[A discriminant model of mental disorders based on wave form patterns obtained by multi-channel near-infrared spectroscopy].

: Multi-channel near-infrared spectroscopy (NIRS) was approved on 2009 as the first advanced medical care modality for use in the field of psychiatry in Japan. We performed NIRS for 185 outpatients in our hospital and 59 healthy subjects to measure hemoglobin concentration changes during verbal fluency tests trying to evaluate the relationships between the wave forms obtained by NIRS and mental disorders. We classified the prefrontal cortex oxy-hemoglobin wave forms obtained from the NIRS into 2 types and sub-classified into 5 wave patterns partly referenced previous papers a) Flat or increasing oxy-Hb form: (1) flat wave pattern, (2) early peak wave pattern, (3) late peak wave pattern during of the task and (4) reascending wave pattern after the task and b) decreasing oxy-Hb form: (5) Decrease wave pattern during the task. Focused on flat or increasing oxy-Hb form, the associations between these 4 wave patterns and psychiatric disorders were confirmed employing the Chi-square test. It was found that the flat wave pattern during the task and depression correlated with a sensitivity of 51.5% and specificity of 90.2%, the late peak wave pattern and bipolar disorder correlated with a sensitivity of 65.9% and specificity of 73.2%, and the re-ascending wave pattern after the task and schizophrenia correlated with a sensitivity of 58.9% and specificity of 94.6%. Our findings suggest that the discriminant model based on wave pattern has the potential to provide information supporting a diagnosis of mental disorder in the setting of clinical laboratory testing.