Spectral Graph Transformer Networks for Brain Surface Parcellation

TL;DR

This paper introduces a Spectral Graph Transformer network that efficiently aligns brain surface graphs in the spectral domain, enabling faster and more accurate cortical parcellation compared to traditional methods.

Contribution

The paper proposes a novel data-driven spectral alignment method using a graph transformer that significantly speeds up brain surface analysis and improves parcellation accuracy.

Findings

Achieved 1400-fold speed improvement over iterative alignment methods.

Improved cortical parcellation performance on 101 brain surfaces.

Validated the effectiveness of the spectral transformer in brain graph analysis.

Abstract

The analysis of the brain surface modeled as a graph mesh is a challenging task. Conventional deep learning approaches often rely on data lying in the Euclidean space. As an extension to irregular graphs, convolution operations are defined in the Fourier or spectral domain. This spectral domain is obtained by decomposing the graph Laplacian, which captures relevant shape information. However, the spectral decomposition across different brain graphs causes inconsistencies between the eigenvectors of individual spectral domains, causing the graph learning algorithm to fail. Current spectral graph convolution methods handle this variance by separately aligning the eigenvectors to a reference brain in a slow iterative step. This paper presents a novel approach for learning the transformation matrix required for aligning brain meshes using a direct data-driven approach. Our alignment and graph processing method provides a fast analysis of brain surfaces. The novel Spectral Graph Transformer (SGT) network proposed in this paper uses very few randomly sub-sampled nodes in the spectral domain to learn the alignment matrix for multiple brain surfaces. We validate the use of this SGT network along with a graph convolution network to perform cortical parcellation. Our method on 101 manually-labeled brain surfaces shows improved parcellation performance over a no-alignment strategy, gaining a significant speed (1400 fold) over traditional iterative alignment approaches.