Automated feet detection for clinical gait assessment

Area of Science:

• Biomedical Engineering
• Computer Vision
• Gait Analysis

Background:

• Accurate clinical gait metrics are crucial for patient assessment and rehabilitation.
• Current gait analysis methods often require specialized equipment and controlled laboratory settings.
• There is a need for accessible, high-accuracy gait assessment tools in real-world environments.

Purpose of the Study:

• To develop and validate a computer vision method for estimating clinical gait metrics in unconstrained environments.
• To enable accurate gait analysis using readily available cameras.
• To provide a tool for therapists and non-specialists for remote gait assessment.

Main Methods:

• Utilized background subtraction to generate subject silhouettes.
• Employed a randomized decision forest for robust foot detection.
• Estimated key gait parameters including stride length, step length, cadence, and walking speed.

Main Results:

• The system demonstrated accurate estimation of clinical gait metrics.
• Validation through error analysis on manually annotated videos confirmed system performance.
• Successful testing in diverse outdoor settings highlighted system robustness.

Conclusions:

• The developed computer vision method provides a significant advancement in accessible gait analysis.
• This technique allows for high-accuracy gait metric estimation in unconstrained, real-world settings.
• It empowers clinical therapists and non-specialists with a versatile tool for remote patient monitoring and assessment.