Discrete vs. Continuous Mapping of Facial Electromyography for Human-Machine-Interface Control: Performance and Training Effects

This study compares two different ways of using facial muscle signals (sEMG) to control a computer keyboard for people with spinal cord injuries who can't use their hands. One system uses distinct muscle movements for each action, while the other allows for more continuous control based on the strength of muscle signals. Participants trained on one system and then switched to the other. The researchers measured how quickly and accurately they could type using each system to see which one was more effective and if training on one system helped with the other. The results showed that the continuous system allowed users to type faster, and training on the continuous system did not improve performance when switching to the discrete system. This suggests that the continuous system may be more effective overall, although the discrete system could be modified for some users.

• 1
  Users of the continuous system achieved higher information transfer rates (ITRs) than those using the discrete system, both initially and after training.
• 2
  When continuous system users switched to the discrete system, their ITRs significantly decreased, while discrete users did not show a significant change when switching to the continuous system.
• 3
  Path efficiency increased with training in the continuous group, indicating that users learned to move the cursor more directly to the target over time.