NIRS-aided differential diagnosis among patients with major depressive disorder, bipolar disorder, and schizophrenia

To establish a clinically applicable neuroimaging-guided diagnostic support system that uses near-infrared spectroscopy (NIRS) for differential diagnosis at the individual level among major depressive disorder (MDD), bipolar disorder (BPD), and schizophrenia (SZ). A total of 192 participants were recruited, including 40 patients with MDD, 38 patients with BPD, 65 patients with SZ, and 49 healthy individuals. We analyzed the spatiotemporal characteristics of hemodynamic responses in the frontotemporal cortex during a verbal fluency test (VFT) measured by NIRS to assess the accuracy of single-subject classification for differential diagnosis among the three psychiatric disorders. The optimal threshold of the frontal centroid value (54 seconds) was utilized on the basis of the findings of the Japanese study. The application of the optimal threshold of the frontal centroid value (54 seconds) allowed for the accurate differentiation of patients with unipolar MDD (72.5%) from BPD (78.9%) or SZ (84.6%). These results suggest that the NIRS-aided differential diagnosis of major psychiatric disorders can be a promising biomarker in Taiwan. Future multi-site studies are needed to validate our findings. •We used NIRS signals for differential diagnosis among major depressive disorder (MDD), bipolar disorder (BPD), and schizophrenia (SZ).•NIRS could accurately differentiate patients with MDD (72.5%) from BPD (78.9%) or SZ (84.6%).•Our results indicated that NIRS could be a promising biomarker in psychiatric clinical practice.