Add application: Brain Glioma Segmenter

[dani-capellan]

brain_glioma_segmenter

Description

The Brain Glioma Segmenter is a software tool developed by researchers from Universidad Politécnica de Madrid and Children’s National Hospital for the segmentation and analysis of brain gliomas in magnetic resonance imaging (MRI). Built in Python, it enables precise quantitative analysis of brain gliomas in MRI scans to support clinical decision-making in both diagnosis and prognosis.

Usage

Folder Structure Overview
The input folder should follow a similar structure as shown below including the naming convention. Do not add any other files or folders into the input directory. XX can be anything. Ensure that the image is preprocessed using the same pipeline and steps applied in the corresponding BraTS 2024 subtask preprocessing.

input_folder/
│
└───XX-XX-XX-XX
    │   XX-XX-XX-XX-t1n.nii.gz
    │   XX-XX-XX-XX-t1c.nii.gz
    │   XX-XX-XX-XX-t2w.nii.gz
    │   XX-XX-XX-XX-t2f.nii.gz

Running Inference
The application accepts 2 positional parameters:

• The input path of a dataset directory (following the structure above).
• The output directory where the segmentation results will be saved.

Command:
jobman submit -i brain-glioma-segmenter -- <INPUT_DIR> <OUTPUT_DIR>

Example:

jobman submit -i brain-glioma-segmenter -- ~/datasets/87f3be56-4725-45c3-9baa-d338de530f73/ ~/persistent-home/results/glioma_seg_output/

Note: The output directory path should be any within the persistent-home (~/persistent-home/), which is shared between all desktops and jobs created by the user; otherwise, the results will be lost after the end of the job. (This tool requires significant resources; ensure your jobman environment is configured to allocate GPUs and at least 16GB of memory if required by your node's setup).

Authors

Daniel Capellán-Martín
Abhijeet Parida
Zhifan Jiang
BIT-UPM
PAI-CNMC-CNH

Contact info

daniel.capellan@upm.es
pabhijeet@childrensnational.org
zjiang@childrensnational.org
mj.ledesma@upm.es
mlingura@childrensnational.org
https://github.com/Pediatric-Accelerated-Intelligence-Lab/BraTS2024-PEDS
https://github.com/Pediatric-Accelerated-Intelligence-Lab/BraTS2023-inferCode

URL

Public dockerfile repository:
https://harbor.eucaim.cancerimage.eu/harbor/projects/4/repositories/brain-glioma-segmenter/info-tab

License

https://spdx.org/licenses/CC-BY-NC-SA-4.0

Visibility of application

Public