Scientists of the University of Coimbra use artificial intelligence to study autism

English version: Diana Taborda

The Institute of Systems and Robotics (ISR) of the Department of Electrical and Computer Engineering (DEEC) - Faculty of Sciences and Technology of the University of Coimbra (FCTUC) - is conducting a project that uses artificial intelligence to study and characterise Autism Spectrum Disorder (ASD).

The project “Move4ASD - Multimodal Motion Analysis Using Machine Learning-based Techniques for Autism Spectrum Disorder Characterization” is being carried out in partnership with the Coimbra Institute for Biomedical Imaging and Translational Research (CIBIT) of the UC Institute for Nuclear Sciences Applied to Health (ICNAS) and the University of Oxford, UK. Its main goal is to “use technological approaches based on artificial intelligence methodologies to characterise autism through the analysis of imitation actions.”, explains João Ruivo Paulo, researcher at ISR and project leader at the UC.

Ruivo Paulo adds that the project “studies the hypothesis that the mirror neuron system is altered in individuals diagnosed with autism, when compared to healthy individuals. Thus, we believe that learning motor activities, such as walking and dancing, and other social interaction activities, may be misinterpreted by individuals diagnosed with this disorder”

According to João Ruivo Paulo, Move4ASD will analyse neurophysiological and behavioural data, collected through electroencephalography (EEG) and three-dimensional movement. “This data will be processed by machine learning techniques to detect differences between healthy individuals and the clinical group participants. Through this group differentiation it will be possible to identify neurophysiological biomarkers of motor patterns that are still poorly understood.”

The team believes that this research “will contribute to a better understanding of the mirror neuron system in autism, thus leading to the development of an automatic tool to identify these biomarkers, which in turn may result in an optimised characterisation of these disorders”.

Besides João Ruivo Paulo, the team includes Professors Paulo Menezes and Gabriel Pires - DEEC/ISR, Miguel Castelo Branco and Teresa Sousa - CIBIT/ICNAS, and Tingting Zhu, researcher at the University of Oxford. The project is funded by the Portuguese Foundation for Science and Technology (FCT).