Diagnosis and characterization of neurodevelopmental disorders are considered challenging processes because of their complexity, multi-factoriality and heterogeneity. The present project will consider two of the most common neurodevelopmental disorders (i.e. autism spectrum disorders (ASD) and language disorders (LD)), with the aim to overcome these difficulties, by: a) deeply investigating their neuronal correlates; b) identifying multi-domain biomarkers (electrophysiological, genetic, environmental and clinical); c) developing a machine learning algorithm for early diagnosis. To achieve the above mentioned aims a multi-domain dataset will be used, comprising data collected from typically developing infants, infants at high risk for ASD and infants at high risk for LD. The data that will be used have been already collected within other trials performed at the Scientific Institute E. Medea.

Inclusion Criteria:

• APGAR scores at birth at 1' and 5' > 7;
• newborn screening otoacoustic emissions in the norm;
• absence of neurological disorders;
• Italian as the language mainly spoken in the family nucleus.

Exclusion Criteria:

• APGAR scores at birth at 1' and 5' <= 7;
• newborn screening otoacoustic emissions not in the norm;
• presence of neurological disorders;
• non-Italian native speaker