Myoelectric interface training enables targeted reduction in abnormal muscle co-activation

Area of Science:

• Neuroscience
• Rehabilitation Medicine
• Biomedical Engineering

Background:

• Abnormal muscle co-activation is a key factor in post-stroke motor impairment.
• A myoelectric computer interface (MyoCI) was developed to retrain arm muscles and reduce abnormal co-activation.
• The impact of MyoCI training on overall arm intermuscular coordination, beyond trained muscles, needed further investigation.

Purpose of the Study:

• To investigate changes in arm muscle synergies after MyoCI training in stroke survivors.
• To determine if MyoCI training affects only targeted muscles or broader intermuscular coordination.
• To assess the relationship between changes in muscle synergies and motor recovery.

Main Methods:

• Thirty-two stroke survivors underwent 6 weeks of MyoCI training.
• Non-negative matrix factorization (NMF) was used to identify muscle synergies during a reaching task before and after training.
• A new metric, the disparity index (DI), was introduced to quantify changes within muscle synergies.

Main Results:

• No consistent changes in the number of muscle synergies were observed across all participants.
• Traditional synergy composition analysis showed no significant changes post-training.
• Participants who responded to MyoCI training exhibited increased disparity within synergies, indicating targeted muscle decoupling.

Conclusions:

• MyoCI training effectively reduced arm impairment by selectively targeting and decoupling specific muscles.
• The nervous system demonstrates a capacity for fractionated motor learning even after stroke.
• MyoCI training successfully enabled stroke survivors to reduce abnormal co-activation in trained muscles.