A Microservices e-Health System for Ecological Frailty Assessment Using Wearables.

Area of Science:

• Gerontology
• Biomedical Engineering
• Artificial Intelligence

Background:

• Aging populations in developed countries face rising healthcare costs and workforce decline.
• Frailty is an early indicator of aging-related decline, necessitating timely detection and intervention.
• Current frailty assessments rely on manual methods focusing primarily on physical aspects.

Purpose of the Study:

• To develop and validate a non-intrusive, sensor-based system for early frailty detection in older adults.
• To integrate the assessment of Instrumental Activities of Daily Living (IADLs) with Basic Activities of Daily Living (BADLs) for a comprehensive frailty evaluation.
• To leverage machine learning for accurate and automated frailty status classification.

Main Methods:

• A microservices architecture system was designed to collect multi-dimensional sensory data during daily living activities.
• The system monitored older adults performing both BADLs and IADLs, capturing physical, cognitive, and social dimensions.
• A machine learning model was trained using the collected sensory data to assess frailty status.

Main Results:

• The developed machine learning model demonstrated superior accuracy in frailty assessment compared to previous methods relying solely on BADLs.
• The system provides an accurate, ecological, and non-intrusive approach to monitoring older adults' health status.
• The proposed method effectively integrates multi-dimensional data for a more holistic frailty evaluation.

Conclusions:

• The sensor-based system offers a flexible and automated solution for early frailty detection, aiding healthcare professionals.
• Integrating IADLs with BADLs in frailty assessment significantly enhances model performance.
• This technology has the potential to mitigate the impact of aging populations on healthcare systems.