AUTHOR=Saner Hugo , Schütz Narayan , Botros Angela , Urwyler Prabitha , Buluschek Philipp , du Pasquier Guillaume , Nef Tobias 

TITLE=Potential of Ambient Sensor Systems for Early Detection of Health Problems in Older Adults

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 7 - 2020

YEAR=2020

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2020.00110

DOI=10.3389/fcvm.2020.00110

ISSN=2297-055X

ABSTRACT=Background: Home monitoring  sensor systems are increasingly used to monitor seniors in their apartments for detection of emergency situations. The aim of this study was to deliver a proof-of-concept for the use of multimodal sensor systems with pervasive computing technology for the detection of clinically relevant health problems over longer time periods. 
Methods: Data was collected with a longitudinal home monitoring study in Switzerland (StrongAge Cohort Study) in a cohort of 24 old and oldest-old, community dwelling adults over a period of one to two years. Physical activity in the apartment, toilet visits, refrigerator use and entrance door openings were quantified using a commercially available passive infrared (PIR) motion sensing system (Domosafety S.A., Switzerland). Heart rate, respiration rate, and sleep quality were recorded with the commercially available EMFIT QS bed sensor device (Emfit Ltd, Finland). Vital signs and contextual data were collected using a wearable sensor on the upper arm (Everion, Biovotion, Switzerland). Sensor data was correlated with health-related data collected from weekly visits of the seniors by health professionals including information about physical, psychological, cognitive and behavior status, health problems, diseases, medication and medical diagnoses.
Results: Twenty of the 24 recruited participants (age 88.9 ± 7.5 years, 79% females) completed the study, 2 participants had to stop study participation because of severe health deterioration, whereas 2 participants died during the course of the study.  A history of chronic disease was present in 12/24 seniors including heart failure, heart rhythm disturbances, pulmonary embolism, severe insulin-dependent diabetes and Parkinson disease. In total, 242’232 person-hours were recorded. During the monitoring period, 963 health status records were reported and repeated clinical assessments of ageing relevant indicators and outcomes were performed. Several episodes of health  deterioration including heart failure worsening and heart rhythm disturbances could be captured  by sensor signals from different sources.  
Conclusions: Our results indicate that monitoring of seniors with a multimodal sensor and pervasive computing system over longer time periods is feasible and well accepted with a great potential for detection of health deterioration. Further studies are necessary to evaluate the full range of the clinical potential of these findings.