A Center of Mass Estimation and Control Strategy for Body-Weight-Support Treadmill Training

Area of Science:

• Neurorehabilitation
• Robotics
• Biomechanics

Background:

• Walking disorders are prevalent after stroke, impacting recovery.
• Current body weight support (BWS) systems offer fixed support, potentially increasing fall risk and reducing patient engagement.

Purpose of the Study:

• To enhance the efficiency and safety of BWS treadmill training for individuals with hemiplegia.
• To develop a dynamic BWS system that adapts to the user's center of mass (CoM) during gait training.

Main Methods:

• A novel strategy was developed to regulate BWS system height, dynamically tracking the subject's center of mass (CoM).
• CoM height was estimated using a long-short term memory (LSTM) network, processing inputs from a locomotion recognition system (walking speed, body-height to leg-length ratio, non-paretic hip/knee angles).

Main Results:

• The CoM height regulation strategy significantly increased hip and knee joint range of motion by 34.54% and 25.64%, respectively, compared to constant support.
• Paretic side stance phase duration increased by 14.6% of the gait cycle, promoting gait symmetry and stability.
• The CoM height tracking closely matched data from non-disabled individuals, indicating improved stability.

Conclusions:

• The proposed CoM height tracking strategy enhances BWS treadmill training for post-stroke hemiplegia.
• This adaptive approach improves joint mobility, gait symmetry, and overall stability, offering a safer and more effective rehabilitation tool.