Skip to content

Latest commit

 

History

History
38 lines (23 loc) · 1010 Bytes

README.md

File metadata and controls

38 lines (23 loc) · 1010 Bytes

MLNet🔍

An Automated Deep Learning Pipeline for EMVI and Response Prediction of Rectal Cancer Using Baseline MRI

Paper Link: https://www.nature.com/articles/s41698-024-00516-x

Installation💻

in your favourite virtual environment:

pip install -r requirements.txt

Training🚀

Firstly, you have to run a nnunet link: https://github.com/MIC-DKFZ/nnUNet/tree/nnunetv1

Then, extract stage features (you can use multifeature_extractor.py)

Then, you need to have a csv file where you have your classification fold information and id. Also, you need to have a label csv file where you store your targeted labels

Also, you can finetune hyperparameters in the config file

Then you can Train with the following command:

CUDA_VISIBLE_DEVICES=0 python trainer.py --load_json config/nnunet.json

Evaluation⚡️

TEST with:

CUDA_VISIBLE_DEVICES=0 python eval_test.py