Validation of a 3D Markerless Motion Capture Tool Using Multiple Pose and Depth Estimations for Quantitative Gait...

Area of Science:

• Biomechanics
• Motion Analysis
• Human Movement Science

Background:

• Gait analysis is crucial for assessing walking and identifying functional deficits.
• Traditional marker-based motion capture is expensive, labor-intensive, and requires expertise.
• Marker-less motion capture offers a potentially more accessible alternative.

Purpose of the Study:

• To evaluate the accuracy and reliability of a 3D marker-less motion capture (3D MMC) system.
• To compare the 3D MMC system against a gold-standard motion capture (MOCAP) system for gait analysis.
• To assess hip and knee joint angles during gait at various speeds.

Main Methods:

• A 3D MMC system utilizing pose and depth estimation was employed.
• Fifteen healthy participants performed gait tasks at 0.7, 1.0, and 1.3 m/s.
• Data from the 3D MMC system were compared with a gold-standard MOCAP system.

Main Results:

• The 3D MMC system demonstrated high accuracy (LCC > 0.96) and excellent inter-session reliability (RMSE < 3°).
• Moderate-to-high accuracy with constant biases was noted during specific gait events.
• Discrepancies were attributed to differing sample rates and kinematic methodologies between systems.

Conclusions:

• The 3D MMC system shows significant potential as a reliable tool for gait analysis.
• This system offers improved usability for both clinical and research settings.
• Future research should address limitations such as participant population and pose estimation model key points.