When a patient starts working with a new human caregiver it takes time and talking to develop trust and acceptance, and we should expect the same for electronic cognitive assistants. Human caregivers also function independently from the patient. While existing electronic cognitive aids simply echo the user's schedule back to them, human caregivers think independently about the best way to cue the user in a given situation. We propose to develop an Independent Living Assistant that interacts with patients in a similar way, by using a human-like face and a "conversational" ability to discuss preferences and options with the user. We will extend a commercial cognitive aid already being used by the VA, by adding physiological sensors to detect panic attacks and other psychological symptoms. The system may detect when users avoid activities, and converse with users during in-vivo exposure sessions to collect physiological data as well as their reported in-vivo experience, in order to determine if and when the patient is ready to move to the next level. Anticipated benefits of the system include increased user independence and reduced long-term caregiver costs. Our Medical Advisory Board includes two VA neuropsychologists who will provide guidance about clinical and commercial considerations.
Keywords: Cognitive Interface, Independent Living Assistant, Ptsd, Tbi, Sensors, Planning, Cueing, Conversation