Transformer-based approach for analyzing movement in children on the Autism Spectrum

• Grant number: 25/02986-8
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2025
• End date: May 31, 2026
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: André Carlos Ponce de Leon Ferreira de Carvalho
• Grantee: Enzo Tonon Morente
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Company: Universidade de São Paulo (USP). Instituto de Ciências Matemáticas e de Computação (ICMC)
• Associated research grant: 20/09835-1 - IARA - Artificial Intelligence in the Remaking of Urban Environments, AP.PCPE

Abstract

This project aims to apply transformer-based architectures to analyze the movements of children withAutism Spectrum Disorder (ASD). The goal is to contribute to the existing literature by testing thesemodels and, with the aid of Explainable AI (XAI) algorithms, understanding the patterns learned bythe models and identifying potential new behavioral biomarkers.Artificial Intelligence (AI) has been increasingly integrated into medicine due to its ability toprocess large volumes of data and identify complex patterns within them. Autism Spectrum Disor-der (ASD) is a neurodevelopmental disorder primarily characterized by deficits in social interaction,atypical movement patterns, and restrictive and repetitive behaviors.Currently, ASD diagnosis is not based on well-defined biomarkers, making the assessment reli-ant on behavioral analyses and, therefore, susceptible to subjectivity. In response to this limitation,machine learning models have been implemented to assist in ASD diagnosis and identify potentialbehavioral biomarkers that might go unnoticed using traditional analysis methods.