Exploring Gaze Pattern Differences Between Autistic and Neurotypical Children: Clustering, Visualisation, and Prediction

TL;DR

This paper investigates whether internal cluster validity indices derived from gaze-point clustering can distinguish ASD from TD children. It applies seven clustering methods across three eye-tracking datasets, then extracts nine indices to create 63 predictive features, evaluated with multiple classifiers. The approach achieves an average AUC of $0.81$ for ASD prediction and shows $0.741$ of indices are significant, supporting the utility of these indices for gaze analysis and potential diagnosis. The findings offer a scalable, automated avenue to analyze gaze patterns in ASD and aid clinical assessment.

Abstract

Autism Spectrum Disorder (ASD) affects children's social and communication abilities, with eye-tracking widely used to identify atypical gaze patterns. While unsupervised clustering can automate the creation of areas of interest for gaze feature extraction, the use of internal cluster validity indices, like Silhouette Coefficient, to distinguish gaze pattern differences between ASD and typically developing (TD) children remains underexplored. We explore whether internal cluster validity indices can distinguish ASD from TD children. Specifically, we apply seven clustering algorithms to gaze points and extract 63 internal cluster validity indices to reveal correlations with ASD diagnosis. Using these indices, we train predictive models for ASD diagnosis. Experiments on three datasets demonstrate high predictive accuracy (81\% AUC), validating the effectiveness of these indices.