Decoding phantom limb movements from intraneural recordings

Area of Science:

• Neuroscience
• Biomedical Engineering
• Prosthetics

Background:

• Limb loss results in sensorimotor deficits, often requiring prosthetic devices.
• Direct neural recording for prosthetic control faces challenges like low signal amplitude and nerve interfacing difficulties.

Purpose of the Study:

• To investigate intraneural multichannel electrodes for accessing motor signals in transfemoral amputees.
• To decode volitional phantom limb movements for advanced prosthetic control.

Main Methods:

• Implantation of transversal intrafascicular multichannel electrodes in distal sciatic nerve branches of two transfemoral amputees.
• Recording multiunit activity associated with phantom knee, ankle, and toe movements.
• Utilizing a Spiking Neural Network (SNN)-based decoder for movement prediction.

Main Results:

• Identified joint- and direction-specific neural activity corresponding to phantom limb movements.
• SNN decoder outperformed conventional methods in predicting attempted movements.
• Integration of intraneural and intermuscular signals further improved decoding accuracy.

Conclusions:

• Intraneural recordings from the sciatic nerve can decode phantom limb movements.
• This technology supports the development of bidirectional, neurally-controlled prosthetic systems.
• Early segregation of motor and sensory maps within the nerve was observed.