Predicting gait and timed up and go time using foot-tapping test and accelerometer

Area of Science:

• Biomedical Engineering
• Gerontology
• Rehabilitation Science

Background:

• Functional mobility is crucial for independence and quality of life.
• Traditional assessments like gait analysis and Timed Up and Go (TUG) require space and trained personnel, limiting routine use.
• A seated foot-tapping test offers a potentially practical, low-burden alternative.

Purpose of the Study:

• To investigate if accelerometer-derived features from unilateral foot-tapping can predict gait and TUG times.
• To assess the feasibility of foot-tapping as a proxy for functional mobility across the adult lifespan.

Main Methods:

• Seventy-eight adults (19-95 years) performed 30-second unilateral foot-tapping trials.
• Triaxial accelerometer signals captured time, frequency, and time-frequency domain features.
• Least Absolute Shrinkage and Selection Operator (LASSO) and Random Forest models predicted gait and TUG times via cross-validation.

Main Results:

• Significant correlations (p<0.05) were found between foot-tapping features and gait/TUG times.
• Time-frequency and time-domain features, including tapping rate and variability, were most associated with outcomes.
• LASSO regression models demonstrated lower Root Mean Square Error (RMSE) compared to Random Forest, particularly for dominant-foot predictions.

Conclusions:

• Accelerometer-based unilateral foot tapping shows promise as a fast, space-efficient, and low-burden complement to conventional mobility assessments.
• Further validation in diverse clinical populations and comparison with other regression techniques are recommended.