Classification subject effects using changes in cerebral blood flow on the Stroop test

Functional near-infrared spectroscopy (fNIRS) is often used as a diagnostic method for mental illness, because the average patterns of changes in cerebral blood flow vary between such patients and healthy subjects. In addition, indoor environments and psychology alter the average patterns of changes in cerebral blood flow. These observations suggest that it may be possible to classify healthy subjects into different groups according to certain characteristics. The accuracy of fNIRS should be improved once it becomes possible to automatically determine the key factors affecting fNIRS data. The present study was performed first to determine whether there are differences in fNIRS data when the test subjects are classified into groups based on the scores related to task performances and questionnaires, and to determine whether there are differences in score when the test subjects are classified into groups based on fNIRS data. Differences were observed in fNIRS data between groups when the subjects were classified based on incorrect answers on the questionnaire and their degree of fatigue. In addition, there were differences in score between groups when the subjects were classified according to fNIRS data. These results suggested that subjects can be classified into groups automatically based on scores related to both task performance and fNIRS data.