Quick Start#
Choose a track and use its own environment. Run the commands from the corresponding software repository root, not this documentation repository. The linked public repository files contain these commands; confirm the session environment with the organisers before running them.
Cancer Quick Start#
Repository: KaleCancer. Use Python 3.11 or 3.12 and uv. From a local checkout, follow the repository quickstart:
uv sync --all-extras
uv run python examples/synthetic/synthetic_survival.py
The synthetic example generates clinical variables, pathology patch features and censored survival times, then trains an intermediate-fusion Cox model. It is a CPU-capable, no-download starter exercise.
Continue with Cancer Getting Started, the Hands-on Tutorial, or the Live-Coding Notebook. The live-coding notebook imports the accompanying synthetic_data_generation.py; review its access notes before attempting local or Colab execution.
Cardiac Quick Start#
Repository: KaleCardiac. Use Python 3.10, 3.11 or 3.12. Install a PyTorch build compatible with your hardware, then follow the repository quickstart:
git clone https://github.com/pykale/cardiac.git
cd cardiac
pip install -e .
python -c "import kalecardiac, torch, kale; print(kalecardiac.__version__)"
The core install includes PyKale. Optional extras cover ECG file formats (.[ecg]), imaging (.[imaging]), interpretation (.[interpret]) and development (.[dev]). The starter examples need no participant dataset.
CardioVAE: chest X-ray + ECG#
python -m examples.cardiovae_multimodal.main --cfg examples/cardiovae_multimodal/configs/quick.yaml
LS-EMVAE: multi-lead ECG#
python -m examples.lsemvae_ecg.main --cfg examples/lsemvae_ecg/configs/quick.yaml
These examples can run on CPU and save predictions, metrics and configuration records under outputs/. Runtime depends on hardware and dependencies. Avoid mixing the local checkout with an older installed package.
Important
Synthetic runs check the pipeline; their scores are not clinical results. Use real clinical data only with appropriate permissions and secure infrastructure, and never upload identifiable or confidential data to public services.