Habitat Analysis: Biomedical Imaging Toolkit (HABIT)

Tumor habitat analysis and intratumoral heterogeneity quantification for clinical and radiomics research. Workflows are driven by YAML configs: preprocessing, habitat segmentation, feature extraction, and optional machine learning.

Language / 语言：English | 简体中文

Documentation

Online docs: https://lichao312214129.github.io/HABIT

Local build: cd docs && make html → docs/build/html/index.html

Suggested learning path

Workflow chapters

Tools & more

Bundled config templates

After cloning or unpacking the repo, use the config/ folder at the project root (sibling to the habit/ Python package). See config/README_CONFIG.md and Configuration reference.

Install & demo data

Two supported install methods (full steps: Installation). Python 3.10–3.14 (numpy 1.26 and 2.x).

• (A) pip (Miniconda / venv / etc.)

  pip install habitat-analysis
  habit --version
  # import name remains: import habit
  

• (B) from Git source

  git clone https://github.com/lichao312214129/HABIT.git
  cd HABIT
  pip install .
  # editable: pip install -e .
  

PyRadiomics is not a default dependency. When you need radiomics features:

pip install "habitat-analysis[radiomics]"
python -m habit.install_radiomics
python -c "import radiomics; print(radiomics.__version__)"

On Windows this installs the matching prebuilt wheel from the HABIT GitHub Release (avoids the broken PyPI sdist). On macOS / Linux it installs pyradiomics from PyPI. Other extras: pip install "habitat-analysis[ml,analysis,registration]".

• Source: GitHub
• Demo data: Quickstart

Support & citation

• Issues: GitHub Issues
• Citation: see CITATION.cff
• License: Apache License 2.0. Free for academic and commercial use; the only obligation is to retain the copyright and license notices and to ship NOTICE with redistributions. When HABIT supports scientific work, the authors request -- but do not require as a license condition -- that you cite it