Automated Health Assessment through Mobile Sensing and Machine Learning of Daily Activities

PROJECT SUMMARY ABSTRACTAdvances in health care have been dramatic since the beginning of the millenniumAs a resultpeople are living longer with age related diseasesand the number of older individuals unable to live independently is rising rapidlyMobile computing offers an unprecedented opportunity to unobtrusively monitor behavior and use the large amount of collected data to automate analysis of functional health in a personandapos s everyday settingsThis project focuses on evaluating the performance and commercial viability of technologies that will meet some of the needs that this coming age wave introduces by automating assessment of a personandapos s functional performanceThe primary objective of this Phase I STTR application is to evaluate the feasibility of assessing an individualandapos s cognitive and mobility based health using behavior patterns as sensed by a smart watchAchieving this objective will provide a foundation for the Phase II application goal of using multiple information sources to automatically generate activity scores and functional health measures from sensor dataBuilding on our prior collaborative workour approach creates a profile of a personandapos s routine behavior through automated real time recognition of complex activities from mobile sensor dataAimWe will evaluate the use of machine learning techniques to map behavior features onto cognitive and mobility health scores provided through app and in person neuropsychological assessmentAimFinallywe will evaluate an interactive visual tool for displaying behavior patterns to provide individuals and their caregivers with insights on their routines and relationship with their health statusAimThis work has important health care implications as functional impairment has been associated with negative outcomes including increased health care utilizationfallsand conversion to dementiaGiven nursing home care coststhe impact of family based careand the importance that people place on staying at homeit is imperative to commercialize technologies that increase functional independence while improving quality of life for both individuals and their caregivers PROJECT NARRATIVE We propose to evaluate the performance and commercial viability of an application that can predict an individualandapos s cognitive and mobility based health based on patterns that are sensed by a smart watchand effectively and efficiently present this information to the caregiverThis work will lay the foundation for a commercial tool to analyze behavior routines and automate assessment of functional healthThis research is relevant to public health because these technologies can extend the functional independence of our aging society through technology assisted health self managementreduce caregiver burdenand improve quality of life