TractoEmbed: Modular Multi-level Embedding framework for white matter tract segmentation

TL;DR

TractoEmbed is a new modular multi-level embedding framework that improves white matter tract segmentation by capturing detailed spatial information and addressing common challenges like class imbalance and structural similarity.

Contribution

It introduces a hierarchical streamline data representation and a modular framework that outperforms existing methods in tract segmentation tasks.

Findings

Outperforms state-of-the-art segmentation methods

Effective across different datasets and age groups

Modular design allows easy integration of additional embeddings

Abstract

White matter tract segmentation is crucial for studying brain structural connectivity and neurosurgical planning. However, segmentation remains challenging due to issues like class imbalance between major and minor tracts, structural similarity, subject variability, symmetric streamlines between hemispheres etc. To address these challenges, we propose TractoEmbed, a modular multi-level embedding framework, that encodes localized representations through learning tasks in respective encoders. In this paper, TractoEmbed introduces a novel hierarchical streamline data representation that captures maximum spatial information at each level i.e. individual streamlines, clusters, and patches. Experiments show that TractoEmbed outperforms state-of-the-art methods in white matter tract segmentation across different datasets, and spanning various age groups. The modular framework directly allows the integration of additional embeddings in future works.