Placement-Dependent Accuracy of a Smartphone-Based Sensor Application Compared to an Accelerometer-Based System for...

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Digital Health

Background:

• Accurate physical activity monitoring is crucial for chronic disease management and post-surgery rehabilitation.
• Stationary cycling and other activities require reliable tracking methods.
• Smartphone-based sensors offer a potential low-cost alternative to dedicated activity trackers.

Purpose of the Study:

• To validate the BeSAFE+ smartphone application for activity recognition and step counting.
• To assess the accuracy of the BeSAFE+ app across five different phone placements.
• To compare the BeSAFE+ app's performance against the SENS Motion® system and observed activity time.

Main Methods:

• A laboratory-based study involving 20 participants.
• Participants performed various activities including walking, running, and cycling (high/low intensity).
• Five smartphone placements (front pocket, back pocket, backpack, armband, fanny pack) were tested, alongside a SENS Motion® tracker.

Main Results:

• The front pocket placement showed the highest accuracy for cycling activities (89-93%).
• The back pocket placement yielded the highest overall classification accuracy for other activities.
• The SENS Motion® tracker generally outperformed smartphone placements, except for running, but several smartphone configurations were comparable.

Conclusions:

• Smartphone-based activity recognition using the BeSAFE+ app can be valid under specific conditions and placements.
• The front and back pocket placements demonstrated promising accuracy for specific activity types.
• Further research may optimize smartphone sensor applications for comprehensive physical activity monitoring.