The ISLES'22 Ensemble Algorithm

Introduction

Algorithm to predict ischemic stroke lesions from MRI data.

This repository contains an ensemble algorithm devised in the context of the ISLES'22 MICCAI Challenge (https://isles22.grand-challenge.org/). The top-3 leading algorithms are used in a majority-voting scheme to predict outputs.

1. Installation
2. Usage
3. Get started
4. Citation
5. About the Ensembled Algorithms
6. Questions
7. Acknowledgement

Installation

1.1) In your conda environment 'myenv' install basic pytorch, torchvision and cudatoolkit.

Note: Mandatory Python version 3.8.0

conda activate myenv
conda install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=11.3 -c pytorch

1.2) Clone this repository.

cd ~
git clone https://github.com/Tabrisrei/ISLES22_Ensemble.git

1.3) Install dependencies for the ensemble algorithm .

cd ISLES22_Ensemble
pip install -e SEALS/
pip install -e FACTORIZER/model/factorizer/
pip install -r requirements.txt

If successfully installed all required packages, you can follow the steps below to download and place the checkpoints.

1.4) Download all content from https://drive.google.com/drive/folders/1GQP5nmtoIfhyu2LHCzPVYnZZ7ZQDX1go?usp=drive_link

1.5)

• Decompress the file "nnUNet_trained_models.tar.gz" in ~/ISLES22_Ensemble/SEALS/data/.
• Decompress the file "ts.tar.gz" in ~/ISLES22_Ensemble/NVAUTO/.
• Decompress "log.tar.gz" in ~/ISLES22_Ensemble/FACTORIZER/model/.

Usage

The algorithms works over skull-stripped MRI images, directly in native-space (no image co-registration is needed). Image modalities required for running the algorithm are DWI (b=1000), ADC and FLAIR.

Note: DWI and ADC images are supposed to be obtained from the same acquisition timepoint. Differences between them (e.g. affine, voxel resolution, number of voxels, etc.) might impact the algorithm performance.

From Python

import sys
ENSEMBLE_PATH = 'path-to-isles-ensemble-repo' 
sys.path.append(ENSEMBLE_PATH)
from isles22_ensemble import predict_ensemble

INPUT_FLAIR = 'path-to-flair.nii.gz'
INPUT_ADC = 'path-to-adc.nii.gz'
INPUT_DWI = 'path-to-dwi.nii.gz'
OUTPUT_PATH = 'path-to-output-folder'

predict_ensemble(isles_ensemble_path=ENSEMBLE_PATH,
                     input_dwi_path=INPUT_DWI,
                    input_adc_path=INPUT_ADC,
                    input_flair_path=INPUT_FLAIR,
                    output_path=OUTPUT_PATH,
                    save_team_outputs=False)

From terminal

Place your MRI images (.nii.gz) in

~/ISLES22_Ensemble/input/images/adc-brain-mri/
~/ISLES22_Ensemble/input/images/dwi-brain-mri/
~/ISLES22_Ensemble/input/images/flair-brain-mri/

Run:

bash Ensemble_launcher.sh

Results will be stored in ~/ISLES22_Ensemble/output/images/stroke-lesion-segmentation/.

Get started

You can try out our algorithm using a real test scan located in ~/data. For doing so, run the example script located in ~/scripts/predict_scan.py. The example data belongs to the ISLES'22 dataset (Hernandez Petzsche et al., Sci Data 2022).

Citation

Please cite the following manuscripts when using The Isles'22 Ensemble:

• de la Rosa, E., Reyes, M., Liew, S. L., Hutton, A., Wiest, R., Kaesmacher, J., ... & Wiestler, B. (2024). A Robust Ensemble Algorithm for Ischemic Stroke Lesion Segmentation: Generalizability and Clinical Utility Beyond the ISLES Challenge. arXiv preprint arXiv:2403.19425.

• Hernandez Petzsche, M. R., de la Rosa, E., Hanning, U., Wiest, R., Valenzuela, W., Reyes, M., ... & Kirschke, J. S. (2022). ISLES 2022: A multi-center magnetic resonance imaging stroke lesion segmentation dataset. Scientific data, 9(1), 762.

About the Ensembled Algorithms

• Algorithm SEALS is based on nnUnet. Git repo https://github.com/Tabrisrei/ISLES22_SEALS

• Algorithm NVAUTO is based on MONAI Auto3dseg. Git repo: https://github.com/mahfuzmohammad/isles22

• Algorithm SWAN is based on FACTORIZER. Git repo: https://github.com/pashtari/factorizer-isles22

Questions

Please contact Ezequiel de la Rosa ([email protected]) or Shengbo Gao ([email protected]).

Acknowledgement

We thank all ISLES'22 challenge participants, collaborators and organizers for allowing this work to happen. We also thank all developers and maintaners of the repositories named below for sharing such valuable resources.

• The code of Team SEALS is adapted from nnUNet.
• The code of NVAUTO is adapted from MONAI
• The code of SWAN is adapted from FACTORIZER