Abstract #0513

The Cerebellum and Brainstem Together Increase Classification Accuracy for Autism Spectrum Disorder over the Whole Brain

Muriel M. K. Bruchhage¹, Leon M. Aksman¹, Andre F. Marquand², MRC AIMS Consortium³, EU TACTICS Consortium⁴, Jan Buitelaar², Declan Murphy⁵, and Steven C. R. Williams¹

¹Centre for Neuroimaging Sciences, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, United Kingdom, ²Department of Cognitive Neuroscience, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Nijmegen, Netherlands, ³Centre for Neuroimaging Sciences, Institute of Psychiatry, Psychology and Neuroscience, King's College London; Cambridge University; Oxford University, London; Cambridge; Oxford, United Kingdom, ⁴Radboud University Medical Center; King's College London; University Medical Center Utrecht; Central Institute of Mental Health Mannheim, Nijmegen; London; Utrecht; Mannheim, Netherlands, ⁵Sackler Institute of Translational Neuroimaging, Department of Forensics and Neurodevelopmental Sciences, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, United Kingdom

Autism spectrum disorder (ASD) has been linked to cerebellar and brainstem dysfunction and abnormal development, but it remains unclear whether these regional abnormalities can help classify the disorder. Performing machine learning based classification using Jacobian determinant based features on two independent male ASD cohorts (adult and paediatric) of different sizes and age range, we demonstrated a consistently higher classification accuracy by up to 15% using the cerebellum and brainstem as regions of interest classifiers over the whole brain. In both cohorts, classification was driven by regional differences in the posterior lateral cerebellum.

This abstract and the presentation materials are available to members only; a login is required.

Join Here