MRI Processing Tools in MATLAB

This repository contains a collection of MATLAB functions designed to simplify common tasks when working with MRI data. These tools were developed by Harrison Brammell while working at the Auburn University MRI Research Center under the guidance of Dr. Adil Bashir.

---

Functions

---

Image Tools (image-tools/)

nii4to3d.m

Splits a 3D or 4D NIfTI file into a series of 3D NIfTI files, one per slice along the third dimension.

Usage:

nii3to4d(imagein)

Parameters:

• imagein: The input 3D or 4D NIfTI image data.

Details:

• Determines image size with size(imagein).
• Creates a new directory ./3d_niis/.
• Iterates through the third dimension (Z-slices) and saves each slice as slice_i.nii using niftiwrite.

---

dicomto4d.m

Searches through a directory for DICOM files and returns a 4D array of size [x, y, z, # of DICOMS].

Usage:

data = dicomto4d(path)
data = dicomto4d(path, order)

Parameters:

• path: Directory path containing DICOM files.
• order (optional): Array specifying the reordering of files for the fourth dimension. Indices correspond to the naturally sorted file order.

Returns:

• data: 4D array containing all DICOM images.

Details:

• Uses natural sorting to order DICOM files numerically (1.dcm, 2.dcm, etc.).
• If order parameter is provided, files are reordered according to the specified indices.
• Creates a 4D array based on the size of the first DICOM file.
• Files must be numbered in the order they should appear in the fourth dimension (unless reordered).
• Designed for DICOM files as produced by Siemens MRI scanners.

---

normClamp.m

Normalizes and clamps a 2D image to specified intensity ranges.

Usage:

normImg = normClamp(image, low, high)

Parameters:

• image: Input 2D image.
• low: Lower clamp value.
• high: Upper clamp value.

Returns:

• normImg: Normalized and clamped image with values between 0 and 1.

Details:

• Converts image to double precision.
• Clamps values below low to low and above high (or infinite) to high.
• Normalizes the result to the range [0, 1].

---

qMRLab Tools (qMRLab-tools/)

These tools require qMRLab v2.4.2 - My fork with specific modifications for work done at the Auburn University MRI Research Center.

process_nifti_exp_mono_t2.m

Processes a NIfTI file using the mono_T2 model with exponential fitting. Utilizes qMRLab v2.4.2.

Prerequisites:

• Must be in the working directory of qMRLab v2.4.2 (only tested version).
• Requires the MATLAB Parallel Computing Toolbox.

Usage:

process_nifti_exp_mono_t2(inputnii, inputmask, outputpath, et, t2, m0, snr)

Parameters:

• inputnii: Path to the input NIfTI file.
• inputmask (optional): Path to the NIfTI mask file.
• outputpath (optional): Directory path to save the output. Defaults to pwd if not specified.
• et (optional): Echo times as a column vector. Defaults to transpose([10:10:70]).
• t2, m0, snr (optional): Additional simulation parameters.

Details:

• Validation: Checks if inputnii exists.
• qMRLab Setup: Runs startup.m and initializes the mono_t2 model.
• Data Processing: Uses exponential fitting and parallel processing.
• Output: Saves fitted results and generates simulation plots (single voxel curve and sensitivity analysis).

---

process_nifti_lin_mono_t2.m

Processes a NIfTI file using the mono_T2 model with linear fitting. Similar to the exponential version but uses linear fitting approach.

Prerequisites:

• Must be in the working directory of qMRLab v2.4.2 (only tested version).
• Requires the MATLAB Parallel Computing Toolbox.

Usage:

process_nifti_lin_mono_t2(inputnii, inputmask, outputpath, et, t2, m0, snr)

Parameters:

• Same as process_nifti_exp_mono_t2.m

Details:

• Sets Model.options.FitType = 'Linear' for linear fitting approach.
• Otherwise identical to the exponential version in terms of workflow and outputs.

---

Deprecated Functions (deprecated/)

dicom2image.m

Converts a DICOM file and returns an image object.

Status: Work in progress / Untested

Usage:

[image] = dicom2image(filename, color)

Parameters:

• filename: Path to the DICOM file.
• color (optional): Colormap to apply to the image (e.g., 'gray', 'jet'). Defaults to 'gray'.

Returns:

• image: Handle to a hidden figure containing the displayed DICOM data.

Details:

• Reads the DICOM file using dicomread.
• Creates a hidden figure and displays the image with the specified colormap.
• Includes a colorbar for reference.

---

Disclaimer

These tools are provided as-is. Please ensure you have the necessary dependencies (e.g., qMRLab v2.4.2) installed and correctly configured for successful execution.

This README documentation was created with the assistance of AI.