Regenerative Peripheral Nerve Interfaces (RPNIs) Improve Functional Outcomes and Intuitive Control of a Prosthetic...

Current commercially available surface electromyography (EMG) prosthetic control approaches are not intuitive and cannot control a large number of movements reliably. Often adding wrist motion compromises the ability to select from multiple grips. However, not having wrist motion contributes to significant compensatory movements. This case study compared functional, biomechanical, and cognitive outcomes between myoelectric control approaches incorporating active wrist rotation using surface EMG and intramuscular EMG signals from regenerative peripheral nerve interfaces (RPNIs) and residual muscles. One female with unilateral transradial amputation performed a series of functional assessments using a multi-grip myoelectric prosthetic hand with four control approaches. Linear discriminant analysis (LDA) pattern recognition (PR) classifiers were trained to decode either surface EMG (surface-PR) or intramuscular EMG from residual muscles and RPNIs (RPNI-PR) into functional grips only or functional grips and wrist rotation. The participant had better performance in the Clothespin Relocation Test (CRT) (2.3x fewer dropped clothespins) and the Coffee Task (2.5x faster, 12 less errors), with lower cognitive load (∆ = 50pts) when using RNPI-PR compared to Surface-PR to control functional grips. Active wrist rotation improved performance for both signal sources in the CRT (∆ = 42%) and reduced some movement compensations for Surface-PR. Overall, prosthetic wrist control using signals from RPNIs and residual muscles provided intuitive control while improving function compared to surface EMG.Clinical relevance- Intramuscular electrodes in RPNIs and residual muscles can provide good control of multiple grips and wrist rotation, without adding cognitive burden compared to surfaces based control.