MVO-identification

This repository contains the source code for the MICCAI 2024 paper:
"Coarse-Grained Mask Regularization for Microvascular Obstruction Identification from non-contrast Cardiac Magnetic Resonance".

Environment Setup

This codebase is tested with the following environment:

• CUDA: 11.0
• PyTorch: 1.7.1
• MMSegmentation: 0.13.0
• Python: 3.8+

To ensure compatibility, it is recommended to use the above versions.

Multi-GPU Training

To train a model with multiple GPUs (e.g., 4 GPUs), use the following command:

./tools/dist_train.sh local_configs/your_model_config.py 4 --work-dir "path/to/save/weights"

• Replace local_configs/your_model_config.py with the path to your model configuration file.
• Replace 4 with the number of GPUs you want to use.
• Replace "path/to/save/weights" with the directory where you want to save the model weights and training logs.

Inference

To run inference on a single GPU:

./tools/dist_test.sh  local_configs/your_model_config.py "your_weight_path" 4 --out "path/to/save/weights" 

• Replace local_configs/your_model_config.py with the path to your model configuration file.
• Replace 4 with the number of GPUs you want to use.
• Replace "path/to/save/weights" with the directory where you want to save the model weights and training logs.