Real-Time Prediction of Leg Joint Kinematics during Steady-State Walking and Task Transitions

Area of Science:

• Biomechanics and Robotics
• Rehabilitation Engineering

Background:

• Locomotor impairments significantly affect mobility.
• Advanced orthoses and prostheses aim to restore function.
• Accurate control of these devices is a key challenge, often requiring reference joint kinematics.

Purpose of the Study:

• To introduce a novel method for estimating reference leg joint kinematics.
• To utilize kinematic data from the contralateral (opposite) limb for estimation.
• To improve control strategies for active orthoses and prostheses.

Main Methods:

• A fusion approach combining Complementary Limb Motion Estimation (CLME) and Adaptive Oscillators (AOs).
• A logistic function is used to integrate the outputs of the two estimators.
• The method was applied to estimate hip joint kinematics during walking.

Main Results:

• The fused method demonstrated comparable performance to CLME during transitional phases.
• Post-transition, the method shifted towards AO performance for improved accuracy.
• Cycle-by-cycle correlation analysis showed enhanced estimation during quasi-steady-state locomotion.

Conclusions:

• The novel fused estimation method offers improved accuracy for leg joint kinematics.
• This technique can enhance the control of assistive devices for locomotion.
• The adaptive nature of the fusion method allows for better performance across different gait phases.