Features Derived from Behavioral Experiments to Distinguish Mental Healthy People from Depressed People

How to distinguish mental healthy people from people suffered from depression objectively is an unsolved question in both pattern recognition and clinical psychology. Traditional diagnosis of depression was drawn by doctors according to the results of questionnaire analysis or scale tests, which were prone to be subjective. In this paper, we selected negative and positive scenes from IAPS as the background, positive and negative faces from Taiwanese Facial Expression Image Database as foreground. Based on the competing effect and the priming effect of background scenes, we designed two experiments by controlling the relative appearing orders of background scenes and foreground faces. By recording subjects’ response times at key-strokes, we quantified the identification differences towards different combinations of background and foreground between depressed people and mental healthy people, and these quantified differences can be used as criteria to derive features distinguishing two kinds of people objectively.