Multimodal model for analysis and support to the diagnosis of Tourette Syndrome in pediatric population: use of convolutional neural networks in functional and structural magnetic resonance imaging

• Grant number: 25/00214-8
• Support Opportunities: Regular Research Grants
• Start date: November 01, 2025
• End date: October 31, 2027
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Marco Antonio Garcia de Carvalho
• Grantee: Marco Antonio Garcia de Carvalho
• Host Institution: Faculdade de Tecnologia (FT). Universidade Estadual de Campinas (UNICAMP). Limeira , SP, Brazil
• City of the host institution: Limeira
• Associated researchers: Kauê Tartarotti Nepomuceno Duarte
• Associated scholarship(s): 25/25237-0 - Multimodal Model for Analysis and Support to the Diagnosis of Tourette Syndrome in Pediatric Population: Use of Convolutional Neural Networks in Functional and Structural Magnetic Resonance Imaging, BP.DR

Abstract

Tourette Syndrome (TS) is a neuropsychiatric disorder, prevalent in approximately 1% of the population, characterized by motor and vocal tics that appear in childhood and can be confused with other conditions such as obsessive-compulsive disorder (OCD) and attention-deficit/hyperactivity disorder (ADHD). This project aims to enhance the diagnosis of TS using advanced imaging processing and machine learning techniques, with a focus on structural magnetic resonance imaging (MRI) and functional MRI (fMRI). The research is structured into three modules; (i) Structural Pre-processing: This module involves image normalization, motion correction, and noise reduction, facing additional challenges due to pediatric data; (ii) Temporal Analysis via Machine Learning: This module employs techniques such as Long Short-Term Memory (LSTM) networks, Statistical Parametric Mapping (SPM), and Time Series Modeling (SLM) to identify dynamic patterns and brain alterations associated with TS; (iii) Data Integration and Classification: This module uses computer vision and artificial intelligence algorithms, such as Convolutional Neural Networks (CNNs), to classify and identify relevant features in brain images. The project is conducted in collaboration with National Taiwan University (NTU), which provided MRI data from 68 participants, divided between TS and healthy control groups. The application of these advanced techniques promises to overcome the limitations of traditional methods, offering a significant advancement in the accuracy and effectiveness of Tourette Syndrome diagnosis. (AU)