Improving brain disorder diagnosis with advanced brain function representation and Kolmogorov-Arnold Networks

Tyler Ward, Abdullah Al Zubaer Imran

Keywords: Brain disorder diagnosis, brain function representation, classification, deep learning, Kolmogorov-Arnold Network, transformer

TL;DR: This work explores the efficacy of Kolmogorov-Arnold Networks (KANs) in replacing multi-layer perceptrons (MLPs) for brain disorder diagnosis based on thorough experimentation with a state-of-the-art transformer-based classification model.

Abstract:

Quantifying functional connectivity (FC), a vital metric for the diagnosis of various brain disorders traditionally relies on the use of a pre-defined brain atlas. However, using such atlases can lead to issues regarding selection bias and lack of regard for specificity. Ad-
dressing this, we propose a novel transformer-based classification network (AFBR-KAN) with effective brain function representation, to aid in diagnosing autism spectrum disorder (ASD). AFBR-KAN leverages Kolmogorov-Arnold Network (KAN) blocks replacing traditional multi-layer perceptron (MLP) components. Thorough experimentation reveals the effectiveness of AFBR-KAN in improving the diagnosis of ASD under various configurations of the model architecture.

Primary Subject Area: Detection and Diagnosis

Secondary Subject Area: Integration of Imaging and Clinical Data

Paper Type: Both

Registration Requirement: Yes

Submission Number: 115