Physiolog

physiolog.org — An open clinical physiology resource for physical therapy students and clinicians.

Overview

Physiolog is built around Clinical Physiology: A Muscle Centered Approach — a textbook that begins at the muscle fiber, asks what it needs to generate tension, and treats circulation, respiration, and renal function as the systems that support movement. It is written primarily for physical therapy students and clinicians, and is also useful for exercise science, athletic training, and anyone studying human movement.

The site serves as:

1. Clinical Physiology Textbook — The textbook is authored in LaTeX (see book/). The published web chapters and downloadable PDF are pinned to the completed 3rd edition while 4th-edition revisions continue in the repository.

2. Interactive Simulations — Dynamic, browser-based physiology simulations that allow hands-on exploration of cardiovascular, respiratory, renal, and other physiological models.

3. AI Learning Assistant — A domain-specific NLP/LLM agent grounded in the textbook content and a curated research library, using retrieval-augmented generation (RAG) to support deep mechanistic understanding.

4. Integrative Research — Ongoing research into clinical physiology as the mechanistic foundation for clinical reasoning, including Models4PT and the Clinical Inference Engine (CIE).

Model and simulation ecosystem

Physiolog is the educational and question-forming end of a broader generative-mechanism research pipeline. Textbook concepts, teaching models, and clinically relevant questions will be translated into explicit computational mechanisms, tested across available simulation engines and public Physiome models, and developed into reproducible models for models4PT.

JSim is being modernized as part of this work. Its support for mixed ODE, PDE, implicit, integral, discrete-event, and procedural models; physical-unit checking; parameter estimation against experimental data; and CellML/SBML interchange make it a useful execution and validation environment alongside HumMod and browser-based teaching models. The goals are complementary:

• build and test generative physiological mechanisms for the research path from Physiolog to models4PT;
• use JSim, HumMod, public Physiome models, and other suitable engines to develop and cross-validate transparent models for the textbook, teaching simulations, and research; and
• preserve model provenance, assumptions, units, scenarios, and engine versions so agreement and disagreement can be interpreted rather than treating any engine as ground truth.

Tech Stack

• Next.js (App Router, TypeScript)
• Tailwind CSS
• Deployed to physiolog.org

Development

# Install dependencies
npm install

# Run development server
npm run dev

Open http://localhost:3000 in your browser.

See EDITING.md for how to edit the book and the site, and how to publish changes to physiolog.org.

Blood-pressure scenarios

Generate the ten fixed blood-pressure scenarios used by the arterial-compliance pilot:

npm run simulation:blood-pressure

The command prints a Markdown table containing each scenario's inputs, systolic and diastolic pressure, pulse pressure, mean arterial pressure, and cardiac output. It exits with an error if an output is non-finite or if the resting and exercise presets differ from their recorded expectations by more than 0.1 mmHg or 0.1 L/min. It also checks the protocol's expected one-factor response directions.

Textbook release policy

The LaTeX on main may contain work toward the 4th edition. Routine deployments restore the textbook source from the exact commit recorded as PUBLISHED_TEXTBOOK_REF in .github/workflows/deploy-pages.yml before generating public chapter HTML. The download link similarly names an editioned, committed PDF. Neither public format advances until a complete new edition is deliberately released; see EDITING.md for the steps.

Simulation development workspace

Simulation work that compares the public teaching models with other model environments should open the local ~/Projects/physiolog-simulations.code-workspace. A portable equivalent is committed at physiolog-simulations.code-workspace. Both workspace files load six sibling roots in one VS Code window:

• physiolog — public teaching simulations and site code.
• hummod-research — private experiment definitions, analysis, selected results, and provenance.
• hummod-standalone — external HumMod 1.6.1 distribution for read-only reference and offline execution; do not commit its files into either repository.
• JSim — model-development and simulation engine under compatibility-focused modernization. Its numerical behavior, model semantics, units, formats, and historical verification baselines must be preserved; see that repository's MODERNIZATION.md for current constraints and status.
• Models4PT — population-level causal knowledge, research models, provenance, and evidence integration for physical therapy and rehabilitation science.
• Clinical-Inference-Engine — the Clinical Inference Engine (CIE), which applies appropriate population knowledge to patient-specific explanatory and probabilistic reasoning while remaining separate from Models4PT.

From the Physiolog directory, open it with:

code ../physiolog-simulations.code-workspace

Start simulation-building agent sessions from a multi-root workspace when they need access to the relevant model repositories. A normal Physiolog-only window remains appropriate for textbook, site, and simulation-interface work that does not need the broader research and clinical-reasoning ecosystem.

Licensing

The website and supporting software in this repository are licensed under the MIT License; see LICENSE. The textbook content in book/, including its text and LaTeX source, remains under the Creative Commons BY-NC-SA 4.0 license stated on the book's copyright page. This is a deliberate split: the MIT license does not grant commercial rights to the textbook content.

Figures created with BioRender may have additional restrictions under the applicable BioRender account license. See the copyright page in book/book.tex before reusing book figures.

Use of AI

The textbook content is authored by Sean Collins; no part of it has been written by AI. The website and supporting software were built with AI assistance (GitHub Copilot). See AI-USE.md for details.

Project Structure

book/                     # LaTeX source for the textbook edition under development
  book.tex                # Root file for the PDF build
  latexml-book.tex        # Root file for the HTML build
  chapter/                # Chapter sources
scripts/
  build-textbook-html.sh  # LaTeX -> per-chapter HTML (npm run build:textbook-html)
src/
  app/
    page.tsx              # Landing page
    layout.tsx            # Root layout with navigation
    textbook/             # Textbook pages
      page.tsx            # Table of contents
      chapters.ts         # Chapter/part structure
      chapter/[chapter]/  # Individual chapters
    simulations/          # Interactive simulations
    ai-assistant/         # AI assistant interface
    research/             # Research page

Contributing

This is a personal academic project. If you are interested in collaborating, please open an issue or contact via the site.