Title: Signatures from neuroimaging studies predict transition to psychosis
Abstract
Currently in the field of schizophrenia research the early detection of at risk mental states for psychosis is an important topic of clinical research. Beside clinical and neuropsychological parameters structual MRI related parameters in focus.
Using support vector machine learning and pattern detection analysis, we demonstrate that the different kinds of at risk states (early at risk state, versus late at risk state) are associated with different kinds of brain alterations. It is possible to predict the risk of transition to psychosis on an individual level with high sensivity and specivity. This pattern recognition analyses proved also to be useful to differentiate between schizophrenia and depression (MDD) using brain aging as intermediate parameter.
Corresponding author:
Prof. Dr. Hans-Jürgen Möller, Department of Psychiatry
Ludwig-Maximilians-University München
Nussbaumstrasse 7, 80336 Munich, Germany
Tel: +49 89 5160 5514, E-mail: hans-juergen.moeller@med.uni-muenchen.de