A Web-Based Non-Intrusive Ambient System to Measure and Classify Activities of Daily Living

摘要: Background: The number of older adults in the global population is increasing. This demographic shift leads to an increasing prevalence age-associated disorders, such as Alzheimer’s disease and other types dementia. With progression disease, risk for institutional care increases, which contrasts with desire most patients stay their home environment. Despite doctors’ caregivers’ awareness patient’s cognitive status, they are often uncertain about its consequences on activities daily living (ADL). To provide effective care, need know how cope ADL, particular, estimation risks associated decline. occurrence, performance, duration different ADL important indicators functional ability. ability these traditionally assessed questionnaires, has disadvantages (eg, lack reliability sensitivity). Several groups have proposed sensor-based systems recognize quantify home. Combined Web technology, can inform caregivers real-time via smartphone). Objective: We hypothesize that a non-intrusive system, does not use body-mounted sensors, video-based imaging, microphone recordings would be better suited dementia patients. Since it require attention compliance, system might well accepted by present passive, Web-based, non-intrusive, assistive technology recognizes classifies ADL. Methods: components this novel were wireless sensors distributed every room participant’s central computer unit (CCU). environmental data acquired 20 days (per participant) then stored processed CCU. In consultation medical experts, eight classified. Results: study, 10 healthy participants (6 women, 4 men; mean age 48.8 years; SD 20.0 range 28-79 years) included. For explorative purposes, one female Alzheimer patient (Montreal Cognitive Assessment score=23, Timed Up Go=19.8 seconds, Trail Making Test A=84.3 B=146 seconds) was measured parallel subjects. total, 1317 performed participants, 1211 classified correctly, 106 missed. led overall sensitivity 91.27% specificity 92.52%. Each subject average 134.8 (SD 75). Conclusions: sensor acquire essential classification living. By analyzing retrieved data, possible distinguish assign patterns subjects' specific identify Web-based allows improve provides valuable information real-time. [J Med Internet Res 2014;16(7):e175]