Development of a Sensor-Based Behavioral Monitoring Solution to Support Dementia Care

Area of Science:

• Gerontology
• Biomedical Engineering
• Computer Science

Background:

• Mobile and wearable technology offers novel methods for behavioral monitoring via sensor data.
• Such technology can significantly aid clinical research and practice for individuals with dementia.
• Effective tools are needed for real-life behavioral measurement in natural settings.

Purpose of the Study:

• To develop and validate algorithms for measuring mobility and activity using sensor data.
• To establish a technical framework for collecting necessary sensor data with off-the-shelf devices.

Main Methods:

• Developed a mobility module using GPS data to calculate spatial, temporal, and count-based metrics.
• Implemented activity measurement including activity bout extraction and daily step counts.
• Collected data from smartwatches and mobile phones, utilizing open-source resources.

Main Results:

• The solution successfully measured travel trajectories and mobility metrics from location data.
• Multimodal activity bouts during travel were accurately extracted.
• Step count data indicated daily activity levels, though smartwatch step counts were higher than phone counts, raising validity concerns.

Conclusions:

• A comprehensive behavioral monitoring solution for real-life settings was created.
• This solution is replicable for various applications requiring mobility and activity insights.
• The study provides a valuable tool for clinical research and practice.