Machine learning improved classification of psychoses using clinical and biological stratification: Update from the bipolar-schizophrenia network for intermediate phenotypes (B-SNIP)

Psychiatry continues to suffer from challenges to diagnostic validity due to lack of biological markers. Distinctions between diagnostic categories are still largely informed on symptom clusters, while underlying biology is complex and largely indiscernible. (Insel et al., 2010; Keshavan et al., 2013) While the reliability of diagnoses has improved with successive revisions of the classificatory systems such as the latest Diagnostic and Statistical Manual of Mental Disorders (DSM-V), whether these disorders have natural biological distinctions is uncertain.

Source: Schizophrenia Research - May 25, 2018 Category: Psychiatry Authors: Suraj Sarvode Mothi, Mukund Sudarshan, Neeraj Tandon, Carol Tamminga, Godfrey Pearlson, John Sweeney, Brett Clementz, Matcheri S. Keshavan Source Type: research