Full-Head Segmentation of MRI with Abnormal Brain Anatomy: Model and Data Release

2025-01-30

AI Summary

Conventional brain MRI segmentation methods frequently misclassify abnormal brain tissues—e.g., in traumatic brain injury or stroke—as cerebrospinal fluid (CSF), leading to critical failures in whole-head anatomical modeling. Method: We propose the first deep learning framework for 3D whole-head segmentation into seven anatomical classes (skin, skull, CSF, gray matter, white matter, etc.), featuring a novel MultiAxial architecture that replaces atlas-based registration with multi-planar feature fusion: three independent 2D U-Nets process sagittal, axial, and coronal slices, followed by feature-level 3D fusion. Annotations combine automated initialization with expert manual refinement to ensure high fidelity. Contribution/Results: We release the first publicly available whole-head MRI segmentation dataset (61 clinical cases). Our model achieves a median Dice score of 0.88 overall (0.898 for brain tissues), significantly outperforming SynthSeg and BrainChop. The model is integrated into the ROAST toolbox, substantially improving transcranial electrical stimulation modeling accuracy.

Technology Category

• Computer Vision: Segmentation
• Humans and AI: Brain-Sensing and Analysis
• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

The goal of this work was to develop a deep network for whole-head segmentation, including clinical MRIs with abnormal anatomy, and compile the first public benchmark dataset for this purpose. We collected 91 MRIs with volumetric segmentation labels for a diverse set of human subjects (4 normal, 32 traumatic brain injuries, and 57 strokes). These clinical cases are characterized by extended cerebrospinal fluid (CSF) in regions normally containing the brain. Training labels were generated by manually correcting initial automated segmentations for skin/scalp, skull, CSF, gray matter, white matter, air cavity, and extracephalic air. We developed a MultiAxial network consisting of three 2D U-Net models that operate independently in sagittal, axial, and coronal planes and are then combined to produce a single 3D segmentation. The MultiAxial network achieved test-set Dice scores of 0.88 (median plus-minus 0.04). For brain tissue, it significantly outperforms existing brain segmentation methods (MultiAxial: 0.898 plus-minus 0.041, SynthSeg: 0.758 plus-minus 0.054, BrainChop: 0.757 plus-minus 0.125). The MultiAxial network gains in robustness by avoiding the need for coregistration with an atlas. It performed well in regions with abnormal anatomy and on images that have been de-identified. It enables more robust current flow modeling when incorporated into ROAST, a widely-used modeling toolbox for transcranial electric stimulation. We are releasing a state-of-the-art model for whole-head MRI segmentation, along with a dataset of 61 clinical MRIs and training labels, including non-brain structures. Together, the model and data may serve as a benchmark for future efforts.

Problem

• Advanced Computer Model
• MRI Image Analysis
• Medical Diagnosis Assistance

Innovation

• MultiAxial Model
• MRI Image Analysis
• ROAST Tool Enhancement