Lumbar Acceleration Gait Estimation: "Step-by-Step" Algorithm Updates and Improvements

Area of Science:

• Digital Health
• Biomedical Engineering
• Wearable Technology

Background:

• Accelerometry-based digital health technologies are increasingly used for gait monitoring due to low participant burden and ease of deployment.
• Remote gait monitoring provides continuous, quantifiable health metrics over extended periods, offering a more comprehensive perspective than traditional clinic visits.
• The SciKit Digital Health (SKDH) package offers a device-agnostic framework for standardizing gait metrics across various devices.

Purpose of the Study:

• To introduce literature-informed enhancements to the SKDH gait algorithm.
• To improve the algorithm's performance against reference standards.
• To reduce the need for manual parameter adjustments across diverse populations.

Main Methods:

• A block-wise refinement process was employed to examine and enhance individual algorithmic components.
• The cumulative impact of these enhancements on the complete gait algorithm and generated metrics was evaluated.
• The study utilized data from healthy adult and pediatric participants.

Main Results:

• The novel gait event estimation method reduced mean absolute error by over 50% compared to the predecessor.
• Intraclass correlation values for gait metric concordance with in-laboratory references improved from 0.50-0.74 to 0.81-0.90.
• Systematic bias in gait speed estimation was rectified, narrowing the difference from the reference from 0.065-0.230 m/s to 0.00-0.03 m/s.

Conclusions:

• The study provides robust evidence supporting the validity of the enhanced gait algorithm.
• A single lumbar accelerometer can capture gait characteristics with high accuracy and reliability.
• The algorithm performs well across various walking speeds and age groups.