Related Experiment Video
Updated: Jul 12, 2025
07:24
Home-Based Monitor for Gait and Activity Analysis
Published on: August 8, 2019
6.8K
Wearable sensor and machine learning estimate tendon load and walking speed during immobilizing boot ambulation
Michelle P Kwon1, Todd J Hullfish1, Casey Jo Humbyrd1
1Department of Orthopaedic Surgery, University of Pennsylvania, Philadelphia, PA, USA.
Scientific Reports
|October 23, 2023
Summary
This study introduces a wearable sensor system to monitor Achilles tendon load and walking speed during recovery. The system accurately predicts these metrics, offering a practical tool for clinical use.
Area of Science:
- Biomechanics
- Wearable Technology
- Rehabilitation Engineering
Background:
- Accurate monitoring of Achilles tendon loading and walking speed is crucial for effective rehabilitation after injury.
- Current monitoring methods can be burdensome and lack clinical practicality for longitudinal assessment.
Purpose of the Study:
- To develop a wearable sensor-based paradigm for accurate and low-burden monitoring of Achilles tendon loading and walking speed.
- To assess the performance of a Least Absolute Shrinkage and Selection Operator (LASSO) regression model for predicting these parameters.
- To investigate the impact of sensor parameter variations on model accuracy.
Main Methods:
- Ten healthy adults ambulated in an immobilizing boot with varying heel wedge angles and walking speeds.
- Data collection included 3D motion capture, ground reaction forces, and 6-axis inertial measurement unit (IMU) signals.
- LASSO regression was employed to predict peak Achilles tendon load and walking speed, with sensor parameter effects analyzed.
Main Results:
- Walking speed prediction models achieved a mean absolute percentage error (MAPE) of 8.81% ± 4.29%, outperforming Achilles tendon load models (MAPE: 34.93% ± 26.3%).
- Subject-specific models demonstrated superior performance compared to non-subject-specific models.
- Altering sensor parameters, such as removing the gyroscope or decreasing sampling frequency, had inconsequential effects on model performance.
Conclusions:
- A wearable sensor paradigm utilizing LASSO regression can accurately predict Achilles tendon loading and walking speed (MAPE ≤ 12.6%) in individuals using an immobilizing boot.
- This approach offers a clinically implementable strategy for longitudinal monitoring of patient loading and activity during Achilles tendon injury recovery.
