Alessandra Retico1, Ilaria Gori1,2, Alessia Giuliano1,3, Piernicola Oliva1,2, Michela Tosetti4, Filippo Muratori3,4, and Sara Calderoni4
1National Institute of Nuclear Physics, Pisa, Italy, 2University of Sassari, Sassari, Italy, 3University of Pisa, Pisa, Italy, 4IRCCS Stella Maris Foundation, Pisa, Italy
Binary classifiers are
widely used to analyze brain MRI features and to identify useful biomarkers of
pathology. Strong challenges arise when dealing with extremely heterogeneous
conditions such as Autism Spectrum Disorders (ASD). We propose the use of the
One-Class Classifier (OCC) method that, in contrast to two-class
classification, is based on the description of the positive class only. A test
of similarity of new cases to the positive examples is then performed, and they
are eventually considered as outliers. The application of OCC to
Freesurfer-based brain MRI features identified a specific endophenotype in
young children with ASD.
