Gait analysis and validation using voxel data.

Area of Science:

• Biomechanics
• Computer Vision
• Medical Imaging

Background:

• Gait analysis is crucial for diagnosing neurological and musculoskeletal disorders.
• Accurate measurement of gait parameters like walking speed, step time, and step length is essential.
• Existing methods for gait analysis can be expensive or limited in scope.

Purpose of the Study:

• To present a novel method for extracting key gait parameters from a 3D voxel reconstruction.
• To validate the accuracy of the proposed method against gold-standard systems.
• To assess the feasibility of using two calibrated camera views for non-invasive gait assessment.

Main Methods:

• A 3D voxel reconstruction was generated using images from two calibrated camera views.
• Gait parameters (walking speed, step time, step length) were extracted from the 3D reconstruction.
• The extracted parameters were validated using a GAITRite Electronic mat and a Vicon motion capture system.

Main Results:

• Excellent agreement was achieved for walking speed measurements.
• Good agreement was observed for step time and step length.
• The results demonstrate the potential of the proposed method despite limitations in frame rate and voxel resolution.

Conclusions:

• The presented method offers a viable approach for extracting gait parameters using readily available camera systems.
• The findings support the use of 3D voxel reconstruction for gait analysis.
• Further refinement could enhance accuracy for step time and step length measurements.