The Effect of Sensor Placement and Number on Physical Activity Recognition and Energy Expenditure Estimation in Older Adults: Validation Study

Abstract

Background Research has shown the feasibility of human activity recognition using wearable accelerometer devices. Different studies have used varying numbers and placements for data collection using sensors. Objective This study aims to compare accuracy performance between multiple and variable placements of accelerometer devices in categorizing the type of physical activity and corresponding energy expenditure in older adults. Methods In total, 93 participants (mean age 72.2 years, SD 7.1) completed a total of 32 activities of daily life in a laboratory setting. Activities were classified as sedentary versus nonsedentary, locomotion versus nonlocomotion, and lifestyle versus nonlifestyle activities (eg, leisure walk vs computer work). A portable metabolic unit was worn during each activity to measure metabolic equivalents (METs). Accelerometers were placed on 5 different body positions: wrist, hip, ankle, upper arm, and thigh. Accelerometer data from each body position and combinations of positions were used to develop random forest models to assess activity category recognition accuracy and MET estimation. Results Model performance for both MET estimation and activity category recognition were strengthened with the use of additional accelerometer devices. However, a single accelerometer on the ankle, upper arm, hip, thigh, or wrist had only a 0.03-0.09 MET increase in prediction error compared with wearing all 5 devices. Balanced accuracy showed similar trends with slight decreases in balanced accuracy for the detection of locomotion (balanced accuracy decrease range 0-0.01), sedentary (balanced accuracy decrease range 0.05-0.13), and lifestyle activities (balanced accuracy decrease range 0.04-0.08) compared with all 5 placements. The accuracy of recognizing activity categories increased with additional placements (accuracy decrease range 0.15-0.29). Notably, the hip was the best single body position for MET estimation and activity category recognition. Conclusions Additional accelerometer devices slightly enhance activity recognition accuracy and MET estimation in older adults. However, given the extra burden of wearing additional devices, single accelerometers with appropriate placement appear to be sufficient for estimating energy expenditure and activity category recognition in older adults.

A Sensor-Based mHealth Platform for Remote Monitoring and Intervention of Frailty Patients at Home

“…For the evaluation of physical activity, the use of inertial sensors based on accelerometry is common [ 40 ]. These devices can provide, in a simple way, numerous metrics related to physical activity such as the number of steps while walking, the time and intensity of physical activity, study of sedentary lifestyle and the estimation of metabolic expenditure [ 41 ]. The main disadvantages of these systems are the lack of transparency in the algorithms used by commercial devices, which in many cases are treated as black boxes, and the lack of a gold standard for carrying out activities for the evaluation.…”

Assessing inertial measurement unit locations for freezing of gait detection and patient preference

“…The best technical set was defined as the IMU set that achieved the highest AUROC. The minimal IMU set was defined as the set with the fewest number of IMUs that still achieved an average AUROC within 5% of the best technical set (22). To account for IMU wearability and assess deployment feasibility, results from the best technical and minimal IMU sets were compared to results from the survey-preferred sets.…”

“…In addition, few studies investigate wearability and patient preferences, which may limit successful deployment. Recently, Davoudi et al conducted a systematic analysis across accelerometer quantities and locations for physical activity recognition in older adults (22). This analysis informed the research community about sensor use in a general population across a variety of activities.…”

Is Complexity of Daily Activity Associated with Physical Function and Life-Space Mobility among Older Adults?

“…Recognizing activities based on body-worn sensors is not a trivial task, and age-related decrease in vigor of movements ( 15 , 16 ) can further increase the difficulty of activity classification ( 15 ). Furthermore, the number and placement of wearable sensors affects the precision of recognizing activity classes ( 17 ). These challenges have been tackled with concurrent trunk and thigh-worn accelerometers, which enables robust classification of postures throughout the day ( 13 , 18 ), while still retaining a reasonable participant burden ( 13 , 17 ).…”

“…Furthermore, the number and placement of wearable sensors affects the precision of recognizing activity classes ( 17 ). These challenges have been tackled with concurrent trunk and thigh-worn accelerometers, which enables robust classification of postures throughout the day ( 13 , 18 ), while still retaining a reasonable participant burden ( 13 , 17 ). Posture assessments may be particularly informative in daily activity behavior containing a large volume of stationary behavior where the distinction between sitting and standing may be of interest ( 19 ).…”