A smart device for non-invasive ADL estimation through multi-environmental sensor fusion.

Area of Science:

• Biomedical Engineering
• Health Informatics
• Sensor Technology

Background:

• Traditional methods for assessing Activities of Daily Living (ADL) suffer from limitations like patient recall bias, privacy concerns, and reliance on wearable device adherence.
• Existing ADL monitoring systems often require invasive methods or face challenges with user compliance and data security.

Purpose of the Study:

• To introduce the Smart Plug Hub (SPH), a novel non-invasive system for accurate estimation of patient ADLs.
• To overcome the limitations of conventional ADL assessment techniques through advanced sensor technology and data processing.

Main Methods:

• The SPH system employs sensor fusion to analyze time-series environmental signals for ADL estimation.
• Optimized resource utilization through "device collaboration" to process segmented time-series environmental data on an edge device.
• Data segmentation and analysis on an edge device (SPH) for efficient processing.

Main Results:

• The SPH system demonstrated significant accuracy in classifying specific ADLs.
• Achieved 75% accuracy in classifying kitchen-based ADLs, such as eating activities.
• Attained 85% accuracy in classifying toilet-based ADLs.

Conclusions:

• The Smart Plug Hub (SPH) offers a promising non-invasive solution for accurate ADL monitoring.
• SPH technology has substantial implications for advancing healthcare and patient care through improved daily living activity monitoring.
• The system's high accuracy in classifying kitchen and toilet ADLs highlights its potential for real-world healthcare applications.