3D Lung Tumour Segmentation

Overview

3D Lung Tumour Segmentation applies deep learning to medical imaging — specifically, segmenting lung tumours from volumetric CT scans using the Medical Segmentation Decathlon lung dataset.

The model is built with MONAI's 3D U-Net architecture and trained with PyTorch Lightning for structured, reproducible training.

Stack

• MONAI — medical imaging transforms and 3D U-Net
• PyTorch Lightning — training loop and checkpointing
• Hydra — configuration management via conf/config.yaml
• FastAPI — inference REST API (app/main.py)

Training

python -m venv venv && source venv/bin/activate
pip install -r requirements.txt
python train.py