RPI: Reliable Peripheral Interfaces

The Reliable Peripheral Interfaces (RPI) effort seeks to demonstrate peripheral-nervous-system (PNS) interfaces that can reliably extract motor-control information for intuitive control of high-performance upper-limb prosthetics. 

This effort includes a variety of PNS-interface approaches such as nerve cuffs, penetrating electrode arrays, regenerative interfaces, tissue-engineered biological constructs, non-penetrating devices, invasive electromyography (EMG) and sensory-input (stimulation) systems. 

Technical areas

1. Create and demonstrate clinically viable reliable-tissue interfaces to peripheral nerves and muscles.  Investments were made in the development of ultra-compliant probes targeting the dorsal and ventral spinal nerve roots, miniature wireless implantable EMG systems, a flat interface nerve electrode (FINE), regenerative electrode technologies, and refinement of Utah slant electrode arrays (USEA) resulting in first-in-human demonstrations.  These peripheral interfaces must enable stable, robust and high-channel count extensive recording of limb-control-signal activity.   
2. Demonstrate the clinically viable tissue-interface electronics and packaging necessary to enable the development and testing of reliable peripheral interfaces designed to control many-DOF prosthetic limbs.  Efforts focused on implantable wireless systems for EMG recording and the high performance electronic microsystems for epineural recording.  
3. Develop clinically viable algorithms and subsystems for reliably decoding limb-control information from the detected peripheral signals. Efforts in sparse PCA, Markov random fields, simultaneous decode of EMG, beam-forming, time-delayed artificial neural nets, and creating motor neuron pool models were launched.   
4. Demonstrate an advanced peripheral-interface system that can accurately detect and decode peripheral limb-control signals in a manner that reliably results in significant functional benefit.  Human-use demonstrations of military-relevant and high-DOF control of a prosthetic limb using implantable EMG, surface EMG, and inter-neural recordings are expected.   
5. Demonstrate clinically viable systems that provide tactile sensory and/or proprioceptive limb feedback via stimulation of the peripheral nervous system.  Reliable stimulation of the peripheral nerves and dorsal root ganglion were demonstrated.