FLAIRBrainSeg: Fine-grained brain segmentation using FLAIR MRI only

2025-04-04

AI Summary

This work addresses the challenge of fine-grained brain parcellation in the absence of T1-weighted MRI, a common clinical limitation. We propose the first end-to-end deep learning method capable of automatically segmenting 132 anatomical structures solely from FLAIR MRI—bypassing conventional image synthesis pipelines. Instead of synthesizing missing T1 data, our approach directly learns the mapping from FLAIR to high-resolution segmentation, incorporating T1-guided pseudo-label supervision and lesion-aware feature fusion to enhance cross-domain robustness. Evaluated on multi-center multiple sclerosis patient data, our method significantly outperforms existing modality-agnostic synthesis-based approaches, achieving an average Dice score improvement of 6.2%. It maintains high accuracy both in-domain and out-of-domain, demonstrating strong generalizability. This establishes a reliable, deployable paradigm for clinical neuroimaging analysis under T1-absent conditions.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper introduces a novel method for brain segmentation using only FLAIR MRIs, specifically targeting cases where access to other imaging modalities is limited. By leveraging existing automatic segmentation methods, we train a network to approximate segmentations, typically obtained from T1-weighted MRIs. Our method, called FLAIRBrainSeg, produces segmentations of 132 structures and is robust to multiple sclerosis lesions. Experiments on both in-domain and out-of-domain datasets demonstrate that our method outperforms modality-agnostic approaches based on image synthesis, the only currently available alternative for performing brain parcellation using FLAIR MRI alone. This technique holds promise for scenarios where T1-weighted MRIs are unavailable and offers a valuable alternative for clinicians and researchers in need of reliable anatomical segmentation.

Problem

• Segments brain using only FLAIR MRI scans
• Targets limited access to other imaging modalities
• Robust for multiple sclerosis lesions analysis

Innovation

• Uses only FLAIR MRI for brain segmentation
• Trains network to approximate T1-weighted segmentations
• Robust to multiple sclerosis lesions