Computer Science Colloquia
Tuesday, August 16, 2011
Robert F. Dickerson
Advisor: Jack Stankovic
Attending Faculty: Alfred Weaver, Chair; John Lach, Minor representative; Kamin Whitehouse; Gabe Robins; and Steve Patek
Olsson Hall, Room 236D, 2:00 PM
Ph.D. Proposal Presentation
System for the Collection, Storage, Analysis, and Reporting of Objective and Subjective Behavioral Measures
Depression is a major health issue that affects over 21 million American men and women each year. Depression often goes unrecognized and untreated and even once treatment begins it is often difficult to monitor its effectiveness. This poses particular challenges for the diagnosis and treatment of depression, particularly for those who avoid visiting a doctor or therapist due to social stigmas or a lack of energy. Currently, depression diagnosis is often based on subjective screening questionnaires or structured clinical interviews that rely on timely in-person visits as well as accurate recollections by the patient. This makes early detection of depression symptoms exceedingly difficult. Yet early detection and treatment of this debilitating disorder has been shown to improve patient outcomes considerably. Assessment and treatment are often hampered by a lack objective data to corroborate patients’ retroactive self-reports about their current functioning; hence an objective symptom-monitoring tool could complement subject self-report measurement and enhance diagnostic accuracy.
We propose the creation of a continuous depression monitoring system for in-home sensing. Before such a system can be realized, we have identified technical challenges that first need to be considered and solved. These include the 1) handling multi-modal inference and 2) evaluation of new caregiver monitoring user interfaces, 3) advancements in emplaced speech emotion monitoring, 4) noninvasive and cheap sleep monitoring, and 5) scaling the system to large number of deployments over long periods of time.