Brain SPECT imaging shows promise in identifying schizophrenia

The research, conducted by the Tri-Institutional Center for Translational Research in Neuroimaging and Data Science and Amen Clinics, analyzed scans from 213 participants. By applying spatially constrained Independent Component Analysis to identify functional network templates, the team mapped how schizophrenia alters brain connectivity. When processed through machine learning models, logistic regression achieved 87% sensitivity, while random forest classifiers reached 88%.

These models pinpointed the middle occipital gyrus, subthalamus, and putamen as critical indicators of the disorder. Dr. Daniel Amen, founder of Amen Clinics, noted that the data reinforces the view of psychiatric conditions as observable brain disorders. The findings suggest that future diagnostic tools could move beyond symptom-based assessments, potentially allowing clinicians to monitor treatment responses and identify the specific circuits underlying hallucinations or cognitive impairment through objective neuroimaging.