BrainLesion Suite: A Flexible and User-Friendly Framework for Modular Brain Lesion Image Analysis

TL;DR

BrainLesion Suite is a flexible, Python-based toolkit that simplifies the creation of modular brain lesion image analysis pipelines, integrating advanced algorithms for preprocessing, synthesis, and performance quantification.

Contribution

It introduces a user-friendly, modular framework that streamlines brain lesion image analysis workflows, incorporating algorithms from the BraTS challenge for missing modality synthesis and lesion inpainting.

Findings

Provides a versatile preprocessing module for multi-modal images

Enables synthesis of missing modalities and lesion inpainting

Includes tools for quantifying segmentation performance

Abstract

BrainLesion Suite is a versatile toolkit for building modular brain lesion image analysis pipelines in Python. Following Pythonic principles, BrainLesion Suite is designed to provide a 'brainless' development experience, minimizing cognitive effort and streamlining the creation of complex workflows for clinical and scientific practice. At its core is an adaptable preprocessing module that performs co-registration, atlas registration, and optional skull-stripping and defacing on arbitrary multi-modal input images. BrainLesion Suite leverages algorithms from the BraTS challenge to synthesize missing modalities, inpaint lesions, and generate pathology-specific tumor segmentations. BrainLesion Suite also enables quantifying segmentation model performance, with tools such as panoptica to compute lesion-wise metrics. Although BrainLesion Suite was originally developed for image analysis pipelines of brain lesions such as glioma, metastasis, and multiple sclerosis, it can be adapted for other biomedical image analysis applications. The individual BrainLesion Suite packages and tutorials are accessible on GitHub.