Title : Integration of digital phenotyping with clinical (endo) phenotypes in children with neuro developmental disorders
Abstract: Neurodevelopmental disorders (NDD) such as attention deficit hyperactivity disorder (ADHD), or social communication disorder, although most commonly studied in childhood, have complex clinical (endo)phenotypes with an important social context and can be lifelong conditions. Furthermore, strong overlaps across NDD phenotypes make both group and distinction of each disorder difficult in real-life practice.
Digital phenotyping in natural environment represents a new approach aimed at measuring human behavior and may, combined with clinical (endo)phenotypes, enhance capability and sensitivity in early identification, diagnosis and management of mental health conditions. Moreover, such a combined approach may easily allow clinicians to perform a more personalized and patient-tailored diagnostic and therapeutic approach. Here, we investigate how digital phenotyping integrated with clinical (endo)phenotypes constitutes a new method for identification and pre-diagnosis of NDD in children aged 7 to 12 years.
A total of 100 children aged 7 to 12 years are evaluated through a new digital neuro-battery designed to assess cognitive and psychopathological domains (digital tests and questionnaires) as well as neurobehavioral functioning (eye- & digit-tracking and EEG). We are examining a large set of features within these multi-modal data using machine learning algorithms to identify prominent predictive variables to achieve dimensionality reduction and provide finer and more sensitive classification of NDD. Our preliminary results are encouraging for ongoing development of our digital screening tools that can be used in ecological environments on touch-screen tablets with cost-effective behavioral trackers. Larger paediatric populations can then be reached to improve early detection, prevention and ultimately intervention.