Frailty assessment based on wavelet analysis during quiet standing balance test

Area of Science:

• Biomedical Engineering
• Gerontology
• Signal Processing

Background:

• Frailty is associated with increased risks of comorbidity, disability, falls, and fractures in the elderly.
• Assessing postural control during quiet standing is crucial for evaluating balance and fall risk in frail older adults.
• Real-time human motion tracking offers an accurate, portable method for kinematic and kinetic measurements.

Purpose of the Study:

• To investigate orientation and acceleration signals using wavelet transform during quiet standing balance tests.
• To differentiate between frail, prefrail, and healthy populations based on sensor data.
• To identify potential biomarkers for frailty using inertial magnetic sensor data.

Main Methods:

• Analysis of tri-axial inertial magnetic sensor data (orientation and acceleration) during quiet standing.
• Application of wavelet decomposition and principal component analysis for time-frequency signal analysis.
• Inclusion of participants from frail (n=14, 79±4 years), prefrail (n=18, 80±3 years), and healthy (n=24, 40±3 years) groups.

Main Results:

• Specific parameters derived from high-frequency components of orientation and acceleration signals, analyzed via wavelet transform, were associated with the frail syndrome.
• The absolute sum of wavelet detail coefficients at high frequencies differentiated frail individuals.

Conclusions:

• Identified signal parameters show potential as clinical markers for detecting frailty syndrome.
• These findings could enhance clinical assessments, rehabilitation strategies, and early identification of elderly individuals with frailty.
• Wavelet analysis of sensor data provides a promising approach for objective frailty assessment.