Social Information pRocessing in Children: an ocUlo-pupillometric tool for Standard evaluation – SIRCUS

Atypical exploration (measured by oculometry) and physiological reactivity (measured by pupillometry) have been reported in ASD and could constitute early markers in children at risk. These oculo-pupillometric parameters can be easily measured in children using eye-tracking methods. These objective parameters could help in the diagnosis and complement the subjective approach of the NDD questionnaires. Indeed, the pupillary response to faces has been shown to be discriminating (in 80% of cases children are correctly classified in the «Controls« or «ASD« group). This discriminating power can be improved by artificial intelligence which would identify and combine the most discriminating parameters to achieve a performance greater than 90%.

Three partners including UMR U1253 (iBrain) (specialized in eye tracking measurements in children with ASD), the Applied Computer Science Laboratory of the University of Tours (LIFAT) (specialized in “machine learning” methods) and the child psychiatry center of the CHRU of Tours (specializing in ASD and TND clinics) will join forces to create a diagnostic aid tool. This project therefore aims to develop a nomadic and efficient tool (based on oculo-pupillometric responses measured by eye tracking) that can be used in clinical routine to help early screening of ASD or differential diagnosis with other TNDs.

Autism spectrum disorder (ASD) as well as other neurodevelopmental disorders (NDD) are a public health priority in France. To help early diagnosis and improve therapeutic care for children, the French government has just set up an early detection program with the creation of "NDD platforms". Thanks to a questionnaire (NDD questionnaire), children suspected of NDD can be directed (from 6 months) to these platforms by non-specialist doctors or pediatricians to be diagnosed and included in a therapeutic program. Among the early indicators examined in these questionnaires, socialization disorders are assessed. In fact, ASD is characterized in particular by social difficulties linked to a disruption in the processing of social stimuli such as faces. Atypical exploration (measured by oculometry) and physiological reactivity (measured by pupillometry) have been reported in ASD and could constitute early markers in children at risk. These oculo-pupillometric parameters can be easily measured in children using eye-tracking methods. These objective parameters could help in the diagnosis and complement the subjective approach of the NDD questionnaires. Indeed, the pupillary response to faces has been shown to be discriminating (in 80% of cases children are correctly classified in the "Controls" or "ASD" group). This discriminating power can be improved by artificial intelligence which would identify and combine the most discriminating parameters to achieve a performance greater than 90%. Three partners including UMR U1253 (iBrain) (specialized in eye tracking measurements in children with ASD), the Applied Computer Science Laboratory of the University of Tours (LIFAT) (specialized in “machine learning” methods) and the child psychiatry center of the CHRU of Tours (specializing in ASD and TND clinics) will join forces to create a diagnostic aid tool. This project therefore aims to develop a nomadic and efficient tool (based on oculo-pupillometric responses measured by eye tracking) that can be used in clinical routine to help early screening of ASD or differential diagnosis with other NDD.