Automated ROI-based brain parcellation analysis of frontal and temporal brain volumes in schizophrenia☆

Abstract 

Structural MRI studies of schizophrenia have yielded a diversity of findings. To help characterize regional gray matter changes in schizophrenia, we used an automated region of interest (ROI)-based approach that targeted frontal and temporal regions in schizophrenia patients. The sample compromised 43 schizophrenia patients (21 chronic patients, 22 unmedicated first episode patients), 20 first episode non-schizophrenia psychosis patients and 47 comparison subjects. Automated regional volume measurement was performed in 22 ROIs, including frontal and temporal cortical subregions and hippocampus. Correlations between volume measures, duration of illness and clinical scores were evaluated. Chronic schizophrenia patients showed gray matter volume differences in left dorsolateral prefrontal cortex (DLPFC) and right supplementary motor area (SMA). First episode psychosis patients presented smaller right anterior cingulate cortex (ACC) and left DLPFC than comparison subjects. Disorganization scores and duration of illness correlated negatively with gray matter volume of DLPFC and SMA in chronic schizophrenia patients. Using an automated ROI-based method, we found volume reductions in lateral and medial frontal regions in both first episode and chronic schizophrenia. The automated ROI-based method can be used as a valid and efficient tool for quantification of regional gray matter volume in schizophrenia in multiple ROIs across the brains of large numbers of subjects.

Keywords: Schizophrenia, Magnetic resonance imaging, Frontal cortex, Brain volume, Structural imaging, First episode schizophrenia