Abstract:
Depression is a familiar psychological disorder caused by a combination of genetic, biolog- ical, environmental, and psychological factors. Untreated depression carries a high cost in terms of relationship problems, family suffering, and loss of work productivity. How- ever, diagnosis and treatment of depression is diffi ult due to varied severity, frequency, and duration of symptoms in depressed individuals. Psychologists use standard scales to detect depression but for that the depressed person needs to be present before the psychologist. Recent study reveals that, depression is reflected in behavioral fluctuation of certain day-to-day activities and physical parameters. It has also been studied that isolation from social activities increases risk of depression while social interaction and support helps greatly in fighting out the problem. In this thesis, a model of depression detection and support system has been designed using extensive user survey on diff rent symptoms and effects of depression that measures diff rent levels of depression based on individuals’ physical state, behavior, and social interaction. The detection system of the model will detect diff rent levels of depression by periodically collecting following data of users: a) heart rate interval through sensors in the smart-watch, b) sleeping pattern through monitoring activities during night, c) movement pattern through GPS location information, and d) communication pattern through monitoring phone calls, email, and social network usage. The support system will receive the level of depression from the detection system. If the depression level is mild, it will play a music track or show images of memorable events. If the depression episode is long, it will send messages to family and friends from a pre-selected list. If the depression level is severe, it will alert family, psychologist or psychological support organizations. This model has been verified by a psycho-social counsellor. Finally an intelligent wearable system has been developed which consists of an Android application installed in a smart-watch synchro- nized with the user’s smart-phone. In the developed module Heart Rate Variability data is collected by the sensor of the smart-watch and analyzed in user’s smart-phone. If the data indicates signs of depression, an SMS alert is sent to an emergency contact person to provide necessary support. Thus the system acts as both depression detection and support system.
