Usage of emotion recognition in military health care

Abstract

Post Traumatic Stress Disorder (PTSD), depression and suicide are major psychiatric problem in both military and civilian situation. These mental diseases combine with emotion change. Recently, the technology of emotion recognition has been developed rapidly and highly. Therefore, we investigate if the emotion recognition by natural speaking voice could detect the emotion change which would occur when exposing mental stress. We used "Sensibility technology ST Emotion" (AGI Japan Inc.) for emotion voice analysis system. This system determines emotional elements as including anger, joy, sorrow, and calmness. It also measures feeling of excitement. Voice data were collected from the personnel of military medical corps participating in a special stressful mission. The voice data were divided into two groups depending on participating period. Some subject's feelings during experimental period were changed. There is a tendency that "joy" of long stay group (Group L) is lower than short stay group (Group S) and "sorrow" of Group L is higher than Group S. The result suggested that the techniques of emotion recognition may be used for screening of mental status in military situation. However, further development is necessary for practical use.