Summaries of Must-Read Clinical Literature, Guidelines, and FDA Actions
Clustering & Supervised Learning for Identifying Schizophrenia Subgroups
In a recent study that sought to identify schizophrenia subgroups using clustering and supervised learning, random forest performance metrics for predicting the group membership of the high and mild symptom burden groups exceeded those of the baseline comparison of the entire schizophrenia population vs normal controls. Symptomatic and MRI data of 167 subjects were used. Among the details:
- Subgroups were defined using hierarchical clustering of clinical data resulting in 2 stable clusters: high symptom burden, predominantly positive symptom burden, and mild symptom burden.
- Cortical thickness estimated were obtained in 78 regions of interest and were input into 3 machine learning models (logistic regression, support vector machine, and random forest) to predict subgroups.
- An analysis of the most important features in the random forest classification demonstrated consistencies with previous findings of regional impairments and symptoms of schizophrenia.
Citation:
Talpalaru A, et al. Identifying schizophrenia subgroups using clustering and supervised learning. [Published online ahead of print August 24, 2019]. Schizophrenia Res. doi: 10.1016/j.schres.2019.05.044.
Schizophrenia is a highly heterogeneous condition with wide variability across individuals in symptom presentation, response to treatment and overall outcomes. While there has been extensive work to characterize the phenomenology of schizophrenia over the past decades, evaluation and treatment decision-making is still largely done in-person and based on assessment by clinicians interacting with patients and families. This study used machine learning methods to perform sub-group and patient-level classification based on neuroimaging-derived neuroanatomical measures. The study methods and findings support future potential for technology and computational modeling to help advance and augment current standards for assessment and care planning for people living with schizophrenia. —Martha Sajatovic, MD, Professor of Psychiatry and of Neurology; Willard Brown Chair in Neurological Outcomes Research; Director, Neurological and Behavioral Outcomes Center, University Hospitals Cleveland Medical Center; Case Western Reserve University School of Medicine.