JMIR Publications

ABSTRACT

Background:

Motion tracking technology is the intermediary conveying activity data to healthcare professionals and the public. Acquiring data through motion tracking is a prerequisite for activity analysis and physical activity intervention. However, the motion tracking data investigation procedure has not yet been systematically reported in relation to users in activity recognition of healthcare application studies.

Objective:

Monitoring and analysing daily living and physical activities can provide vital information for disease prevention and healthy life promotion. This study aims at 1) systematically reviewing motion tracking of daily living and physical activities and activity identification in healthcare application research, 2) identifying the interaction between devices, participants, and environments by positioning designers' perspectives, 3) analysing the process of the health application research.

Methods:

A systematic review was conducted to investigate motion tracking data and the research process of daily living and physical activities in healthcare and physical wellness applications. The review was performed in four electronic databases: Scopus, Web of Science, EBSCO, and PubMed. By applying Actor Network Theory and Data-enabled Design, we analysed motion tracking data and the interaction of humans, devices, and environments in healthcare and physical wellness application research scenarios.

Results:

Fifty-five articles were included for full-text appraisal after the abstract screening process. The studies with accelerometer and gyroscope sensors comprised the core of the investigations (n=43; 78.2%), in which the studies using a portable device (n=11; 20%) and using multiple sensors that can achieve motion tracking (n=16; 29.1%). The sensors placements are implemented on full body (n=16; 29.1%) and in physical spaces (n=17; 30.9%). Researchers focused on lab-based activities in healthcare application study and aimed for professional use, such as activity identification and joint functionality.

Conclusions:

The findings revealed the motion data of participants are isolated from real-world scenarios. The application research for public health monitoring may not match the diversity of motion data in real-life settings, and the interaction between devices and participants is limited. Thus, we propose a designer-involved research process of motion data on daily living and physical activity for healthcare applications combining the DED framework.