Walking-age analyzer for healthcare applications.

Area of Science:

• Biomedical Engineering
• Gerontology
• Human Movement Science

Background:

• Assessing gait and mobility is crucial for understanding age-related changes and health status.
• Current methods for gait analysis can be complex and may not capture subtle age-related differences effectively.

Purpose of the Study:

• To develop and validate a system for analyzing and identifying walking-age patterns using wearable sensors.
• To categorize individuals into distinct walking-age groups based on their gait characteristics.
• To explore the potential of gait analysis for early detection of health or mobility issues.

Main Methods:

• Collected 3-D motion data (accelerometer and gyroscope) from 79 volunteers aged 10-83 years.
• Constructed a walking pattern database and applied feature extraction and clustering techniques (K-means).
• Utilized low-pass filtering and empirical mode decomposition for signal processing and analysis.

Main Results:

• Identified three distinct walking-age groups: children (≤10 years), adults (20s-60s), and elders (70s-80s).
• Demonstrated that simple walking pattern signals can effectively categorize individuals into these age groups.
• Showcased the ability to detect individuals whose walking patterns suggest a higher 'walking age' than their chronological age.

Conclusions:

• Walking-age pattern analysis using wearable sensors is a viable method for age categorization.
• Deviations towards a higher 'walking age' may indicate underlying health problems, such as joint weakness, musculoskeletal issues, or fall risk.
• This system offers a potential tool for non-invasive health monitoring and early detection of mobility impairments.