petermcgor/conflunet repository overview

⁠ConfLUNet Docker Image

This Docker image provides a ready-to-use environment for ConfLUNet, a deep learning model that performs instance segmentation of multiple sclerosis lesions in FLAIR MRI images. The model is based on the nnUNet framework.

⁠Quick Start

To use this image, you need a CUDA-capable GPU with appropriate drivers and nvidia-docker installed.

Basic usage pattern:

docker run -it --gpus all --shm-size=40gb \
  -v /path/to/your/data:/opt/conflunet_resources \
  petermcgor/conflunet:0.2.0 [command]

⁠Prerequisites

Your system should have:

• NVIDIA GPU with CUDA support
• NVIDIA Container Toolkit installed
• Adequate disk space and RAM
• Shared memory allocation (recommended: 40GB, adjustable based on needs)

⁠Data Directory Structure

Your mounted volume must contain the following nnUNet directory structure:

/path/to/your/data/
├── nnUNet_raw/         # Raw data following nnUNet format
├── nnUNet_preprocessed/ # Preprocessed data
└── nnUNet_results/     # Training results

⁠Common Commands

Preprocessing data:

docker run -it --gpus all --shm-size=40gb \
  -v /path/to/your/data:/opt/conflunet_resources \
  petermcgor/conflunet:0.2.0 \
  conflunet_plan_and_preprocess --dataset_id [ID] --check_dataset_integrity

Training:

docker run -it --gpus all --shm-size=40gb \
  -v /path/to/your/data:/opt/conflunet_resources \
  petermcgor/conflunet:0.2.0 \
  conflunet_train --dataset_id [ID] --fold [FOLD] --model_name [NAME]

⁠Additional Information

For detailed documentation, dataset preparation guidelines, and model information, visit the ConfLUNet GitHub repository⁠.

⁠Version Information

• Base image: pytorch/pytorch:2.1.2-cuda11.8-cudnn8-runtime
• ConfLUNet version: 0.2.0
• PyTorch version: 2.1.2
• CUDA version: 11.8

⁠License

This project is licensed under the MIT License - see the LICENSE⁠ file in the GitHub repository for details.

⁠Support

For issues, questions, or contributions, please visit our GitHub repository⁠.