Simulation of normal and pathological gaits using a fusion knowledge strategy

Area of Science:

• Biomechanics
• Robotics
• Medical Engineering

Background:

• Gait distortion is a key indicator of various pathological disorders.
• Traditional gait analysis relies heavily on physician expertise, lacking objective quantification.
• Existing models often fail to capture the complexity of human locomotion.

Purpose of the Study:

• To develop a novel human gait model integrating Center of Gravity (CoG) trajectory and learned heel paths.
• To enable the reproduction of both normal and pathological kinematic gait patterns.
• To enhance diagnostic and prognostic capabilities for gait-related disorders.

Main Methods:

• Approximating Center of Gravity (CoG) trajectory using an extended physical compass pendulum model with energy accumulators.
• Incorporating learned heel paths from real-world data to refine the physical model.
• Estimating joint trajectories via classical inverse kinematics.

Main Results:

• The model achieved a high correlation coefficient of 0.96 when compared with standard gait patterns.
• Successfully simulated neuromuscular diseases like Parkinson's (phases 2-4) with an average correlation of 0.92.
• Reproduced clinical signs such as Crouch gait with high accuracy.

Conclusions:

• The fused gait model offers a robust and accurate method for analyzing human locomotion.
• This approach provides a more objective tool for diagnosing and understanding gait pathologies.
• The model holds potential for improving clinical interpretation and patient outcomes in neurology and orthopedics.