Simultaneous Matrix Diagonalization for Structural Brain Networks Classification

TL;DR

This paper introduces a novel method using simultaneous approximate diagonalization of brain network matrices to improve the stability and accuracy of classifying brain diseases like Alzheimer's.

Contribution

It proposes a new approach for brain disease classification that leverages simultaneous matrix diagonalization to extract stable eigenfeatures from connectome data.

Findings

Outperforms baseline methods in Alzheimer's detection

Provides more stable eigenstructure features

Achieves competitive accuracy with state-of-the-art techniques

Abstract

This paper considers the problem of brain disease classification based on connectome data. A connectome is a network representation of a human brain. The typical connectome classification problem is very challenging because of the small sample size and high dimensionality of the data. We propose to use simultaneous approximate diagonalization of adjacency matrices in order to compute their eigenstructures in more stable way. The obtained approximate eigenvalues are further used as features for classification. The proposed approach is demonstrated to be efficient for detection of Alzheimer's disease, outperforming simple baselines and competing with state-of-the-art approaches to brain disease classification.