SIAM: Head and Brain MRI Segmentation from Few High-Quality Templates via Synthetic Training

TL;DR

SIAM is a novel 3D brain and head MRI segmentation model trained on only six high-quality templates, using synthetic data to accurately segment multiple structures across various imaging modalities.

Contribution

It introduces a synthetic training approach that reduces template requirements and extends segmentation to both brain and extra-cerebral tissues.

Findings

Matches or outperforms state-of-the-art methods in multiple datasets

Extends automated segmentation to non-brain structures

Shows superior consistency and sensitivity across contrasts

Abstract

Synthetic training has recently advanced brain MRI segmentation by enabling contrast-agnostic models trained entirely on generated data. However, most existing approaches rely on hundreds of automatically labeled templates, introducing systematic biases and limiting their flexibility to incorporate new anatomical structures. We present the Segment It All Model (SIAM), a 3D whole-head segmentation framework for 16 anatomical structures, trained using only six high-quality, manually annotated templates. SIAM extends domain randomization to both intensity and shape domains: synthetic image generation ensures contrast variability, while high-resolution spatial transformations model anatomical differences in cortical thickness and deep nuclei morphology. Unlike prior synthetic models, SIAM simultaneously segments brain as well as extra-cerebral tissues, including cerebrospinal fluid, vessels, dura mater, skull, and skin, enabling fully automated, preprocessing-free analysis. Evaluation across eight heterogeneous datasets (N=301), that include multiple contrasts (T1-weighted, T2-weighted, CT) and span a wide range of ages, demonstrates that SIAM matches or outperforms state-of-the-art methods for brain structures, in addition to extending automated segmentation to non-brain structures. The model also exhibits superior consistency across contrasts and repeated acquisitions, together with improved sensitivity to subtle gray matter atrophy. We openly release the model and the label templates at https://github.com/romainVala/SIAM.