Open-H-RF reference Python library for ultrasound file I/O and beamforming — which every `.hdf5` file and `reconstruct.py` depend on. The dataset targets three Open-H-RF task categories: (1) generalized reconstruction (FSA channel data paired with a synthesized full synthetic-aperture B-mode as ground truth), (2) compressed sensing / adaptive transmit (retrospective sub-selection of the 128 transmit events — derived from the same captures, no additional files), and (3) segmentation (binary echogenicity masks on a subset of phantoms). All data is simulated — no clinical, phantom-hardware, or animal acquisition is involved. A similar data-generation approach was used in Sharifzadeh et al. (2024). Users of this dataset are kindly requested to cite that work: @@ -82,13 +82,13 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/concordia/data/image_0005.hdf5 \ - --config hf://nvidia/OpenH-RF/concordia/pipeline.yaml \ + --dataset hf://nvidia/Open-H-RF/concordia/data/image_0005.hdf5 \ + --config hf://nvidia/Open-H-RF/concordia/pipeline.yaml \ --n-frames 1 \ --save-as png ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/concordia/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/concordia/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -107,7 +107,7 @@ Pre-processing applied before packaging: anti-alias FIR decimation (factor 5, 10 ## Dataset Quantification -**Current OpenH-RF release:** 2,000 HDF5 files; 82.39 GB (82,390,286,336 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 2,000 HDF5 files; 82.39 GB (82,390,286,336 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. This dataset comprises 2,000 FSA captures across five echogenicity classes: diff --git a/datasets/concordia/reconstruct.py b/datasets/concordia/reconstruct.py index faf5864..64a9343 100644 --- a/datasets/concordia/reconstruct.py +++ b/datasets/concordia/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the concordia dataset of OpenH-RF. +"""Example reconstruction script for the concordia dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/concordia +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/concordia B-mode reconstruction of full synthetic aperture (FSA) channel data, shown alongside the capture's label and its scatterer field. @@ -59,7 +59,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/concordia/data/image_0753.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/concordia/data/image_0753.hdf5" FRAME = 0 # Frame index (default: 0) OUT = HERE / "assets" / "reference_capture.png" SAVE_YAML = None # Optionally write the pipeline to a reusable pipeline.yaml diff --git a/datasets/dartmouth-uct/README.md b/datasets/dartmouth-uct/README.md index 14630c1..34052b4 100644 --- a/datasets/dartmouth-uct/README.md +++ b/datasets/dartmouth-uct/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - usct - sound-speed-estimation - attenuation @@ -23,7 +23,7 @@ size_categories: ![Ring-array USCT reflectivity reconstruction of a digital breast phantom, next to its ground-truth sound-speed and attenuation maps](assets/phantom_179604449_z200.png) -*A 2D-sim breast cross-section, [`data/2d/phantom_179604449_z200.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/dartmouth-uct/data/2d/phantom_179604449_z200.hdf5), reconstructed as a reflectivity image beside the ground-truth sound-speed and attenuation maps stored in the file.* +*A 2D-sim breast cross-section, [`data/2d/phantom_179604449_z200.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/dartmouth-uct/data/2d/phantom_179604449_z200.hdf5), reconstructed as a reflectivity image beside the ground-truth sound-speed and attenuation maps stored in the file.* ## Dataset Description @@ -96,7 +96,7 @@ Both collections are generated from the **same underlying digital breast phantom ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/dartmouth-uct/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/dartmouth-uct/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. Set `ZEA_FILE` at the top of the script and run `python reconstruct.py`; the figure at the top of this card is its output. `SOS_MAP = True` swaps the constant-sound-speed delays for a straight-ray integral through the ground-truth map. @@ -115,7 +115,7 @@ All data are packaged in the **`zea` HDF5 file format** (one `.hdf5` file per ac ## Dataset Quantification -**Current OpenH-RF release:** 2,625 HDF5 files; 257.62 GB (257,619,787,776 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 2,625 HDF5 files; 257.62 GB (257,619,787,776 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** **2,149 (2D-sim)** (1,859 dense + 290 fatty) + **476 (3D-sim)** = **2,625 acquisitions**. - **Single-Tx channel-capture frames:** 2,149 × 256 (2D-sim) + 476 × 64 (3D-sim) ≈ **5.8 × 10⁵ frames**. diff --git a/datasets/dartmouth-uct/reconstruct.py b/datasets/dartmouth-uct/reconstruct.py index d832681..edd6276 100644 --- a/datasets/dartmouth-uct/reconstruct.py +++ b/datasets/dartmouth-uct/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the dartmouth-uct dataset of OpenH-RF. +"""Example reconstruction script for the dartmouth-uct dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/dartmouth-uct +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/dartmouth-uct Delay-and-sum reflectivity reconstruction of ring-array USCT channel data. @@ -64,7 +64,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/dartmouth-uct/data/2d/phantom_179604449_z200.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/dartmouth-uct/data/2d/phantom_179604449_z200.hdf5" FOV = None # square field of view [m] (default: the ground-truth map footprint) NUM_PIXELS = None # output image is num_pixels x num_pixels (default: ground-truth resolution) SOS_MAP = False # use the ground-truth sound-speed map for straight-ray corrected delays diff --git a/datasets/kaist-snubh-barreleye/README.md b/datasets/kaist-snubh-barreleye/README.md index 47e75b7..63a93b4 100644 --- a/datasets/kaist-snubh-barreleye/README.md +++ b/datasets/kaist-snubh-barreleye/README.md @@ -8,7 +8,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - breast - in-vivo - plane-wave @@ -23,7 +23,7 @@ size_categories: ![DAS B-mode reconstruction of a biopsy-proven invasive ductal carcinoma (S01_D1)](assets/main.png) -*Delay-and-sum reconstruction of a biopsy-proven invasive ductal carcinoma, [`data/S01_D1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/kaist-snubh-barreleye/data/S01_D1.hdf5).* +*Delay-and-sum reconstruction of a biopsy-proven invasive ductal carcinoma, [`data/S01_D1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/kaist-snubh-barreleye/data/S01_D1.hdf5).* ## Dataset Description @@ -70,13 +70,13 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/kaist-snubh-barreleye/data/S01_D1.hdf5 \ - --config hf://nvidia/OpenH-RF/kaist-snubh-barreleye/pipeline.yaml \ + --dataset hf://nvidia/Open-H-RF/kaist-snubh-barreleye/data/S01_D1.hdf5 \ + --config hf://nvidia/Open-H-RF/kaist-snubh-barreleye/pipeline.yaml \ --n-frames 1 \ --save-as png ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/kaist-snubh-barreleye/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/kaist-snubh-barreleye/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). This is a single-frame acquisition, so `zea process` outputs a `.png` rather than a `.gif` — this requires a `zea` build newer than the currently pinned 0.1.6 (single-frame PNG output landed after that release). @@ -84,7 +84,7 @@ This is a single-frame acquisition, so `zea process` outputs a `.png` rather tha [zea v0.1.6](https://github.com/tue-bmd/zea) -All data is delivered in the **zea HDF5** format (OpenH-RF spec). One HDF5 file per acquisition; one image track per file. +All data is delivered in the **zea HDF5** format (Open-H-RF spec). One HDF5 file per acquisition; one image track per file. ``` hdf5/ @@ -94,12 +94,12 @@ hdf5/ ``` **`original/` — raw RF channel data.** -- **Channel reordering** of the scanner's raw export into the OpenH-RF convention `(n_frames=1, n_tx=9, n_ax=Ns, n_el=192, n_ch=1)`. +- **Channel reordering** of the scanner's raw export into the Open-H-RF convention `(n_frames=1, n_tx=9, n_ax=Ns, n_el=192, n_ch=1)`. - No demodulation, decimation, band-pass filtering, or value clipping — `raw_data` is bit-faithful to the scanner export. ## Dataset Quantification -**Current OpenH-RF release:** 70 HDF5 files; 802.10 MB (802,095,104 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 70 HDF5 files; 802.10 MB (802,095,104 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Subjects:** 35 (S01–S37 with S11 and S29 unused, all female; age groups in *Subject Metadata*). - **Acquisitions:** 70 (2 repeat acquisitions per subject, named `_D1` / `_D2`). diff --git a/datasets/kaist-snubh-barreleye/reconstruct.py b/datasets/kaist-snubh-barreleye/reconstruct.py index f45a392..3c29e41 100644 --- a/datasets/kaist-snubh-barreleye/reconstruct.py +++ b/datasets/kaist-snubh-barreleye/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the kaist-snubh-barreleye dataset of OpenH-RF. +"""Example reconstruction script for the kaist-snubh-barreleye dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/kaist-snubh-barreleye +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/kaist-snubh-barreleye B-mode reconstruction of 9-angle plane-wave compounded in-vivo breast channel data. The 1-12 MHz band-pass rejects a persistent sub-MHz band before coherent @@ -35,8 +35,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/kaist-snubh-barreleye/data/S01_D1.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/kaist-snubh-barreleye/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/kaist-snubh-barreleye/data/S01_D1.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/kaist-snubh-barreleye/pipeline.yaml" OUT = HERE / "assets" / "reconstruction.png" diff --git a/datasets/mosaic-nuevosono/README.md b/datasets/mosaic-nuevosono/README.md index 43d6729..7b00682 100644 --- a/datasets/mosaic-nuevosono/README.md +++ b/datasets/mosaic-nuevosono/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - IVUS - tracked-ultrasound language: @@ -18,7 +18,7 @@ language: ![IVUS pullback: B-mode, segmentation overlay and pullback trajectory](assets/pullback.gif) -*Pullback through [`data/22_12_10_52.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/mosaic-nuevosono/data/22_12_10_52.hdf5), reconstructed from the raw channel data. Left to right: B-mode, the same frame with the lumen and intima-media segmentation, and the linear-encoder pullback position with the current frame marked.* +*Pullback through [`data/22_12_10_52.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/mosaic-nuevosono/data/22_12_10_52.hdf5), reconstructed from the raw channel data. Left to right: B-mode, the same frame with the lumen and intima-media segmentation, and the linear-encoder pullback position with the current frame marked.* ## Dataset Description @@ -52,7 +52,7 @@ Intended for IVUS tracking and segmentation applications, including lesion detec ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/mosaic-intelligence/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/mosaic-intelligence/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. The script overlays the segmentation masks on frames spread evenly across the pullback: @@ -139,6 +139,6 @@ A `zea.Pipeline` reconstructs the IVUS B-mode from the raw channel data (RF → ## Ethical Considerations -Porcine animal study data only; contains no human subjects or PHI. Collected and released in compliance with applicable institutional animal care approvals and OpenH-RF governance and data-sharing policies. +Porcine animal study data only; contains no human subjects or PHI. Collected and released in compliance with applicable institutional animal care approvals and Open-H-RF governance and data-sharing policies. All contributed data, labels and metadata are released by Mosaic Intelligence Labs and NuevoSono; pre-existing hardware, software, simulation and platform intellectual property remains the property of the respective owners. The license is also recorded in each file's `metadata/credit` field. diff --git a/datasets/mosaic-nuevosono/reconstruct.py b/datasets/mosaic-nuevosono/reconstruct.py index 1374a01..66d6784 100644 --- a/datasets/mosaic-nuevosono/reconstruct.py +++ b/datasets/mosaic-nuevosono/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the mosaic-nuevosono dataset of OpenH-RF. +"""Example reconstruction script for the mosaic-nuevosono dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/mosaic-nuevosono +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/mosaic-nuevosono B-mode reconstruction of rotational in-vivo IVUS channel data. @@ -55,8 +55,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/mosaic-nuevosono/data/22_12_10_52.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/mosaic-nuevosono/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/mosaic-nuevosono/data/22_12_10_52.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/mosaic-nuevosono/pipeline.yaml" NUM_FRAMES = 5 # Number of frames to overlay, spaced evenly across the pullback FRAMES = None # Explicit frame indices to overlay (overrides NUM_FRAMES) OUT = None diff --git a/datasets/nv-raw2insights-us/README.md b/datasets/nv-raw2insights-us/README.md index 549d6be..f3828fc 100644 --- a/datasets/nv-raw2insights-us/README.md +++ b/datasets/nv-raw2insights-us/README.md @@ -8,7 +8,7 @@ tags: - ultrasound - rf - iq - - openh-rf + - open-h-rf - beamforming - sound-speed-estimation - phase-aberration @@ -24,7 +24,7 @@ size_categories: ![DBUA B-mode reconstruction and estimated sound speed using NV-Raw2Insights-US data](assets/dbua-reconstruction.gif) -*DBUA results using synthetic validation sample [`data/nv_r2i_us_validation_0084.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/nv-raw2insights-us/data/nv_r2i_us_validation_0084.hdf5): B-mode (left) and estimated sound speed (right). Bulk-speed calibration is followed by 400 spatial-refinement iterations, with fixed display scales. These are DBUA reconstructions, not predictions from an NV-Raw2Insights-US model.* +*DBUA results using synthetic validation sample [`data/nv_r2i_us_validation_0084.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/nv-raw2insights-us/data/nv_r2i_us_validation_0084.hdf5): B-mode (left) and estimated sound speed (right). Bulk-speed calibration is followed by 400 spatial-refinement iterations, with fixed display scales. These are DBUA reconstructions, not predictions from an NV-Raw2Insights-US model.* @@ -34,7 +34,7 @@ NV-Raw2Insights-US is a **simulated full-synthetic-aperture (FSA)** ultrasound d Each sample is a single-frame FSA acquisition from a **180-element linear array** simulated with **k-Wave** over a heterogeneous tissue phantom containing cysts. Every acquisition provides raw baseband **IQ channel data** alongside co-registered ground truth: a **speed-of-sound map**, a **binary cyst segmentation mask**, a **DAS B-mode** and a **focused-transmit B-mode**, and a per-sample **phase-aberration** value. All data are **synthetic**; no human or animal subjects are involved. -This is the **zea-format (OpenH-RF) release**; the same simulations are also published in a Hugging Face `datasets`/Arrow build at [`nvidia/NV-Raw2Insights-US`](https://huggingface.co/datasets/nvidia/NV-Raw2Insights-US). +This is the **zea-format (Open-H-RF) release**; the same simulations are also published in a Hugging Face `datasets`/Arrow build at [`nvidia/NV-Raw2Insights-US`](https://huggingface.co/datasets/nvidia/NV-Raw2Insights-US). ## Dataset Contributor(s) @@ -60,7 +60,7 @@ Developing and benchmarking methods that operate on raw ultrasound channel data: ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/nv-raw2insights-us/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), which uses the [zea library](https://github.com/tue-bmd/zea) and streams the data from the Hugging Face Hub. It reconstructs both a plain DAS B-mode (the `pipeline.yaml` in this folder) and a sound-speed-corrected B-mode using the ground-truth `sos_map`. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/nv-raw2insights-us/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), which uses the [zea library](https://github.com/tue-bmd/zea) and streams the data from the Hugging Face Hub. It reconstructs both a plain DAS B-mode (the `pipeline.yaml` in this folder) and a sound-speed-corrected B-mode using the ground-truth `sos_map`. ## Dataset Format @@ -87,7 +87,7 @@ Per-file contents: ## Dataset Quantification -**Current OpenH-RF release:** 923 HDF5 files; 214.13 GB (214,133,178,368 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 923 HDF5 files; 214.13 GB (214,133,178,368 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **923 samples (830 train / 93 validation)**, one HDF5 per sample, all in `data/`; the split is carried in the filename. - Each sample is a single-frame FSA acquisition (frame index always 0). diff --git a/datasets/nv-raw2insights-us/reconstruct.py b/datasets/nv-raw2insights-us/reconstruct.py index 53817f7..7679639 100644 --- a/datasets/nv-raw2insights-us/reconstruct.py +++ b/datasets/nv-raw2insights-us/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the nv-raw2insights-us dataset of OpenH-RF. +"""Example reconstruction script for the nv-raw2insights-us dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/nv-raw2insights-us +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/nv-raw2insights-us B-mode reconstruction of simulated full synthetic aperture (FSA) channel data. @@ -39,13 +39,13 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/nv-raw2insights-us/data/nv_r2i_us_validation_0084.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/nv-raw2insights-us/data/nv_r2i_us_validation_0084.hdf5" OUT = HERE / "assets" / "nv_raw2insights_us_reconstructed.png" CONFIG = HERE / "pipeline.yaml" def coords_to_imshow_mm(coords): - """openh-rf per-pixel coordinates (z, x, 3), last axis [x, y, z] in metres + """open-h-rf per-pixel coordinates (z, x, 3), last axis [x, y, z] in metres -> mpl imshow extent [left, right, bottom, top] in mm.""" c = np.asarray(coords) x, z = c[..., 0] * 1e3, c[..., 2] * 1e3 @@ -182,7 +182,7 @@ def main(): axes[6].set_xlabel("Lateral [mm]") axes[6].set_ylabel("Depth [mm]") - fig.suptitle(f"openh-rf sample (phase error: {phase_err[0]:.2f} rad)", fontsize=14, y=1.02) + fig.suptitle(f"open-h-rf sample (phase error: {phase_err[0]:.2f} rad)", fontsize=14, y=1.02) plt.tight_layout() Path(OUT).parent.mkdir(parents=True, exist_ok=True) plt.savefig(OUT, dpi=150, bbox_inches="tight") diff --git a/datasets/oslo/A_cardiac/README.md b/datasets/oslo/A_cardiac/README.md index 86348d2..170b0c1 100644 --- a/datasets/oslo/A_cardiac/README.md +++ b/datasets/oslo/A_cardiac/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - cardiac - in-vivo size_categories: @@ -20,7 +20,7 @@ size_categories: ![Cine loop of a parasternal long-axis view](assets/hero.gif) -*All 50 frames of [`Verasonics_P2-4_parasternal_long_subject_1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* +*All 50 frames of [`Verasonics_P2-4_parasternal_long_subject_1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* ## Dataset Description @@ -46,7 +46,7 @@ In-vivo human cardiac channel-capture data acquired with a Verasonics Vantage 25 ## Intended Usage -Generalized reconstruction and adaptive beamforming of cardiac ultrasound (RFP task 6.1). Suitable for B-mode reconstruction, aperture-domain processing, and deep-learning beamforming research on in-vivo cardiac data. (OpenH-RF RFP task 6.1 Generalized Reconstruction). +Generalized reconstruction and adaptive beamforming of cardiac ultrasound (RFP task 6.1). Suitable for B-mode reconstruction, aperture-domain processing, and deep-learning beamforming research on in-vivo cardiac data. (Open-H-RF RFP task 6.1 Generalized Reconstruction). ## Dataset Characterization @@ -62,11 +62,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/A_cardiac/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/A_cardiac/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline, display window and sector limits of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). +This `pipeline.yaml` holds the pipeline, display window and sector limits of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). ## Dataset Format @@ -76,7 +76,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 2 HDF5 files; 3.53 GB (3,533,897,728 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 2 HDF5 files; 3.53 GB (3,533,897,728 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 2 - **Total channel-capture frames:** 75 diff --git a/datasets/oslo/B_carotid/README.md b/datasets/oslo/B_carotid/README.md index da265fd..ad2ea3d 100644 --- a/datasets/oslo/B_carotid/README.md +++ b/datasets/oslo/B_carotid/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - vascular - carotid - in-vivo @@ -21,7 +21,7 @@ size_categories: ![Focused B-mode of a carotid cross-section](assets/hero.png) -*First frame of [`L7_FI_carotid_cross_1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* +*First frame of [`L7_FI_carotid_cross_1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* ## Dataset Description @@ -47,7 +47,7 @@ In-vivo human carotid-artery channel-capture data acquired with a Verasonics Van ## Intended Usage -Generalized reconstruction and adaptive beamforming of vascular ultrasound (RFP task 6.1). (OpenH-RF RFP task 6.1 Generalized Reconstruction). +Generalized reconstruction and adaptive beamforming of vascular ultrasound (RFP task 6.1). (Open-H-RF RFP task 6.1 Generalized Reconstruction). ## Dataset Characterization @@ -63,11 +63,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/B_carotid/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/B_carotid/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). +This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). ## Dataset Format @@ -77,7 +77,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 3 HDF5 files; 219.48 MB (219,480,064 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 3 HDF5 files; 219.48 MB (219,480,064 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 3 - **Total channel-capture frames:** 6 diff --git a/datasets/oslo/C_verasonics_phantom/README.md b/datasets/oslo/C_verasonics_phantom/README.md index 0a73e2c..05e199e 100644 --- a/datasets/oslo/C_verasonics_phantom/README.md +++ b/datasets/oslo/C_verasonics_phantom/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - phantom - cirs size_categories: @@ -20,7 +20,7 @@ size_categories: ![Phased-array sector B-mode of a cyst phantom](assets/hero.png) -*First frame of [`FI_P4_cysts_center.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* +*First frame of [`FI_P4_cysts_center.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* ## Dataset Description @@ -46,7 +46,7 @@ Tissue-mimicking and table-top phantom channel-capture data acquired on a Veraso ## Intended Usage -Generalized reconstruction, image-quality assessment (resolution, contrast, dynamic range), and beamforming research (RFP task 6.1). Several acquisitions provide multiple transmit schemes on the same target for cross-method comparison. (OpenH-RF RFP task 6.1 Generalized Reconstruction). +Generalized reconstruction, image-quality assessment (resolution, contrast, dynamic range), and beamforming research (RFP task 6.1). Several acquisitions provide multiple transmit schemes on the same target for cross-method comparison. (Open-H-RF RFP task 6.1 Generalized Reconstruction). ## Dataset Characterization @@ -62,11 +62,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/C_verasonics_phantom/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/C_verasonics_phantom/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline, display window and sector limits of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). +This `pipeline.yaml` holds the pipeline, display window and sector limits of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). ## Dataset Format @@ -76,7 +76,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 15 HDF5 files; 1.06 GB (1,058,734,080 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 15 HDF5 files; 1.06 GB (1,058,734,080 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 15 - **Total channel-capture frames:** 27 diff --git a/datasets/oslo/D_alpinion_phantom/README.md b/datasets/oslo/D_alpinion_phantom/README.md index 9f47e6e..5fb9fcf 100644 --- a/datasets/oslo/D_alpinion_phantom/README.md +++ b/datasets/oslo/D_alpinion_phantom/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - phantom - alpinion size_categories: @@ -20,7 +20,7 @@ size_categories: ![Plane-wave compounded B-mode of hypoechoic cysts](assets/hero.png) -*First frame of [`Alpinion_L3-8_CPWC_hypoechoic.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* +*First frame of [`Alpinion_L3-8_CPWC_hypoechoic.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* ## Dataset Description @@ -46,7 +46,7 @@ Phantom channel-capture data acquired on an Alpinion E-Cube 12R research scanner ## Intended Usage -Generalized reconstruction and image-quality assessment on a second hardware platform (RFP task 6.1); cross-vendor robustness studies. (OpenH-RF RFP task 6.1 Generalized Reconstruction). +Generalized reconstruction and image-quality assessment on a second hardware platform (RFP task 6.1); cross-vendor robustness studies. (Open-H-RF RFP task 6.1 Generalized Reconstruction). ## Dataset Characterization @@ -62,11 +62,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/D_alpinion_phantom/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/D_alpinion_phantom/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). +This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). ## Dataset Format @@ -76,7 +76,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 4 HDF5 files; 466.42 MB (466,419,712 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 4 HDF5 files; 466.42 MB (466,419,712 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 4 - **Total channel-capture frames:** 4 diff --git a/datasets/oslo/E_simulation/README.md b/datasets/oslo/E_simulation/README.md index 0d671c2..a02a0d0 100644 --- a/datasets/oslo/E_simulation/README.md +++ b/datasets/oslo/E_simulation/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - simulation - field-ii - synthetic @@ -21,7 +21,7 @@ size_categories: ![B-mode of the PICMUS numerical phantom](assets/hero.png) -*First frame of [`PICMUS_numerical_calib_v2.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* +*First frame of [`PICMUS_numerical_calib_v2.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5), reconstructed from the raw channel data with `reconstruct.py`.* ## Dataset Description @@ -47,7 +47,7 @@ Physics-based synthetic channel-capture data generated with the Field II ultraso ## Intended Usage -Generalized reconstruction, beamformer development and validation with known ground truth (RFP task 6.1); resolution/contrast/dynamic-range characterization; training/validation of learned reconstruction methods. (OpenH-RF RFP task 6.1 Generalized Reconstruction). +Generalized reconstruction, beamformer development and validation with known ground truth (RFP task 6.1); resolution/contrast/dynamic-range characterization; training/validation of learned reconstruction methods. (Open-H-RF RFP task 6.1 Generalized Reconstruction). ## Dataset Characterization @@ -63,11 +63,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/E_simulation/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/E_simulation/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). +This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). ## Dataset Format @@ -77,7 +77,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 11 HDF5 files; 2.98 GB (2,980,642,816 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 11 HDF5 files; 2.98 GB (2,980,642,816 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 11 - **Total channel-capture frames:** 17 diff --git a/datasets/oslo/F_motion/README.md b/datasets/oslo/F_motion/README.md index dd2757d..d0007f2 100644 --- a/datasets/oslo/F_motion/README.md +++ b/datasets/oslo/F_motion/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - elastography - shear-wave - arfi @@ -21,7 +21,7 @@ size_categories: ![Shear waves travelling outward after an acoustic push](assets/hero.gif) -*First push-track sequence (frames 1-49) of [`SWE_L7_type_III.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/F_motion/SWE_L7_type_III.hdf5): B-mode (left) and axial displacement between consecutive frames (right), a lag-one autocorrelation (Kasai) estimate on the IQ beamformed with `pipeline.yaml`. The displacement estimate is not part of the reference pipeline.* +*First push-track sequence (frames 1-49) of [`SWE_L7_type_III.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/F_motion/SWE_L7_type_III.hdf5): B-mode (left) and axial displacement between consecutive frames (right), a lag-one autocorrelation (Kasai) estimate on the IQ beamformed with `pipeline.yaml`. The displacement estimate is not part of the reference pipeline.* ## Dataset Description @@ -47,7 +47,7 @@ Shear-wave elastography (SWE) and acoustic-radiation-force-impulse (ARFI) push-t ## Intended Usage -Motion estimation (RFP task 6.4): shear-wave velocity estimation, ARFI displacement tracking, and high-frame-rate motion reconstruction. (OpenH-RF RFP task 6.4 Motion Estimation). +Motion estimation (RFP task 6.4): shear-wave velocity estimation, ARFI displacement tracking, and high-frame-rate motion reconstruction. (Open-H-RF RFP task 6.4 Motion Estimation). ## Dataset Characterization @@ -63,11 +63,11 @@ The example acquisition can be processed with the `pipeline.yaml` definition in ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/oslo/F_motion/SWE_L7_type_III.hdf5 \ - --config hf://nvidia/OpenH-RF/oslo/F_motion/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/oslo/F_motion/SWE_L7_type_III.hdf5 \ + --config hf://nvidia/Open-H-RF/oslo/F_motion/pipeline.yaml ``` -This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). These plane-wave tracking acquisitions use the non-focused `compound` pipeline, and the reference reconstruction shows a single tracking frame. +This `pipeline.yaml` holds the pipeline and display window of that acquisition. Alternatively, all acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py) at the root of this collection, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` definitions at the collection root and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub; `parameters.yaml` picks the pipeline, display window and dynamic range per acquisition (see the [collection card](../README.md#processing-the-dataset)). These plane-wave tracking acquisitions use the non-focused `compound` pipeline, and the reference reconstruction shows a single tracking frame. ## Dataset Format @@ -77,7 +77,7 @@ All acquisitions are stored in the **zea** HDF5 file format. Each `.hdf5` file i ## Dataset Quantification -**Current OpenH-RF release:** 4 HDF5 files; 354.42 MB (354,418,688 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 4 HDF5 files; 354.42 MB (354,418,688 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Number of acquisitions:** 4 - **Total channel-capture frames:** 800 diff --git a/datasets/oslo/README.md b/datasets/oslo/README.md index f034e93..11f433e 100644 --- a/datasets/oslo/README.md +++ b/datasets/oslo/README.md @@ -9,7 +9,7 @@ language: tags: - ultrasound - rf - - openh-rf + - open-h-rf - beamforming - channel-data size_categories: @@ -20,11 +20,11 @@ size_categories: ![B-modes from five of the six sub-datasets: in-vivo cardiac, in-vivo carotid, Verasonics and Alpinion phantoms, and simulation](assets/hero.png) -*First frames reconstructed from the raw channel data with `reconstruct.py`. Top: [`A_cardiac/Verasonics_P2-4_parasternal_long_subject_1`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5), [`A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1.hdf5), [`C_verasonics_phantom/FI_P4_cysts_center`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5). Bottom: [`B_carotid/L7_FI_carotid_cross_1`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5), [`C_verasonics_phantom/L7_FI_Verasonics_CIRS`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/C_verasonics_phantom/L7_FI_Verasonics_CIRS.hdf5), [`D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5), [`E_simulation/PICMUS_numerical_calib_v2`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5). Panels are scaled to a common height per row, not to a common physical scale.* +*First frames reconstructed from the raw channel data with `reconstruct.py`. Top: [`A_cardiac/Verasonics_P2-4_parasternal_long_subject_1`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_parasternal_long_subject_1.hdf5), [`A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1.hdf5), [`C_verasonics_phantom/FI_P4_cysts_center`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/C_verasonics_phantom/FI_P4_cysts_center.hdf5). Bottom: [`B_carotid/L7_FI_carotid_cross_1`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/B_carotid/L7_FI_carotid_cross_1.hdf5), [`C_verasonics_phantom/L7_FI_Verasonics_CIRS`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/C_verasonics_phantom/L7_FI_Verasonics_CIRS.hdf5), [`D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/D_alpinion_phantom/Alpinion_L3-8_CPWC_hypoechoic.hdf5), [`E_simulation/PICMUS_numerical_calib_v2`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/oslo/E_simulation/PICMUS_numerical_calib_v2.hdf5). Panels are scaled to a common height per row, not to a common physical scale.* ## Dataset Description -Contributed by the **University of Oslo (UiO), Department of Informatics** (the USTB team) to the [OpenH-RF](https://github.com/open-h/OpenH-RF) initiative. All data is pre-beamformed (raw channel-capture) ultrasound stored in the **zea** HDF5 format and released under **CC BY 4.0**. +Contributed by the **University of Oslo (UiO), Department of Informatics** (the USTB team) to the [Open-H-RF](https://github.com/open-h/Open-H-RF) initiative. All data is pre-beamformed (raw channel-capture) ultrasound stored in the **zea** HDF5 format and released under **CC BY 4.0**. 39 acquisitions packaged into six application sub-datasets. Each sub-dataset folder contains its zea `.hdf5` files, a Hugging Face–style `README.md` data card, and one reference B-mode PNG per acquisition. The **`reconstruct.py`, `pipeline.yaml`, and the CC BY 4.0 `LICENCE` live once at the submission root** — the reconstruction reconstructs every sub-dataset (the pipeline is identical across folders), and the single LICENCE covers the whole collection (each data card also declares `license: cc-by-4.0` in its YAML frontmatter). @@ -57,7 +57,7 @@ Total: **39 acquisitions**, 8.61 GB of stored HDF5 data. ## Processing the Dataset -Every acquisition is reconstructable from the file alone — all acquisition parameters live in the zea `/scan` and `/probe` groups. The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/oslo/reconstruct.py), as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline*.yaml` and `parameters.yaml` definitions in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub: set `PATHS` to the acquisition to reconstruct (default: the apical four-chamber cardiac scan above) and `REFOCUS = True` to also emit the REFoCUS variant. See [Reconstruction Details](#reconstruction-details) for how each acquisition's pipeline is chosen. Each sub-dataset folder also holds a `pipeline.yaml` for its example acquisition, so it can be rendered with a single `zea process` command (see the sub-dataset cards). +Every acquisition is reconstructable from the file alone — all acquisition parameters live in the zea `/scan` and `/probe` groups. The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/oslo/reconstruct.py), as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline*.yaml` and `parameters.yaml` definitions in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub: set `PATHS` to the acquisition to reconstruct (default: the apical four-chamber cardiac scan above) and `REFOCUS = True` to also emit the REFoCUS variant. See [Reconstruction Details](#reconstruction-details) for how each acquisition's pipeline is chosen. Each sub-dataset folder also holds a `pipeline.yaml` for its example acquisition, so it can be rendered with a single `zea process` command (see the sub-dataset cards). ## Dataset Format @@ -65,7 +65,7 @@ Every acquisition is reconstructable from the file alone — all acquisition par ## Dataset Quantification -**Current OpenH-RF release:** 39 HDF5 files; 8.61 GB (8,613,593,088 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 39 HDF5 files; 8.61 GB (8,613,593,088 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. ## Reconstruction Details diff --git a/datasets/oslo/reconstruct.py b/datasets/oslo/reconstruct.py index 2fe65fa..c2de06a 100644 --- a/datasets/oslo/reconstruct.py +++ b/datasets/oslo/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the oslo dataset of OpenH-RF. +"""Example reconstruction script for the oslo dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/oslo +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/oslo B-mode reconstruction for every acquisition in the USTB channel-capture collection. @@ -59,7 +59,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -PATHS = "hf://nvidia/OpenH-RF/oslo/A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1.hdf5" +PATHS = "hf://nvidia/Open-H-RF/oslo/A_cardiac/Verasonics_P2-4_apical_four_chamber_subject_1.hdf5" REFOCUS = False # Also emit the secondary REFoCUS reconstruction where flagged for it diff --git a/datasets/politorino/README.md b/datasets/politorino/README.md index 5be1344..67a1b53 100644 --- a/datasets/politorino/README.md +++ b/datasets/politorino/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - medical-imaging - beamforming - fascicle tracking @@ -21,7 +21,7 @@ size_categories: ![Reconstructed cineloop from PAT02_w1.hdf5](assets/PAT02_w1.gif) -*Cine loop of treadmill walking at 2 km/h, [`data/PAT02/PAT02_w1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/politorino/data/PAT02/PAT02_w1.hdf5), reconstructed from the raw channel data with the `pipeline.yaml` in this folder.* +*Cine loop of treadmill walking at 2 km/h, [`data/PAT02/PAT02_w1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/politorino/data/PAT02/PAT02_w1.hdf5), reconstructed from the raw channel data with the `pipeline.yaml` in this folder.* ## Dataset Description @@ -62,7 +62,7 @@ This data was employed for an initial study evaluating fascicle tracking algorit ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/politorino/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/politorino/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. The script reconstructs a B-mode from the raw channel data and overlays the stored fascicle tracking on it, writing a `.png`. The constants at the top of the script select what is drawn: @@ -89,7 +89,7 @@ Per-sample contents of the HDF5: ## Dataset Quantification -**Current OpenH-RF release:** 30 HDF5 files; 169.95 GB (169,950,314,496 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 30 HDF5 files; 169.95 GB (169,950,314,496 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - 30 files - 6 acquisitions of 9.6s at 500 fps for each of the 5 subjects - **Stored HDF5 size:** 169.95 GB (169,950,314,496 bytes); 5.67 GB per file on average. Individual file sizes vary. diff --git a/datasets/politorino/reconstruct.py b/datasets/politorino/reconstruct.py index 06dd0a3..f96d2ed 100644 --- a/datasets/politorino/reconstruct.py +++ b/datasets/politorino/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the politorino dataset of OpenH-RF. +"""Example reconstruction script for the politorino dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/politorino +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/politorino B-mode reconstruction of high-frame-rate fascicle tracking channel data, with the stored fascicle tracking overlaid: the superficial and deep aponeuroses as @@ -43,7 +43,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/politorino/data/PAT02/PAT02_w1.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/politorino/data/PAT02/PAT02_w1.hdf5" OUT = HERE / "assets" / "PAT02_w1_reconstructed.png" SAVE_PIPELINE = None # Optionally write the pipeline to a reusable pipeline.yaml FRAME = 0 diff --git a/datasets/resolvestroke/README.md b/datasets/resolvestroke/README.md index 3aa5ca8..9766374 100644 --- a/datasets/resolvestroke/README.md +++ b/datasets/resolvestroke/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - 3d - matrix-probe - diverging-wave @@ -26,11 +26,11 @@ size_categories: Transcranial contrast-enhanced power Doppler of a human subject (SP03-Left, bolus +10 s), x-z maximum-intensity projection -*3D power Doppler of the transcranial CEUS acquisition [`clinical/SP03-Left/SP03-Left.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/resolvestroke/clinical/SP03-Left/SP03-Left.hdf5) at bolus +10 s, x-z maximum-intensity projection, reconstructed from the released channel data.* +*3D power Doppler of the transcranial CEUS acquisition [`clinical/SP03-Left/SP03-Left.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/resolvestroke/clinical/SP03-Left/SP03-Left.hdf5) at bolus +10 s, x-z maximum-intensity projection, reconstructed from the released channel data.* ## Dataset Description -[Resolve Stroke](https://www.resolvestroke.com/) develops SYLVER, a software-driven ultrasound platform that supports clinical assessment in patients suspected of, or at risk for, cerebral perfusion abnormalities by improving visualization of cerebral vasculature and providing complementary information on cerebral perfusion. SYLVER images with a 32×32 matrix array probe at 2 MHz using diverging-wave transmits. CE and FDA clearances are expected in 2026. This directory holds the data Resolve Stroke contributes to OpenH-RF: pre-beamformed RF/IQ channel data from an imaging phantom, a flow phantom, and in-vivo transcranial acquisitions, all de-identified and released under CC BY 4.0. The in-vivo data comes from the CPP-approved SCULPT clinical study. Two transmit sequences are included: a saddle sequence that images a 2D plane for real-time B-mode, and a 4 kHz single-aperture volume sequence for CEUS and blood-flow measurement. +[Resolve Stroke](https://www.resolvestroke.com/) develops SYLVER, a software-driven ultrasound platform that supports clinical assessment in patients suspected of, or at risk for, cerebral perfusion abnormalities by improving visualization of cerebral vasculature and providing complementary information on cerebral perfusion. SYLVER images with a 32×32 matrix array probe at 2 MHz using diverging-wave transmits. CE and FDA clearances are expected in 2026. This directory holds the data Resolve Stroke contributes to Open-H-RF: pre-beamformed RF/IQ channel data from an imaging phantom, a flow phantom, and in-vivo transcranial acquisitions, all de-identified and released under CC BY 4.0. The in-vivo data comes from the CPP-approved SCULPT clinical study. Two transmit sequences are included: a saddle sequence that images a 2D plane for real-time B-mode, and a 4 kHz single-aperture volume sequence for CEUS and blood-flow measurement. Each dataset directory holds its data card (`README.md`), a `reconstruct.py` + `pipeline.yaml` beamforming recipe and, for the CEUS datasets, a power-Doppler reconstruction; the HDF5 files are on the Hub and the preview images in [`assets/`](assets/). @@ -53,7 +53,7 @@ Each dataset directory holds its data card (`README.md`), a `reconstruct.py` + ` ## Processing the Dataset -Each sub-dataset has its own `reconstruct.py` and `pipeline.yaml`, as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF/tree/main/datasets/resolvestroke), built on the [zea library](https://github.com/tue-bmd/zea); the CEUS datasets add a power-Doppler reconstruction. See the data card of each sub-dataset for how to run them. +Each sub-dataset has its own `reconstruct.py` and `pipeline.yaml`, as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF/tree/main/datasets/resolvestroke), built on the [zea library](https://github.com/tue-bmd/zea); the CEUS datasets add a power-Doppler reconstruction. See the data card of each sub-dataset for how to run them. The Python scripts carry their own `SPDX-License-Identifier: Apache-2.0` header; the dataset itself is CC BY 4.0. @@ -63,7 +63,7 @@ The Python scripts carry their own `SPDX-License-Identifier: Apache-2.0` header; ## Dataset Quantification -**Current OpenH-RF release:** 43 HDF5 files; 110.62 GB (110,619,394,048 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 43 HDF5 files; 110.62 GB (110,619,394,048 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. ## Contents diff --git a/datasets/resolvestroke/clinical/README.md b/datasets/resolvestroke/clinical/README.md index 0b24417..2690be3 100644 --- a/datasets/resolvestroke/clinical/README.md +++ b/datasets/resolvestroke/clinical/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - 3d - matrix-probe - contrast-enhanced @@ -24,7 +24,7 @@ size_categories: Microvascular image (x-z MIP of the computed reference) for the 20 clinical acquisitions -*Reference microvascular image (`mvi`, x-z maximum-intensity projection of the computed reference) for the 20 acquisitions in [`clinical/`](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/clinical).* +*Reference microvascular image (`mvi`, x-z maximum-intensity projection of the computed reference) for the 20 acquisitions in [`clinical/`](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/clinical).* ## Dataset Description @@ -71,12 +71,12 @@ Contrast-enhanced ultrasound flow/perfusion quantification, microbubble detectio ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/resolvestroke/clinical/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/resolvestroke/clinical/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. -The reconstruction recipe is identical for all 20 acquisitions, so the scripts and pipeline YAMLs live once in this directory. Each script has a `SUBJECT` constant at the top (default `"SP02-Left-2"`); set it to any acquisition name from the inventory, e.g. `SUBJECT = "SP07-Right"`, and the `hf://` input path `hf://nvidia/OpenH-RF/resolvestroke/clinical//.hdf5` and the output filenames follow from it. Swap `ZEA_FILE` (`INPUT` in `reconstruct_PD_3d.py`) for a local path to run against your own copy. +The reconstruction recipe is identical for all 20 acquisitions, so the scripts and pipeline YAMLs live once in this directory. Each script has a `SUBJECT` constant at the top (default `"SP02-Left-2"`); set it to any acquisition name from the inventory, e.g. `SUBJECT = "SP07-Right"`, and the `hf://` input path `hf://nvidia/Open-H-RF/resolvestroke/clinical//.hdf5` and the output filenames follow from it. Swap `ZEA_FILE` (`INPUT` in `reconstruct_PD_3d.py`) for a local path to run against your own copy. ```bash -uv run --project /path/to/OpenH-RF python reconstruct.py +uv run --project /path/to/Open-H-RF python reconstruct.py ``` `reconstruct_PD_3d.py` (with `pipeline_PD_3d.yaml`) renders the power-Doppler reconstruction and `display_references.py` the computed reference maps; both are described below. @@ -95,7 +95,7 @@ The files are in the zea HDF5 format, root `zea_version` 0.1.6. `data/raw_data` ## Dataset Quantification -**Current OpenH-RF release:** 20 HDF5 files; 105.16 GB (105,159,196,672 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 20 HDF5 files; 105.16 GB (105,159,196,672 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - Acquisitions: 20 (10 subjects × 2) - Frames per acquisition: 20000 (5 clips × 4000 frames) @@ -129,26 +129,26 @@ One row per acquisition. Power Doppler: x-z maximum-intensity projection of each AcquisitionSubjectSidePower Doppler, five clips (x-z MIP)Reference maps (mvi · velocity_radial_avg, x-z and y-z) - SP01-Left-1SP01LeftSP01-Left-1 power Doppler of the five clipsSP01-Left-1 reference maps - SP01-Left-2SP01LeftSP01-Left-2 power Doppler of the five clipsSP01-Left-2 reference maps - SP02-Left-1SP02LeftSP02-Left-1 power Doppler of the five clipsSP02-Left-1 reference maps - SP02-Left-2SP02LeftSP02-Left-2 power Doppler of the five clipsSP02-Left-2 reference maps - SP03-LeftSP03LeftSP03-Left power Doppler of the five clipsSP03-Left reference maps - SP03-RightSP03RightSP03-Right power Doppler of the five clipsSP03-Right reference maps - SP04-Left-1SP04LeftSP04-Left-1 power Doppler of the five clipsSP04-Left-1 reference maps - SP04-Left-2SP04LeftSP04-Left-2 power Doppler of the five clipsSP04-Left-2 reference maps - SP05-LeftSP05LeftSP05-Left power Doppler of the five clipsSP05-Left reference maps - SP05-RightSP05RightSP05-Right power Doppler of the five clipsSP05-Right reference maps - SP06-LeftSP06LeftSP06-Left power Doppler of the five clipsSP06-Left reference maps - SP06-RightSP06RightSP06-Right power Doppler of the five clipsSP06-Right reference maps - SP07-LeftSP07LeftSP07-Left power Doppler of the five clipsSP07-Left reference maps - SP07-RightSP07RightSP07-Right power Doppler of the five clipsSP07-Right reference maps - SP08-LeftSP08LeftSP08-Left power Doppler of the five clipsSP08-Left reference maps - SP08-RightSP08RightSP08-Right power Doppler of the five clipsSP08-Right reference maps - SP09-Left-1SP09LeftSP09-Left-1 power Doppler of the five clipsSP09-Left-1 reference maps - SP09-Left-2SP09LeftSP09-Left-2 power Doppler of the five clipsSP09-Left-2 reference maps - SP10-LeftSP10LeftSP10-Left power Doppler of the five clipsSP10-Left reference maps - SP10-RightSP10RightSP10-Right power Doppler of the five clipsSP10-Right reference maps + SP01-Left-1SP01LeftSP01-Left-1 power Doppler of the five clipsSP01-Left-1 reference maps + SP01-Left-2SP01LeftSP01-Left-2 power Doppler of the five clipsSP01-Left-2 reference maps + SP02-Left-1SP02LeftSP02-Left-1 power Doppler of the five clipsSP02-Left-1 reference maps + SP02-Left-2SP02LeftSP02-Left-2 power Doppler of the five clipsSP02-Left-2 reference maps + SP03-LeftSP03LeftSP03-Left power Doppler of the five clipsSP03-Left reference maps + SP03-RightSP03RightSP03-Right power Doppler of the five clipsSP03-Right reference maps + SP04-Left-1SP04LeftSP04-Left-1 power Doppler of the five clipsSP04-Left-1 reference maps + SP04-Left-2SP04LeftSP04-Left-2 power Doppler of the five clipsSP04-Left-2 reference maps + SP05-LeftSP05LeftSP05-Left power Doppler of the five clipsSP05-Left reference maps + SP05-RightSP05RightSP05-Right power Doppler of the five clipsSP05-Right reference maps + SP06-LeftSP06LeftSP06-Left power Doppler of the five clipsSP06-Left reference maps + SP06-RightSP06RightSP06-Right power Doppler of the five clipsSP06-Right reference maps + SP07-LeftSP07LeftSP07-Left power Doppler of the five clipsSP07-Left reference maps + SP07-RightSP07RightSP07-Right power Doppler of the five clipsSP07-Right reference maps + SP08-LeftSP08LeftSP08-Left power Doppler of the five clipsSP08-Left reference maps + SP08-RightSP08RightSP08-Right power Doppler of the five clipsSP08-Right reference maps + SP09-Left-1SP09LeftSP09-Left-1 power Doppler of the five clipsSP09-Left-1 reference maps + SP09-Left-2SP09LeftSP09-Left-2 power Doppler of the five clipsSP09-Left-2 reference maps + SP10-LeftSP10LeftSP10-Left power Doppler of the five clipsSP10-Left reference maps + SP10-RightSP10RightSP10-Right power Doppler of the five clipsSP10-Right reference maps ## Subject Metadata @@ -177,7 +177,7 @@ A display-only linear TGC (1.5 dB/cm) adds a depth-dependent dB gain to the rend Run (a GPU is strongly recommended: about 14 min per acquisition with `uv sync --extra gpu`, hours with the CPU-only JAX of the plain `uv sync`): ```bash -uv run --project /path/to/OpenH-RF python reconstruct_PD_3d.py +uv run --project /path/to/Open-H-RF python reconstruct_PD_3d.py ``` ## Computed References @@ -204,7 +204,7 @@ Notes: ![Computed reference maps montage, SP02-Left-2](../assets/clinical_SP02-Left-2_references_montage.png) ```bash -uv run --project /path/to/OpenH-RF python display_references.py +uv run --project /path/to/Open-H-RF python display_references.py ``` ## Ethical Considerations diff --git a/datasets/resolvestroke/clinical/display_references.py b/datasets/resolvestroke/clinical/display_references.py index cbf6a67..4baa1eb 100644 --- a/datasets/resolvestroke/clinical/display_references.py +++ b/datasets/resolvestroke/clinical/display_references.py @@ -10,8 +10,8 @@ pick one with ``SUBJECT`` below, or pass ``--input`` (a Hub ``hf://`` path or a local file). - uv run --project /path/to/OpenH-RF python display_references.py - uv run --project /path/to/OpenH-RF python display_references.py --input SP07-Right.hdf5 + uv run --project /path/to/Open-H-RF python display_references.py + uv run --project /path/to/Open-H-RF python display_references.py --input SP07-Right.hdf5 """ import argparse @@ -29,7 +29,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus; override with --input. SUBJECT = "SP02-Left-2" -DEFAULT_INPUT = f"hf://nvidia/OpenH-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" +DEFAULT_INPUT = f"hf://nvidia/Open-H-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" # name, colormap, contrast limits (None = auto), signed (diverging max-|.| MIP) MAPS = [ diff --git a/datasets/resolvestroke/clinical/reconstruct.py b/datasets/resolvestroke/clinical/reconstruct.py index ad4832b..357cccd 100644 --- a/datasets/resolvestroke/clinical/reconstruct.py +++ b/datasets/resolvestroke/clinical/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the resolvestroke/clinical datasets of OpenH-RF. +"""Example reconstruction script for the resolvestroke/clinical datasets of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/clinical +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/clinical B-mode reconstruction of a matrix-probe diverging-wave clinical CEUS flow acquisition. All 20 clinical acquisitions (``SP01``-``SP10``, Left/Right) share the same probe, @@ -40,7 +40,7 @@ from zea.beamform.pixelgrid import polar_pixel_grid HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/clinical/pipeline.yaml" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/clinical/pipeline.yaml" TGC_DB_PER_CM = 1.5 # display-only linear TGC (dB gain per cm of depth) # --- Inputs ----------------------------------------------------------------- @@ -48,7 +48,7 @@ # subdirectory / file stem on the Hub, e.g. "SP07-Right"). Defaults stream straight # from the published corpus. Swap ZEA_FILE for a local path to run against your own copy. SUBJECT = "SP02-Left-2" -ZEA_FILE = f"hf://nvidia/OpenH-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" +ZEA_FILE = f"hf://nvidia/Open-H-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" FRAME = 0 OUT = HERE / "assets" / f"{SUBJECT}_bmode.png" diff --git a/datasets/resolvestroke/clinical/reconstruct_PD_3d.py b/datasets/resolvestroke/clinical/reconstruct_PD_3d.py index 11eeb1b..1aeae1a 100644 --- a/datasets/resolvestroke/clinical/reconstruct_PD_3d.py +++ b/datasets/resolvestroke/clinical/reconstruct_PD_3d.py @@ -30,11 +30,11 @@ referencing them is loaded, which is why the YAML can list them by name. Runtime: about 14 min for a full acquisition on an RTX 4000 Ada with GPU JAX -(``uv sync --extra gpu`` in the OpenH-RF repo; the plain ``uv sync`` installs a CPU-only +(``uv sync --extra gpu`` in the Open-H-RF repo; the plain ``uv sync`` installs a CPU-only JAX and this script then takes hours). Lower ``N_FRAMES`` for a quick look. Usage: - KERAS_BACKEND=jax uv run --project /path/to/OpenH-RF python reconstruct_PD_3d.py + KERAS_BACKEND=jax uv run --project /path/to/Open-H-RF python reconstruct_PD_3d.py """ import os @@ -54,7 +54,7 @@ from zea.ops import Operation HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/clinical/pipeline_PD_3d.yaml" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/clinical/pipeline_PD_3d.yaml" # Clip definitions (matching the combined file structure) CLIPS = [ @@ -76,7 +76,7 @@ # subdirectory / file stem on the Hub, e.g. "SP07-Right"). Defaults stream straight # from the published corpus. Swap INPUT for a local path to run against your own copy. SUBJECT = "SP02-Left-2" -INPUT = f"hf://nvidia/OpenH-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" +INPUT = f"hf://nvidia/Open-H-RF/resolvestroke/clinical/{SUBJECT}/{SUBJECT}.hdf5" N_FRAMES = N_FRAMES_BF OUT = HERE / f"{SUBJECT}_PD_montage.png" @@ -88,7 +88,7 @@ # Custom pipeline operations (registered locally — no zea PR required). # --------------------------------------------------------------------------- # # These ops are defined here, not in zea: a pipeline.yaml naming them resolves -# only once this module is imported. See https://github.com/open-h/OpenH-RF +# only once this module is imported. See https://github.com/open-h/Open-H-RF @ops_registry("tissue_highpass") class TissueHighpass(Operation): """Slow-time high-pass (wall) filter to suppress stationary tissue clutter. diff --git a/datasets/resolvestroke/phantom_flow/README.md b/datasets/resolvestroke/phantom_flow/README.md index 2862eb8..cd025e6 100644 --- a/datasets/resolvestroke/phantom_flow/README.md +++ b/datasets/resolvestroke/phantom_flow/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - 3d - matrix-probe - contrast-enhanced @@ -23,7 +23,7 @@ size_categories: ![Power-Doppler 3D MIP montage with reference mvi](../assets/phantom_flow_PD_montage.png) -*Power Doppler of the five clips (x-z and y-z maximum-intensity projections), with the reference `mvi` in the last column, reconstructed from [`phantom_flow.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/resolvestroke/phantom_flow/phantom_flow.hdf5).* +*Power Doppler of the five clips (x-z and y-z maximum-intensity projections), with the reference `mvi` in the last column, reconstructed from [`phantom_flow.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/resolvestroke/phantom_flow/phantom_flow.hdf5).* ## Dataset Description @@ -68,10 +68,10 @@ Contrast-enhanced ultrasound flow quantification, microbubble detection and trac ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/resolvestroke/phantom_flow/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/resolvestroke/phantom_flow/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. ```bash -uv run --project /path/to/OpenH-RF python reconstruct.py +uv run --project /path/to/Open-H-RF python reconstruct.py ``` `reconstruct_PD_3d.py` (with `pipeline_PD_3d.yaml`) renders the power-Doppler reconstruction and `display_references.py` the computed reference maps; both are described below. @@ -90,7 +90,7 @@ The file is in the zea HDF5 format, root `zea_version` 0.1.6, validated `complia ## Dataset Quantification -**Current OpenH-RF release:** 1 HDF5 file; 5.25 GB (5,250,744,320 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 1 HDF5 file; 5.25 GB (5,250,744,320 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - Frames: 20000 (5 clips × 4000 frames) - Frame rate: 4000 Hz (within each clip) @@ -136,7 +136,7 @@ The resulting montage is shown at the top of this card. Run (a GPU is strongly recommended: about 14 min per acquisition with `uv sync --extra gpu`, hours with the CPU-only JAX of the plain `uv sync`): ```bash -uv run --project /path/to/OpenH-RF python reconstruct_PD_3d.py +uv run --project /path/to/Open-H-RF python reconstruct_PD_3d.py ``` ## Computed References @@ -162,7 +162,7 @@ Notes: ![Computed reference maps montage](../assets/phantom_flow_references_montage.png) -Run: `uv run --project /path/to/OpenH-RF python display_references.py` +Run: `uv run --project /path/to/Open-H-RF python display_references.py` ## Known Issues diff --git a/datasets/resolvestroke/phantom_flow/display_references.py b/datasets/resolvestroke/phantom_flow/display_references.py index ea989f3..6bef1af 100644 --- a/datasets/resolvestroke/phantom_flow/display_references.py +++ b/datasets/resolvestroke/phantom_flow/display_references.py @@ -6,7 +6,7 @@ `coordinates`) and renders, for each map, two orthogonal maximum-intensity projections (x-z and y-z) on the true Cartesian geometry. - uv run --project /path/to/OpenH-RF python display_references.py + uv run --project /path/to/Open-H-RF python display_references.py """ import argparse @@ -24,7 +24,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus; override with --input. -DEFAULT_INPUT = "hf://nvidia/OpenH-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" +DEFAULT_INPUT = "hf://nvidia/Open-H-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" # name, colormap, contrast limits (None = auto), signed (diverging max-|.| MIP) # Columns left -> right; mask first. diff --git a/datasets/resolvestroke/phantom_flow/reconstruct.py b/datasets/resolvestroke/phantom_flow/reconstruct.py index cc76b1a..5a1e363 100644 --- a/datasets/resolvestroke/phantom_flow/reconstruct.py +++ b/datasets/resolvestroke/phantom_flow/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the resolvestroke/phantom_flow dataset of OpenH-RF. +"""Example reconstruction script for the resolvestroke/phantom_flow dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/phantom_flow +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/phantom_flow B-mode reconstruction of a matrix-probe diverging-wave pipe phantom flow acquisition. @@ -37,12 +37,12 @@ from zea.beamform.pixelgrid import polar_pixel_grid HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/phantom_flow/pipeline.yaml" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/phantom_flow/pipeline.yaml" # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" FRAME = 0 OUT = HERE / "assets" / f"{Path(ZEA_FILE).stem}_bmode.png" diff --git a/datasets/resolvestroke/phantom_flow/reconstruct_PD_3d.py b/datasets/resolvestroke/phantom_flow/reconstruct_PD_3d.py index 5dcae95..109c74e 100644 --- a/datasets/resolvestroke/phantom_flow/reconstruct_PD_3d.py +++ b/datasets/resolvestroke/phantom_flow/reconstruct_PD_3d.py @@ -30,11 +30,11 @@ referencing them is loaded, which is why the YAML can list them by name. Runtime: about 14 min for a full acquisition on an RTX 4000 Ada with GPU JAX -(``uv sync --extra gpu`` in the OpenH-RF repo; the plain ``uv sync`` installs a CPU-only +(``uv sync --extra gpu`` in the Open-H-RF repo; the plain ``uv sync`` installs a CPU-only JAX and this script then takes hours). Lower ``N_FRAMES`` for a quick look. Usage: - KERAS_BACKEND=jax uv run --project /path/to/OpenH-RF python reconstruct_PD_3d.py + KERAS_BACKEND=jax uv run --project /path/to/Open-H-RF python reconstruct_PD_3d.py """ import os @@ -54,7 +54,7 @@ from zea.ops import Operation HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/phantom_flow/pipeline_PD_3d.yaml" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/phantom_flow/pipeline_PD_3d.yaml" # Clip definitions (matching the combined file structure) CLIPS = [ @@ -74,7 +74,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap INPUT for a local path to # run against your own copy. -INPUT = "hf://nvidia/OpenH-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" +INPUT = "hf://nvidia/Open-H-RF/resolvestroke/phantom_flow/phantom_flow.hdf5" N_FRAMES = N_FRAMES_BF OUT = HERE / f"{Path(INPUT).stem}_PD_montage.png" @@ -86,7 +86,7 @@ # Custom pipeline operations (registered locally — no zea PR required). # --------------------------------------------------------------------------- # # These ops are defined here, not in zea: a pipeline.yaml naming them resolves -# only once this module is imported. See https://github.com/open-h/OpenH-RF +# only once this module is imported. See https://github.com/open-h/Open-H-RF @ops_registry("tissue_highpass") class TissueHighpass(Operation): """Slow-time high-pass (wall) filter to suppress stationary tissue clutter. diff --git a/datasets/resolvestroke/phantom_mp/README.md b/datasets/resolvestroke/phantom_mp/README.md index 23aa4fc..54b931b 100644 --- a/datasets/resolvestroke/phantom_mp/README.md +++ b/datasets/resolvestroke/phantom_mp/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - 3d - matrix-probe - phantom @@ -21,7 +21,7 @@ size_categories: ![Reference B-mode (two perpendicular sectors)](../assets/phantom_mp_bmode.png) -*B-mode of one frame, x-z and y-z sectors, reconstructed from [`phantom_mp.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/resolvestroke/phantom_mp/phantom_mp.hdf5).* +*B-mode of one frame, x-z and y-z sectors, reconstructed from [`phantom_mp.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/resolvestroke/phantom_mp/phantom_mp.hdf5).* ## Dataset Description @@ -56,10 +56,10 @@ Beamforming and reconstruction research (RFP task group 6.1, Generalized Reconst ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/resolvestroke/phantom_mp/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/resolvestroke/phantom_mp/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. ```bash -uv run --project /path/to/OpenH-RF python reconstruct.py +uv run --project /path/to/Open-H-RF python reconstruct.py ``` The Python scripts carry their own `SPDX-License-Identifier: Apache-2.0` header; the dataset itself is CC BY 4.0. @@ -72,7 +72,7 @@ Single zea HDF5 file (`phantom_mp.hdf5`), one track holding the raw channel data ## Dataset Quantification -**Current OpenH-RF release:** 1 HDF5 file; 33.75 MB (33,751,040 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 1 HDF5 file; 33.75 MB (33,751,040 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - Frames: 1000 (single acquisition) - Train / val / test split: N/A diff --git a/datasets/resolvestroke/phantom_mp/reconstruct.py b/datasets/resolvestroke/phantom_mp/reconstruct.py index 4cbc589..c6b3793 100644 --- a/datasets/resolvestroke/phantom_mp/reconstruct.py +++ b/datasets/resolvestroke/phantom_mp/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the resolvestroke/phantom_mp dataset of OpenH-RF. +"""Example reconstruction script for the resolvestroke/phantom_mp dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/phantom_mp +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/phantom_mp B-mode reconstruction of a matrix-probe diverging-wave phantom acquisition. @@ -35,12 +35,12 @@ from zea.beamform.pixelgrid import polar_pixel_grid HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/phantom_mp/pipeline.yaml" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/phantom_mp/pipeline.yaml" # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/resolvestroke/phantom_mp/phantom_mp.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/resolvestroke/phantom_mp/phantom_mp.hdf5" FRAME = 0 OUT = HERE / "assets" / f"{Path(ZEA_FILE).stem}_bmode.png" diff --git a/datasets/resolvestroke/saddle/README.md b/datasets/resolvestroke/saddle/README.md index 2b8b9b4..375156f 100644 --- a/datasets/resolvestroke/saddle/README.md +++ b/datasets/resolvestroke/saddle/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - matrix-probe - diverging-wave - b-mode @@ -24,13 +24,13 @@ size_categories: ![Saddle-array B-modes for all 21 datasets](../assets/saddle_bmode_montage.png) -*Saddle-array B-modes of all 21 files in [`saddle/data/`](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/saddle/data), one panel per dataset; the phantom (`PMP01`) shows a regular column of point targets.* +*Saddle-array B-modes of all 21 files in [`saddle/data/`](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/saddle/data), one panel per dataset; the phantom (`PMP01`) shows a regular column of point targets.* ## Dataset Description Single-frame anatomical B-mode acquisitions from the "saddle" imaging sequence of Resolve Stroke's SYLVER ultrasound device, using a 32×32 matrix probe. Each raw acquisition contains, alongside the multi-thousand-frame contrast-enhanced ultrasound (CEUS) sequence, one wide-angle diverging-wave frame (the *saddle* sequence: `x_ang` −24°…+24° in nine steps, no elevation steering, cylindrical elevation focus) received on the full 1024-element aperture through four consecutive 256-element receive events per transmit. Beamformed, this single frame gives a sector B-mode of the imaging plane: the structural reference view acquired at the same probe placement as the contrast (CEUS) recording. -This directory contains one such reference B-mode per dataset: 20 clinical transcranial acquisitions (SCULPT study) and 1 static matrix-probe imaging phantom (21 files total). It is the structural companion to the OpenH-RF Resolve Stroke clinical CEUS clip submission, and the two use the same anonymized subject codes. +This directory contains one such reference B-mode per dataset: 20 clinical transcranial acquisitions (SCULPT study) and 1 static matrix-probe imaging phantom (21 files total). It is the structural companion to the Open-H-RF Resolve Stroke clinical CEUS clip submission, and the two use the same anonymized subject codes. ## Dataset Contributor(s) @@ -51,7 +51,7 @@ This directory contains one such reference B-mode per dataset: 20 clinical trans ## Intended Usage -Matrix-probe diverging-wave beamforming research, anatomical B-mode reconstruction, and structural reference for the companion transcranial CEUS flow/perfusion datasets (OpenH-RF request-for-proposals task group 6.2, Blood Flow). +Matrix-probe diverging-wave beamforming research, anatomical B-mode reconstruction, and structural reference for the companion transcranial CEUS flow/perfusion datasets (Open-H-RF request-for-proposals task group 6.2, Blood Flow). ## Dataset Characterization @@ -61,10 +61,10 @@ Matrix-probe diverging-wave beamforming research, anatomical B-mode reconstructi ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/resolvestroke/saddle/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/resolvestroke/saddle/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. ```bash -uv run --project /path/to/OpenH-RF python reconstruct.py +uv run --project /path/to/Open-H-RF python reconstruct.py ``` `reconstruct.py` streams `PMP01.hdf5` from the Hub by default; set `ZEA_FILE` at the top of the script to another of the 21 files (or a local path). The B-mode PNG is written to `assets/_bmode.png`. @@ -83,7 +83,7 @@ The hardware time-gain compensation is baked into `raw_data`; `scan/tgc_gain_cur ## Dataset Quantification -**Current OpenH-RF release:** 21 HDF5 files; 175.70 MB (175,702,016 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 21 HDF5 files; 175.70 MB (175,702,016 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - Files: 21 (20 clinical + 1 phantom) - Frames per file: 1 diff --git a/datasets/resolvestroke/saddle/reconstruct.py b/datasets/resolvestroke/saddle/reconstruct.py index e9f16f4..a2b113f 100644 --- a/datasets/resolvestroke/saddle/reconstruct.py +++ b/datasets/resolvestroke/saddle/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the resolvestroke/saddle dataset of OpenH-RF. +"""Example reconstruction script for the resolvestroke/saddle dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/resolvestroke/saddle +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/resolvestroke/saddle B-mode reconstruction of a matrix-probe diverging-wave acquisition. @@ -33,8 +33,8 @@ from zea import Config, File, Pipeline HERE = Path(__file__).parent -CONFIG = "hf://nvidia/OpenH-RF/resolvestroke/saddle/pipeline.yaml" -HF_DIR = "hf://nvidia/OpenH-RF/resolvestroke/saddle/data" +CONFIG = "hf://nvidia/Open-H-RF/resolvestroke/saddle/pipeline.yaml" +HF_DIR = "hf://nvidia/Open-H-RF/resolvestroke/saddle/data" # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Any of the 21 files works: diff --git a/datasets/siemens-healthineers/README.md b/datasets/siemens-healthineers/README.md index ab47e6f..106430c 100644 --- a/datasets/siemens-healthineers/README.md +++ b/datasets/siemens-healthineers/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - medical-imaging - beamforming - in-vivo @@ -21,7 +21,7 @@ size_categories: ![DAS B-mode reconstructions of four anatomies from the Siemens FSA channel data](assets/anatomies.png) -*Four of the anatomies in this dataset, each beamformed from `data/raw_data`: [`Subject_03_acq_018`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/siemens-healthineers/data/Subject_03_acq_018.hdf5), [`Subject_02_acq_004`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/siemens-healthineers/data/Subject_02_acq_004.hdf5), [`Subject_03_acq_008`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/siemens-healthineers/data/Subject_03_acq_008.hdf5), [`Subject_05_acq_031`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/siemens-healthineers/data/Subject_05_acq_031.hdf5).* +*Four of the anatomies in this dataset, each beamformed from `data/raw_data`: [`Subject_03_acq_018`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/siemens-healthineers/data/Subject_03_acq_018.hdf5), [`Subject_02_acq_004`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/siemens-healthineers/data/Subject_02_acq_004.hdf5), [`Subject_03_acq_008`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/siemens-healthineers/data/Subject_03_acq_008.hdf5), [`Subject_05_acq_031`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/siemens-healthineers/data/Subject_05_acq_031.hdf5).* ## Dataset Description @@ -61,11 +61,11 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/siemens-healthineers/data/Subject_03_acq_008.hdf5 \ - --config hf://nvidia/OpenH-RF/siemens-healthineers/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/siemens-healthineers/data/Subject_03_acq_008.hdf5 \ + --config hf://nvidia/Open-H-RF/siemens-healthineers/pipeline.yaml ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/siemens-healthineers/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/siemens-healthineers/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -105,7 +105,7 @@ All spatial maps carry per-pixel Cartesian `coordinates` in meters, last axis `[ ## Dataset Quantification -**Current OpenH-RF release:** 590 HDF5 files; 228.54 GB (228,538,449,920 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 590 HDF5 files; 228.54 GB (228,538,449,920 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **590 acquisitions (4,539 frames) from 15 subjects**, one continuous dataset: 51 acquisitions / 411 frames (Subjects 01–03) from the first acquisition round and 539 acquisitions / 4,128 frames (Subjects 04–15) from the second. - **File naming:** one HDF5 per acquisition, `Subject_XX_acq_NNN.hdf5` (acquisitions numbered per subject in acquisition order). diff --git a/datasets/siemens-healthineers/reconstruct.py b/datasets/siemens-healthineers/reconstruct.py index ffbbccb..7b297b6 100644 --- a/datasets/siemens-healthineers/reconstruct.py +++ b/datasets/siemens-healthineers/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the siemens-healthineers dataset of OpenH-RF. +"""Example reconstruction script for the siemens-healthineers dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/siemens-healthineers +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/siemens-healthineers B-mode reconstruction of in-vivo IQ channel data from a clinical scanner. @@ -35,11 +35,11 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/siemens-healthineers/data/Subject_01_acq_006.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/siemens-healthineers/data/Subject_01_acq_006.hdf5" CONFIG = HERE / "pipeline.yaml" OUT = None # PNG path (default: _recon.png) FRAME = 0 -HF_CONFIG = "hf://nvidia/OpenH-RF/siemens-healthineers/pipeline.yaml" # where CONFIG is published +HF_CONFIG = "hf://nvidia/Open-H-RF/siemens-healthineers/pipeline.yaml" # where CONFIG is published def main(): diff --git a/datasets/stanford-murine/README.md b/datasets/stanford-murine/README.md index 2da4292..572b22b 100644 --- a/datasets/stanford-murine/README.md +++ b/datasets/stanford-murine/README.md @@ -8,7 +8,7 @@ task_categories: tags: - ultrasound - rf -- openh-rf +- open-h-rf - beamforming - sound-speed-estimation - aberration-correction @@ -25,7 +25,7 @@ size_categories: ![B-mode reconstructions of a rat liver and the ATS 549 phantom](assets/hero.png) -*Frame 0 reconstructed with `reconstruct.py` (L12-3v, 60 dB). Top: the exposed liver of [Rat3](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/stanford-murine/data/RatExperiments/VerasonicsAcq/Rat3/ExposedLiver/DATA_Tracks_20190319_115804.hdf5), `multifocal` track. Bottom: the ATS 549 phantom's [lesions](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/stanford-murine/data/SoSExperiments/august13_2020/L12_3v/ATS_phantom/lesion/DATA_Tracks_20200813_164505.hdf5) and [point targets](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/stanford-murine/data/SoSExperiments/august13_2020/L12_3v/ATS_phantom/point_target/DATA_Tracks_20200813_164745.hdf5), `hadamard` track. The point targets curve into arcs because the phantom is beamformed at 1540 m/s while its own sound speed is lower — the mismatch this dataset is built to estimate and correct.* +*Frame 0 reconstructed with `reconstruct.py` (L12-3v, 60 dB). Top: the exposed liver of [Rat3](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/stanford-murine/data/RatExperiments/VerasonicsAcq/Rat3/ExposedLiver/DATA_Tracks_20190319_115804.hdf5), `multifocal` track. Bottom: the ATS 549 phantom's [lesions](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/stanford-murine/data/SoSExperiments/august13_2020/L12_3v/ATS_phantom/lesion/DATA_Tracks_20200813_164505.hdf5) and [point targets](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/stanford-murine/data/SoSExperiments/august13_2020/L12_3v/ATS_phantom/point_target/DATA_Tracks_20200813_164745.hdf5), `hadamard` track. The point targets curve into arcs because the phantom is beamformed at 1540 m/s while its own sound speed is lower — the mismatch this dataset is built to estimate and correct.* ## Dataset Description @@ -38,7 +38,7 @@ The source data are available from Figshare: The original Figshare releases were published under Apache 2.0; this converted release is CC BY 4.0. -The converted dataset is hosted at [nvidia/OpenH-RF/stanford-murine](https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/stanford-murine). +The converted dataset is hosted at [nvidia/Open-H-RF/stanford-murine](https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/stanford-murine). ## Dataset Contributor(s) @@ -109,13 +109,13 @@ The acquisitions can be processed with the `pipeline_multifocal.yaml`, `pipeline ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/stanford-murine/data/RatExperiments/VerasonicsAcq/Rat3/ExposedLiver/DATA_Tracks_20190319_115804.hdf5 \ - --config hf://nvidia/OpenH-RF/stanford-murine/pipeline_synthetic_aperture.yaml \ + --dataset hf://nvidia/Open-H-RF/stanford-murine/data/RatExperiments/VerasonicsAcq/Rat3/ExposedLiver/DATA_Tracks_20190319_115804.hdf5 \ + --config hf://nvidia/Open-H-RF/stanford-murine/pipeline_synthetic_aperture.yaml \ --track synthetic_aperture \ --n-frames 1 ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/stanford-murine/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). It reconstructs the first frame of every track in `ZEA_FILE` with the same pipelines and writes one image per track to `assets/`. On top of the YAML, the script inserts a baseband FIR low-pass after demodulation, with its cutoff at half the probe bandwidth (4.0 MHz for L12-3v, 3.0 MHz for L12-5, 1.3 MHz for C5-2v). The cutoff depends on the probe, so it is not part of the pipeline YAMLs and `zea process` runs without it; the difference is small on the rat acquisitions and clearly visible at depth on the phantoms. +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/stanford-murine/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). It reconstructs the first frame of every track in `ZEA_FILE` with the same pipelines and writes one image per track to `assets/`. On top of the YAML, the script inserts a baseband FIR low-pass after demodulation, with its cutoff at half the probe bandwidth (4.0 MHz for L12-3v, 3.0 MHz for L12-5, 1.3 MHz for C5-2v). The cutoff depends on the probe, so it is not part of the pipeline YAMLs and `zea process` runs without it; the difference is small on the rat acquisitions and clearly visible at depth on the phantoms. ## Dataset Format @@ -141,7 +141,7 @@ Reconstructed B-mode images and segmentation masks are not stored in the convert ## Dataset Quantification -**Current OpenH-RF release:** 245 HDF5 files; 328.99 GB (328,985,804,800 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 245 HDF5 files; 328.99 GB (328,985,804,800 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. The original four-file demo (not the complete HF release summarized above) contains four HDF5 acquisition bundles: one Rat9 bundle and one phantom bundle for each of C5-2v, L12-3v, and L12-5. Each bundle has three sequence tracks, and each track retains every frame in its source MAT file. That demo has 12 tracks and 64 frames and occupies approximately 5.5 GiB; HDF5 writer and compression versions may change the exact byte count. diff --git a/datasets/stanford-murine/reconstruct.py b/datasets/stanford-murine/reconstruct.py index b7723b0..25aad5b 100644 --- a/datasets/stanford-murine/reconstruct.py +++ b/datasets/stanford-murine/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the stanford-murine dataset of OpenH-RF. +"""Example reconstruction script for the stanford-murine dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/stanford-murine +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/stanford-murine B-mode reconstruction of murine full synthetic aperture, multifocal and Hadamard-encoded channel data. @@ -40,10 +40,10 @@ # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. ZEA_FILE = ( - "hf://nvidia/OpenH-RF/stanford-murine/data/RatExperiments/VerasonicsAcq/" + "hf://nvidia/Open-H-RF/stanford-murine/data/RatExperiments/VerasonicsAcq/" "Rat3/ExposedLiver/DATA_Tracks_20190319_115804.hdf5" ) -CONFIG_DIR = "hf://nvidia/OpenH-RF/stanford-murine" # holds the pipeline_*.yaml configs +CONFIG_DIR = "hf://nvidia/Open-H-RF/stanford-murine" # holds the pipeline_*.yaml configs OUT_DIR = HERE / "assets" # every PNG is written here diff --git a/datasets/strasbourg-basel/README.md b/datasets/strasbourg-basel/README.md index 705f839..a392940 100644 --- a/datasets/strasbourg-basel/README.md +++ b/datasets/strasbourg-basel/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - tracked-ultrasound - robot-tracked - probe-pose @@ -29,7 +29,7 @@ size_categories: BoneSRF B-mode of phantom 2, whole-bone sweep

-*B-mode of a fractured-femur phantom, frame 45 of [`data/phantom2_wholebone.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/strasbourg-basel/BoneSRF/data/phantom2_wholebone.hdf5), beamformed from the recovered channel data.* +*B-mode of a fractured-femur phantom, frame 45 of [`data/phantom2_wholebone.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/strasbourg-basel/BoneSRF/data/phantom2_wholebone.hdf5), beamformed from the recovered channel data.* ## Dataset Description @@ -37,7 +37,7 @@ size_categories: BoneSRF

-**BoneSRF** (Bone Surface Reflection) is an ultrasound channel-data dataset for bone-surface and fracture-reflection research. It contains three 3D-printed, fractured femur phantoms, scanned with a handheld point-of-care probe, inverted back to pre-beamformed RF and packaged in the OpenH-RF `zea` format. +**BoneSRF** (Bone Surface Reflection) is an ultrasound channel-data dataset for bone-surface and fracture-reflection research. It contains three 3D-printed, fractured femur phantoms, scanned with a handheld point-of-care probe, inverted back to pre-beamformed RF and packaged in the Open-H-RF `zea` format. Each phantom was swept three times (`distal`, `proximal`, `wholebone`), giving nine `zea` HDF5 files. Every scan is robot-tracked: the probe was mounted on a robotic arm and its pose recorded separately. Each file also carries that phantom's CT scan and multi-label segmentation. @@ -49,7 +49,7 @@ This is phantom data, not simulated, clinical or in-vivo. It is intended for ful 1. **Acquisition.** Each phantom was scanned with a Clarius L20HD3 linear array, producing the probe's own beamformed RF output. This is the physically acquired data, not simulated. 2. **Inversion.** The beamformed RF is inverted back into pre-beamformed, per-element channel data with [`das-inverse` (`clarius` branch)](https://github.com/sankethvedula/das-inverse/tree/clarius), a CGLS solver over a `zea.inverse.DASOperator` forward model of the probe's focused, walking-sub-aperture transmit sequence (`invert_clarius_beamformed.py`). The result is not sensor-captured, but numerically consistent with the real beamformed acquisition it was inverted from. -3. **Packaging.** The inverted channel data, transmit-sequence metadata, probe geometry, per-frame probe pose, and the phantom's CT + segmentation are written out as one `zea` HDF5 file per scan, matching the OpenH-RF format spec. +3. **Packaging.** The inverted channel data, transmit-sequence metadata, probe geometry, per-frame probe pose, and the phantom's CT + segmentation are written out as one `zea` HDF5 file per scan, matching the Open-H-RF format spec. ### Probe tracking @@ -91,7 +91,7 @@ The CT and segmentation data inside the files are released under the same CC BY ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/strasbourg-basel/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/strasbourg-basel/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. The script loads the saved `zea.Pipeline` from `pipeline.yaml`, runs it on one frame of one scan, and writes `assets/.png`. @@ -105,7 +105,7 @@ The constants at the top of the script select what is reconstructed: These values reconstruct `phantom2_wholebone` at its reference frame (45) on the CPU: ```python -SCAN = "hf://nvidia/OpenH-RF/strasbourg-basel/BoneSRF/data/phantom2_wholebone.hdf5" +SCAN = "hf://nvidia/Open-H-RF/strasbourg-basel/BoneSRF/data/phantom2_wholebone.hdf5" FRAME = None DEVICE = "cpu" ``` @@ -226,7 +226,7 @@ The CT is reference imaging of the physical phantom in scanner (LPS) space. It i ## Dataset Quantification -**Current OpenH-RF release:** 9 HDF5 files; 200.75 GB (200,745,025,536 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 9 HDF5 files; 200.75 GB (200,745,025,536 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. Nine acquisitions, one continuous sweep each; 1,427 frames in total. No train / val / test split (each file is a single reference acquisition). Every scan has one tracked probe pose per frame. diff --git a/datasets/strasbourg-basel/reconstruct.py b/datasets/strasbourg-basel/reconstruct.py index 064aabf..10e929f 100644 --- a/datasets/strasbourg-basel/reconstruct.py +++ b/datasets/strasbourg-basel/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the strasbourg-basel dataset of OpenH-RF. +"""Example reconstruction script for the strasbourg-basel dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/strasbourg-basel +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/strasbourg-basel B-mode reconstruction of focused, walking-aperture bone channel data (BoneSRF). @@ -54,7 +54,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -SCAN = "hf://nvidia/OpenH-RF/strasbourg-basel/BoneSRF/data/phantom2_proximal.hdf5" +SCAN = "hf://nvidia/Open-H-RF/strasbourg-basel/BoneSRF/data/phantom2_proximal.hdf5" FRAME = None # frame to beamform (default: that scan's reference frame) DEVICE = None # CUDA device ID (e.g. 'cuda:0', 'auto:1', or 'cpu') CT = True # also write assets/ct_.png from the CT embedded in the file diff --git a/datasets/technion/bladder/README.md b/datasets/technion/bladder/README.md index d36f9ec..4d144d8 100644 --- a/datasets/technion/bladder/README.md +++ b/datasets/technion/bladder/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - beamforming - bladder - 3d @@ -21,7 +21,7 @@ size_categories: ![Transverse suprapubic view of the bladder, one 10-frame cine loop](assets/cine.gif) -*Transverse suprapubic view: one cine loop (10 frames) from [`data/a1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/technion/bladder/data/a1.hdf5).* +*Transverse suprapubic view: one cine loop (10 frames) from [`data/a1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/technion/bladder/data/a1.hdf5).* ## Dataset Description @@ -37,7 +37,7 @@ Real, **in-vivo human** pre-beamformed ultrasound **channel data** for bladder i ## Dataset Creation Date -Source data 2018; converted to the OpenH-RF (zea) format 07/16/2026. +Source data 2018; converted to the Open-H-RF (zea) format 07/16/2026. ## License / Terms of Use @@ -61,12 +61,12 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/technion/bladder/data/a1.hdf5 \ - --config hf://nvidia/OpenH-RF/technion/bladder/pipeline.yaml \ + --dataset hf://nvidia/Open-H-RF/technion/bladder/data/a1.hdf5 \ + --config hf://nvidia/Open-H-RF/technion/bladder/pipeline.yaml \ --n-frames 10 ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/technion/bladder/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/technion/bladder/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -76,7 +76,7 @@ zea file format, one HDF5 file per sweep (`data/.hdf5`, e.g. `a1.hdf5`, ## Dataset Quantification -**Current OpenH-RF release:** 14 HDF5 files; 83.71 GB (83,713,785,856 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 14 HDF5 files; 83.71 GB (83,713,785,856 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Frames / sweeps / subjects:** 1,508 frames · 14 sweeps · 7 subjects. - **Train / val / test split:** N/A (contributor to define). diff --git a/datasets/technion/bladder/reconstruct.py b/datasets/technion/bladder/reconstruct.py index 772fa36..f7dcc7a 100644 --- a/datasets/technion/bladder/reconstruct.py +++ b/datasets/technion/bladder/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the technion/bladder dataset of OpenH-RF. +"""Example reconstruction script for the technion/bladder dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/technion/bladder +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/technion/bladder B-mode reconstruction of in-vivo bladder channel data on a polar scanline grid. @@ -32,8 +32,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/technion/bladder/data/a1.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/technion/bladder/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/technion/bladder/data/a1.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/technion/bladder/pipeline.yaml" FRAME = 10 OUT = HERE / "assets" / "bmode.png" diff --git a/datasets/technion/cardiac/README.md b/datasets/technion/cardiac/README.md index 1af40ef..e82e2b2 100644 --- a/datasets/technion/cardiac/README.md +++ b/datasets/technion/cardiac/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - beamforming - cardiac - 3d @@ -21,7 +21,7 @@ size_categories: ![Apical four-chamber view, one 32-frame cardiac cine loop](assets/cine.gif) -*Apical four-chamber view: one cine loop (32 frames) from [`data/c1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/technion/cardiac/data/c1.hdf5).* +*Apical four-chamber view: one cine loop (32 frames) from [`data/c1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/technion/cardiac/data/c1.hdf5).* ## Dataset Description @@ -37,7 +37,7 @@ Real, **in-vivo human** pre-beamformed ultrasound **channel data** for cardiac i ## Dataset Creation Date -Acquired 2018; converted to the OpenH-RF (zea) format 07/16/2026. +Acquired 2018; converted to the Open-H-RF (zea) format 07/16/2026. ## License / Terms of Use @@ -61,12 +61,12 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/technion/cardiac/data/c1.hdf5 \ - --config hf://nvidia/OpenH-RF/technion/cardiac/pipeline.yaml \ + --dataset hf://nvidia/Open-H-RF/technion/cardiac/data/c1.hdf5 \ + --config hf://nvidia/Open-H-RF/technion/cardiac/pipeline.yaml \ --n-frames 32 ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/technion/cardiac/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/technion/cardiac/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -76,7 +76,7 @@ zea file format, one HDF5 file per cine loop (`data/.hdf5`, e.g. ## Dataset Quantification -**Current OpenH-RF release:** 25 HDF5 files; 14.99 GB (14,992,998,400 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 25 HDF5 files; 14.99 GB (14,992,998,400 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Frames / cines / subjects:** 777 frames · 25 cine loops · 6 subjects (a–f). - **Train / val / test split:** N/A (contributor to define; the `f2` patient set is a natural held-out cine). diff --git a/datasets/technion/cardiac/reconstruct.py b/datasets/technion/cardiac/reconstruct.py index df8d5b7..6e75ed6 100644 --- a/datasets/technion/cardiac/reconstruct.py +++ b/datasets/technion/cardiac/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the technion/cardiac dataset of OpenH-RF. +"""Example reconstruction script for the technion/cardiac dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/technion/cardiac +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/technion/cardiac B-mode reconstruction of in-vivo cardiac channel data on a polar scanline grid. @@ -32,8 +32,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/technion/cardiac/data/c1.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/technion/cardiac/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/technion/cardiac/data/c1.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/technion/cardiac/pipeline.yaml" FRAME = 8 OUT = HERE / "assets" / "bmode.png" diff --git a/datasets/technion/phantom/README.md b/datasets/technion/phantom/README.md index af779c2..5e38ac9 100644 --- a/datasets/technion/phantom/README.md +++ b/datasets/technion/phantom/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - iq - - openh-rf + - open-h-rf - beamforming - phantom - 3d @@ -21,7 +21,7 @@ size_categories: ![Tissue-mimicking phantom with point targets and an anechoic cyst](assets/bmode.png) -*Frame 6 of [`data/ph.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/technion/phantom/data/ph.hdf5), reconstructed by `reconstruct.py`.* +*Frame 6 of [`data/ph.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/technion/phantom/data/ph.hdf5), reconstructed by `reconstruct.py`.* ## Dataset Description @@ -37,7 +37,7 @@ Pre-beamformed ultrasound **channel data** from a tissue-mimicking phantom, acqu ## Dataset Creation Date -Source data 2018; converted to the OpenH-RF (zea) format 07/16/2026. +Source data 2018; converted to the Open-H-RF (zea) format 07/16/2026. ## License / Terms of Use @@ -61,12 +61,12 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/technion/phantom/data/ph.hdf5 \ - --config hf://nvidia/OpenH-RF/technion/phantom/pipeline.yaml \ + --dataset hf://nvidia/Open-H-RF/technion/phantom/data/ph.hdf5 \ + --config hf://nvidia/Open-H-RF/technion/phantom/pipeline.yaml \ --n-frames 1 ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/technion/phantom/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/technion/phantom/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -76,7 +76,7 @@ zea file format, a single HDF5 file `data/ph.hdf5`. Source complex samples repac ## Dataset Quantification -**Current OpenH-RF release:** 1 HDF5 file; 659.95 MB (659,947,520 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 1 HDF5 file; 659.95 MB (659,947,520 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Frames / acquisitions:** 12 frames · 1 acquisition. - **Stored HDF5 size:** 659.95 MB (659,947,520 bytes). diff --git a/datasets/technion/phantom/reconstruct.py b/datasets/technion/phantom/reconstruct.py index 8d7f243..f25cc06 100644 --- a/datasets/technion/phantom/reconstruct.py +++ b/datasets/technion/phantom/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the technion/phantom dataset of OpenH-RF. +"""Example reconstruction script for the technion/phantom dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/technion/phantom +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/technion/phantom B-mode reconstruction of phantom channel data on a polar scanline grid. @@ -32,8 +32,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/technion/phantom/data/ph.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/technion/phantom/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/technion/phantom/data/ph.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/technion/phantom/pipeline.yaml" FRAME = 6 OUT = HERE / "assets" / "bmode.png" diff --git a/datasets/tel-aviv/README.md b/datasets/tel-aviv/README.md index af6f550..503ad80 100644 --- a/datasets/tel-aviv/README.md +++ b/datasets/tel-aviv/README.md @@ -3,14 +3,14 @@ name: tel-aviv license: cc-by-4.0 pretty_name: Ilovitsh Lab Mice Tumors task_categories: [image-segmentation] -tags: [ultrasound, rf, openh-rf] +tags: [ultrasound, rf, open-h-rf] --- # Ilovitsh Lab Mice Tumors & Water-Bead Phantoms ![Reconstructed cineloop from 01-scan-3.hdf5](assets/01-scan-3.gif) -*Cine loop of the rotational sweep through a mouse tumor, [`mouse_tumor/seg/01-scan-3.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tel-aviv/mouse_tumor/seg/01-scan-3.hdf5), reconstructed from the raw channel data with `mouse_tumor/pipeline.yaml`.* +*Cine loop of the rotational sweep through a mouse tumor, [`mouse_tumor/seg/01-scan-3.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tel-aviv/mouse_tumor/seg/01-scan-3.hdf5), reconstructed from the raw channel data with `mouse_tumor/pipeline.yaml`.* ## Dataset Description @@ -45,7 +45,7 @@ Segmentation of tumors in mice and water-bead phantoms. ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/tel-aviv/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definitions in `mouse_tumor/` and `phantom/` and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub and overlays the stored segmentation on the reconstructed B-mode. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/tel-aviv/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definitions in `mouse_tumor/` and `phantom/` and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub and overlays the stored segmentation on the reconstructed B-mode. ## Dataset Format @@ -53,7 +53,7 @@ The acquisitions can be processed with the `reconstruct.py` [script](https://git ## Dataset Quantification -**Current OpenH-RF release:** 22 HDF5 files; 4.36 GB (4,357,947,392 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 22 HDF5 files; 4.36 GB (4,357,947,392 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Samples / frames:** 22 total `.hdf5` acquisitions. Each acquisition contains 144 angular frames covering a 180° rotation (1.25° increments) of 5 plane waves. - **Train / val / test split:** N/A diff --git a/datasets/tel-aviv/reconstruct.py b/datasets/tel-aviv/reconstruct.py index 24c105f..7c4dba6 100644 --- a/datasets/tel-aviv/reconstruct.py +++ b/datasets/tel-aviv/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the tel-aviv dataset of OpenH-RF. +"""Example reconstruction script for the tel-aviv dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/tel-aviv +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/tel-aviv B-mode reconstruction of multi-angle plane-wave acquisitions of mouse tumors and water-bead phantoms. @@ -36,9 +36,9 @@ # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. The phantom acquisitions live under # tel-aviv/phantom and carry no segmentation. -ZEA_FILE = "hf://nvidia/OpenH-RF/tel-aviv/mouse_tumor/seg/01-scan-3.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/tel-aviv/mouse_tumor/seg/01-scan-3.hdf5" CONFIG = HERE / "mouse_tumor" / "pipeline.yaml" # this is what the run loads -HF_CONFIG = "hf://nvidia/OpenH-RF/tel-aviv/mouse_tumor/pipeline.yaml" # where CONFIG is published +HF_CONFIG = "hf://nvidia/Open-H-RF/tel-aviv/mouse_tumor/pipeline.yaml" # where CONFIG is published FRAME = 46 OUT = HERE / "assets" / "01-scan-3_frame046.png" diff --git a/datasets/tue-aaa/README.md b/datasets/tue-aaa/README.md index 2212611..806d8a6 100644 --- a/datasets/tue-aaa/README.md +++ b/datasets/tue-aaa/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - beamforming - motion-estimation - abdominal-aortic-aneurysm @@ -20,7 +20,7 @@ language: ![Reconstructed cineloop from AAA_subject11.hdf5](assets/AAA_subject11.gif) -*Cine loop of [`data/AAA_subject11.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-aaa/data/AAA_subject11.hdf5), reconstructed from the raw channel data.* +*Cine loop of [`data/AAA_subject11.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-aaa/data/AAA_subject11.hdf5), reconstructed from the raw channel data.* ## Dataset Description @@ -58,11 +58,11 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/tue-aaa/data/AAA_subject11.hdf5 \ - --config hf://nvidia/OpenH-RF/tue-aaa/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/tue-aaa/data/AAA_subject11.hdf5 \ + --config hf://nvidia/Open-H-RF/tue-aaa/pipeline.yaml ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/tue-aaa/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/tue-aaa/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -72,7 +72,7 @@ zea file format. Subject metadata is stored under `metadata/subject` (`age`, `se ## Dataset Quantification -**Current OpenH-RF release:** 15 HDF5 files; 53.08 GB (53,077,606,400 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 15 HDF5 files; 53.08 GB (53,077,606,400 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Samples / frames:** 15 acquisitions (one HDF5 file each), 500 frames per acquisition — 7,500 frames total - **Stored HDF5 size:** 53.08 GB (53,077,606,400 bytes). diff --git a/datasets/tue-aaa/reconstruct.py b/datasets/tue-aaa/reconstruct.py index e362fa9..b2053f8 100644 --- a/datasets/tue-aaa/reconstruct.py +++ b/datasets/tue-aaa/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the tue-aaa dataset of OpenH-RF. +"""Example reconstruction script for the tue-aaa dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/tue-aaa +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/tue-aaa B-mode reconstruction of in-vivo abdominal aortic aneurysm curved-array channel data. @@ -58,9 +58,9 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/tue-aaa/data/AAA_subject11.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/tue-aaa/data/AAA_subject11.hdf5" OUT = HERE / "assets" / "AAApatient01_bmode.png" -HF_CONFIG = "hf://nvidia/OpenH-RF/tue-aaa/pipeline.yaml" +HF_CONFIG = "hf://nvidia/Open-H-RF/tue-aaa/pipeline.yaml" DEVICE = None # CUDA device ID (e.g. 'cuda:0', 'auto:1', or 'cpu') diff --git a/datasets/tue-cardiac/README.md b/datasets/tue-cardiac/README.md index a720882..98aec6d 100644 --- a/datasets/tue-cardiac/README.md +++ b/datasets/tue-cardiac/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - echocardiography - cardiac - plax @@ -33,7 +33,7 @@ size_categories: -*Parasternal long-axis cine loops of four volunteers, pulse-inversion harmonic tracks: focused ([`subject-002`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-cardiac/data/subject-002.hdf5)), wide ([`subject-004`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-cardiac/data/subject-004.hdf5)), focused ([`subject-006`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-cardiac/data/subject-006.hdf5)) and wide ([`subject-009`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-cardiac/data/subject-009.hdf5)).* +*Parasternal long-axis cine loops of four volunteers, pulse-inversion harmonic tracks: focused ([`subject-002`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-cardiac/data/subject-002.hdf5)), wide ([`subject-004`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-cardiac/data/subject-004.hdf5)), focused ([`subject-006`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-cardiac/data/subject-006.hdf5)) and wide ([`subject-009`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-cardiac/data/subject-009.hdf5)).* ## Dataset Description @@ -95,7 +95,7 @@ The dataset must not be used as a clinically validated diagnostic product. ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/tue-cardiac/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipelines/*.yaml` definitions in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/tue-cardiac/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipelines/*.yaml` definitions in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. Each file holds the eight transmit-encoding tracks. Set `ZEA_FILE`, `TRACK` and `FRAME` at the top of the script, together with the matching pipeline under `pipelines/` as `CONFIG` (`PIPELINE_FOR_TRACK` lists which track uses which; see also [Data Validation](#data-validation)). The image is written to `assets/`. @@ -178,7 +178,7 @@ The source tensor order is preserved as `(n_frames, n_tx, n_ax, n_el, n_ch) = (1 Suggested citation for the dataset: -> Penninga, S. W., & van Sloun, R. J. G. (2026). *TU/e Cardiac RF Multi-Transmit* [Data set]. Eindhoven University of Technology, OpenH-RF. +> Penninga, S. W., & van Sloun, R. J. G. (2026). *TU/e Cardiac RF Multi-Transmit* [Data set]. Eindhoven University of Technology, Open-H-RF. ```bibtex @misc{penninga_tue_cardiac_plax_2026, @@ -186,13 +186,13 @@ Suggested citation for the dataset: author = {Penninga, Simon W. and van Sloun, Ruud J. G.}, year = {2026}, publisher = {Biomedical Diagnostics Lab, Eindhoven University of Technology}, - howpublished = {OpenH-RF dataset}, + howpublished = {Open-H-RF dataset}, } ``` ## Dataset Quantification -**Current OpenH-RF release:** 12 HDF5 files; 199.18 GB (199,180,025,844 bytes) stored; root `zea_version` **0.1.4**. +**Current Open-H-RF release:** 12 HDF5 files; 199.18 GB (199,180,025,844 bytes) stored; root `zea_version` **0.1.4**. - **Participants/files:** 12 distinct people represented by 12 HDF5 files. - **Tracks/acquisitions:** 8 tracks per file; 96 acquisitions total. - **Frames:** 100 per track; 9,600 frames total. diff --git a/datasets/tue-cardiac/reconstruct.py b/datasets/tue-cardiac/reconstruct.py index 44dd32f..087596a 100644 --- a/datasets/tue-cardiac/reconstruct.py +++ b/datasets/tue-cardiac/reconstruct.py @@ -1,8 +1,8 @@ #!/usr/bin/env python3 # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the tue-cardiac dataset of OpenH-RF. +"""Example reconstruction script for the tue-cardiac dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/tue-cardiac +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/tue-cardiac B-mode reconstruction of one track and frame of multi-transmit cardiac channel data. @@ -50,10 +50,10 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/tue-cardiac/data/subject-012.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/tue-cardiac/data/subject-012.hdf5" TRACK = "focused_fund" # one of TRACKS; must match CONFIG below FRAME = 0 -CONFIG = "hf://nvidia/OpenH-RF/tue-cardiac/pipelines/pipeline.yaml" +CONFIG = "hf://nvidia/Open-H-RF/tue-cardiac/pipelines/pipeline.yaml" OUT = HERE / "assets" / f"{Path(ZEA_FILE).stem}_{TRACK}_frame-{FRAME:03d}.png" DEVICE = None # e.g. cpu, cuda:0, auto:1 diff --git a/datasets/tue-carotid/README.md b/datasets/tue-carotid/README.md index 4149ca1..5ef0862 100644 --- a/datasets/tue-carotid/README.md +++ b/datasets/tue-carotid/README.md @@ -8,7 +8,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - carotid - in-vivo language: @@ -20,7 +20,7 @@ language: ![Longitudinal view of a carotid bifurcation](assets/5_long_bifur_R_0000.gif) -*One cardiac cycle of a longitudinal bifurcation scan, [`data/5_long_bifur_R_0000.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tue-carotid/data/5_long_bifur_R_0000.hdf5).* +*One cardiac cycle of a longitudinal bifurcation scan, [`data/5_long_bifur_R_0000.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tue-carotid/data/5_long_bifur_R_0000.hdf5).* ## Dataset Description @@ -76,11 +76,11 @@ The acquisitions can be processed with the `pipeline.yaml` definition in this fo ```bash zea process \ - --dataset hf://nvidia/OpenH-RF/tue-carotid/data/5_long_bifur_R_0000.hdf5 \ - --config hf://nvidia/OpenH-RF/tue-carotid/pipeline.yaml + --dataset hf://nvidia/Open-H-RF/tue-carotid/data/5_long_bifur_R_0000.hdf5 \ + --config hf://nvidia/Open-H-RF/tue-carotid/pipeline.yaml ``` -Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/tue-carotid/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF). +Alternatively, you can use the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/tue-carotid/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF). ## Dataset Format @@ -88,7 +88,7 @@ Alternatively, you can use the `reconstruct.py` [script](https://github.com/open ## Dataset Quantification -**OpenH-RF release:** 80 HDF5 files; 435.40 GB (435,403,030,528 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Open-H-RF release:** 80 HDF5 files; 435.40 GB (435,403,030,528 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Samples / frames / acquisitions:** 80 acquisitions (one HDF5 file each) from 10 subjects, 150 frames per acquisition — 12,000 frames total. - **Transmit events per frame:** 149 — 128 focused lines interleaved with 21 plane waves. diff --git a/datasets/tue-carotid/reconstruct.py b/datasets/tue-carotid/reconstruct.py index 662bd57..5781013 100644 --- a/datasets/tue-carotid/reconstruct.py +++ b/datasets/tue-carotid/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the tue-carotid dataset of OpenH-RF. +"""Example reconstruction script for the tue-carotid dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/tue-carotid +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/tue-carotid B-mode reconstruction of in-vivo carotid channel data. @@ -33,8 +33,8 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/tue-carotid/data/5_long_bifur_R_0000.hdf5" -CONFIG = "hf://nvidia/OpenH-RF/tue-carotid/pipeline.yaml" +ZEA_FILE = "hf://nvidia/Open-H-RF/tue-carotid/data/5_long_bifur_R_0000.hdf5" +CONFIG = "hf://nvidia/Open-H-RF/tue-carotid/pipeline.yaml" FRAME = 0 OUT = HERE / "assets" / "zea_carotid_2023_sample.png" diff --git a/datasets/tumunich/README.md b/datasets/tumunich/README.md index 08632aa..5acc4cf 100644 --- a/datasets/tumunich/README.md +++ b/datasets/tumunich/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - 3d - robotic - freehand @@ -31,7 +31,7 @@ size_categories: -*Acquisition setup (left), a single-frame reconstruction (middle) and a panoramic reconstruction compounded along the robot-tracked sweep (right) of the CIRS 054GS phantom, [`data/cirs_phantom/synth_apert_sweep_1.hdf5`](https://huggingface.co/datasets/nvidia/OpenH-RF/blob/main/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5).* +*Acquisition setup (left), a single-frame reconstruction (middle) and a panoramic reconstruction compounded along the robot-tracked sweep (right) of the CIRS 054GS phantom, [`data/cirs_phantom/synth_apert_sweep_1.hdf5`](https://huggingface.co/datasets/nvidia/Open-H-RF/blob/main/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5).* ## Dataset Description @@ -72,9 +72,9 @@ Primary task: **robotic tracked ultrasound acquisition** — compounding the per ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/tumunich/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/tumunich/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. -`reconstruct.py` beamforms a single frame and compares it to the stored Verasonics B-mode; [`reconstruct_panorama.py`](https://github.com/open-h/OpenH-RF/blob/main/datasets/tumunich/reconstruct_panorama.py), with `pipeline_panorama.yaml`, compounds a tracked sweep into the panoramic reconstruction shown above, using the per-frame robot poses. +`reconstruct.py` beamforms a single frame and compares it to the stored Verasonics B-mode; [`reconstruct_panorama.py`](https://github.com/open-h/Open-H-RF/blob/main/datasets/tumunich/reconstruct_panorama.py), with `pipeline_panorama.yaml`, compounds a tracked sweep into the panoramic reconstruction shown above, using the per-frame robot poses. ## Dataset Format @@ -88,7 +88,7 @@ Zero-valued rows pad the top of each stored frame so the image starts at the tra ## Dataset Quantification -**Current OpenH-RF release:** 11 HDF5 files; 62.25 GB (62,245,175,296 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 11 HDF5 files; 62.25 GB (62,245,175,296 bytes) stored; root `zea_version` **0.1.4**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - **Phantoms:** 3 (vascular access arm phantom, CIRS 054GS, CIRS 074 thyroid) - **Files:** 11 zea HDF5 files across 3 sub-dataset folders diff --git a/datasets/tumunich/reconstruct.py b/datasets/tumunich/reconstruct.py index 2a7d300..f05f5b2 100644 --- a/datasets/tumunich/reconstruct.py +++ b/datasets/tumunich/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the tumunich dataset of OpenH-RF. +"""Example reconstruction script for the tumunich dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/tumunich +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/tumunich B-mode reconstruction of walking-aperture robotic ultrasound sweep channel data. @@ -50,7 +50,7 @@ HERE = Path(__file__).parent CONFIG = HERE / "pipeline.yaml" # written by build_pipeline(); this is what the run loads -HF_CONFIG = "hf://nvidia/OpenH-RF/tumunich/pipeline.yaml" # where CONFIG is published +HF_CONFIG = "hf://nvidia/Open-H-RF/tumunich/pipeline.yaml" # where CONFIG is published PARAMETERS = { @@ -61,7 +61,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -ZEA_FILE = "hf://nvidia/OpenH-RF/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5" +ZEA_FILE = "hf://nvidia/Open-H-RF/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5" OUT = HERE / "assets" / f"{Path(ZEA_FILE).stem}_reconstructed.png" FRAME = 0 # Zero-based frame index to reconstruct DYNAMIC_RANGE = [-40, 0] # dB range shown @@ -73,7 +73,7 @@ def coords_to_imshow_mm(coords): - """openh-rf per-pixel coordinates (z, x, 3), last axis [x, y, z] in metres + """open-h-rf per-pixel coordinates (z, x, 3), last axis [x, y, z] in metres -> mpl imshow extent [left, right, bottom, top] in mm.""" x = coords[..., 0] z = coords[..., 2] @@ -81,7 +81,7 @@ def coords_to_imshow_mm(coords): # These ops are defined here, not in zea: a pipeline.yaml naming them resolves -# only once this module is imported. See https://github.com/open-h/OpenH-RF +# only once this module is imported. See https://github.com/open-h/Open-H-RF @ops_registry("power_compress") class PowerCompress(Operation): """Power (gamma) compression: raises the normalised envelope to ``exponent``.""" diff --git a/datasets/tumunich/reconstruct_panorama.py b/datasets/tumunich/reconstruct_panorama.py index f80f6a8..0e73ace 100644 --- a/datasets/tumunich/reconstruct_panorama.py +++ b/datasets/tumunich/reconstruct_panorama.py @@ -1,5 +1,5 @@ # SPDX-License-Identifier: Apache-2.0 -"""Extended-field-of-view reconstruction for the tumunich dataset of OpenH-RF. +"""Extended-field-of-view reconstruction for the tumunich dataset of Open-H-RF. The robot translates the probe 70 mm along its own lateral axis during the acquisition, so frames taken at different points in the sweep see overlapping @@ -43,7 +43,7 @@ # --- Inputs ----------------------------------------------------------------- # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. -INPUT = "hf://nvidia/OpenH-RF/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5" +INPUT = "hf://nvidia/Open-H-RF/tumunich/data/cirs_phantom/synth_apert_sweep_1.hdf5" OUTPUT = None # PNG to write (default: _panorama.png next to the sample) N_FRAMES = 20 # Frames to compound, spread evenly over the travel BLEND_TAPER = 0.25 # Fraction of each frame's width ramped down at either edge diff --git a/datasets/twente-cavitation/README.md b/datasets/twente-cavitation/README.md index 925faf2..aaec9c5 100644 --- a/datasets/twente-cavitation/README.md +++ b/datasets/twente-cavitation/README.md @@ -7,7 +7,7 @@ task_categories: tags: - ultrasound - rf - - openh-rf + - open-h-rf - cavitation language: - en @@ -51,7 +51,7 @@ The dataset contains data over a large pressure range, from very low pressures u ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/twente-cavitation/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/twente-cavitation/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the `pipeline.yaml` definition in this folder and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. Set `ZEA_FILE` at the top of the script to pick an acquisition and `N_FRAMES` to set how many frames are averaged; the map is written to `assets/.png`. @@ -63,7 +63,7 @@ Set `ZEA_FILE` at the top of the script to pick an acquisition and `N_FRAMES` to ## Dataset Quantification -**Current OpenH-RF release:** 19 HDF5 files; 10.85 GB (10,850,533,376 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. +**Current Open-H-RF release:** 19 HDF5 files; 10.85 GB (10,850,533,376 bytes) stored; root `zea_version` **0.1.6**. Sizes include all HDF5 contents and use decimal units (MB = 10^6 bytes, GB = 10^9 bytes, TB = 10^12 bytes), not decoded-array memory or original-source download sizes. - 19 aquisitions of 400 frames each, totalling 7600 frames - No train/validation/test split is defined; all acquisitions are provided in full. diff --git a/datasets/twente-cavitation/reconstruct.py b/datasets/twente-cavitation/reconstruct.py index 34d6b58..30fb276 100644 --- a/datasets/twente-cavitation/reconstruct.py +++ b/datasets/twente-cavitation/reconstruct.py @@ -1,7 +1,7 @@ # SPDX-License-Identifier: Apache-2.0 -"""Example reconstruction script for the twente-cavitation dataset of OpenH-RF. +"""Example reconstruction script for the twente-cavitation dataset of Open-H-RF. -Dataset link: https://huggingface.co/datasets/nvidia/OpenH-RF/tree/main/twente-cavitation +Dataset link: https://huggingface.co/datasets/nvidia/Open-H-RF/tree/main/twente-cavitation Passive acoustic map (PAM) reconstruction of one cavitation acquisition. @@ -78,12 +78,12 @@ # Defaults stream straight from the published corpus. Swap any of these for a # local path to run against your own copy. ZEA_FILE = ( - "hf://nvidia/OpenH-RF/twente-cavitation/data/cavitation_bubbles_1000kPa_01mL_per_min.hdf5" + "hf://nvidia/Open-H-RF/twente-cavitation/data/cavitation_bubbles_1000kPa_01mL_per_min.hdf5" ) CONFIG = HERE / "pipeline.yaml" # written by write_config(); this is what the run loads N_FRAMES = 10 # number of frames to average OUT = HERE / "assets" / f"{Path(ZEA_FILE).stem}.png" -HF_CONFIG = "hf://nvidia/OpenH-RF/twente-cavitation/pipeline.yaml" # where CONFIG is published +HF_CONFIG = "hf://nvidia/Open-H-RF/twente-cavitation/pipeline.yaml" # where CONFIG is published def build_pipeline() -> Pipeline: diff --git a/datasets/twente-microbubblesim/README.md b/datasets/twente-microbubblesim/README.md index 899fa4e..4a9df69 100644 --- a/datasets/twente-microbubblesim/README.md +++ b/datasets/twente-microbubblesim/README.md @@ -9,7 +9,7 @@ tags: - rf - microbubbles - synthetic-data - - openh-rf + - open-h-rf language: - en size_categories: @@ -56,7 +56,7 @@ The dataset is intended for studying the effect of transmit waveforms on deep-le ## Processing the Dataset -The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/OpenH-RF/blob/main/datasets/twente-microbubblesim/reconstruct.py) as provided in the [OpenH-RF GitHub repository](https://github.com/open-h/OpenH-RF), together with the pipeline definitions in `pipeline/` and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. +The acquisitions can be processed with the `reconstruct.py` [script](https://github.com/open-h/Open-H-RF/blob/main/datasets/twente-microbubblesim/reconstruct.py) as provided in the [Open-H-RF GitHub repository](https://github.com/open-h/Open-H-RF), together with the pipeline definitions in `pipeline/` and the [zea library](https://github.com/tue-bmd/zea). The script streams the data from the Hugging Face Hub. Every pulse track has its own verified pipeline file, `pipeline/pipeline_track__