Meeting Banner
Abstract #3701

On Using Structural Network Patterns for Prediction of Genetic Risks in Schizophrenia

Madhura A. Ingalhalikar1, Luke Bloy1, Drew Parker1, Raquel Gur2, Ruben Gur2, Ragini Verma1

1Section of Biomedical Image Analysis, University of Pennsylvania, Philadelphia, PA, United States; 2Brain Behavior Laboratory, University of Pennsylvania, Philadelphia, PA, United States


This study investigates the presence of endophenotypic brain patterns in the family members of patients with schizophrenia via a structural network analysis. High dimensional gender specific classifiers based on local and global network properties were constructed for patients diagnosed with schizophrenia or schizoaffective disorder and healthy controls. The classifier associated a distributed network connectivity score (DNCS) with each of the asymptomatic family member. Forty percent of the FMs were classified closer to patients. Furthermore, females displayed enhanced genetic susceptibility based on the specificity of the classifier and the DNCS scores of the family members.