Markerless vs. Marker-Based Gait Analysis: A Proof of Concept Study.

Area of Science:

• Biomechanics
• Medical imaging
• Rehabilitation technology

Background:

• Human gait analysis is crucial for evaluating neurological diseases and neuromotor disorders.
• Marker-based motion capture is the gold standard but is costly, time-consuming, and can affect natural movement.
• Markerless videography systems are being developed as alternatives, but quantitative 3D comparisons are scarce.

Purpose of the Study:

• To introduce and evaluate a novel RGB video-based markerless system for 3D gait analysis.
• To quantitatively compare the performance of the markerless system against a traditional marker-based motion capture system.

Main Methods:

• Developed a markerless 3D gait analysis system using computer vision and deep learning.
• Acquired simultaneous data from both markerless and marker-based systems for 16 participants walking indoors.
• Compared spatio-temporal parameters and joint angles between the two systems.

Main Results:

• The markerless system achieved comparable spatio-temporal gait parameters to the marker-based system.
• Joint angle measurements were similar, with only minor underestimation of ankle and knee flexion.
• The findings support the viability of markerless techniques in gait analysis.

Conclusions:

• Markerless 3D gait analysis using deep learning is a feasible and promising alternative to marker-based systems.
• This technology can potentially reduce costs and improve the naturalness of gait assessments in clinical settings.
• Further validation can pave the way for wider adoption in medicine and rehabilitation.