Maximizing the Information Transfer Rate of a Myoelectric Classification System for Individuals With Spinal Cord...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Spinal cord injury (SCI) severely impacts independence and quality of life.
• Myoelectric control systems offer potential for assistive technologies but often overlook SCI-related neurological diversity.

Purpose of the Study:

• To demonstrate that optimizing myoelectric control system parameters (gesture number, gesture rate) enhances information transfer rate (ITR) for individuals with SCI.
• To characterize the relationship between optimal design parameters and specific patterns of neurological impairment.

Main Methods:

• Recruited ten uninjured and ten participants with SCI.
• Utilized an 8-channel myoelectric control system to test various gesture numbers and rates.
• Measured information transfer rate (ITR) to balance gesture complexity and classification accuracy.

Main Results:

• Optimized myoelectric control systems significantly improved ITR in the SCI group (from 25.8±10.9 to 31.8±8.0 bits/min, p=0.002).
• Found significant correlations between optimal gesture number and SCI impairment metrics.
• Demonstrated that ITR can be increased by personalizing systems to individual SCI characteristics.

Conclusions:

• Personalized optimization of myoelectric control systems is crucial for improving assistive technology performance in individuals with SCI.
• Understanding the link between impairment patterns and optimal system parameters can lead to more effective neuroprosthetics.
• This study highlights the potential for tailored myoelectric interfaces to restore function and enhance independence post-SCI.