This program seeks to develop technology to reliably extract information from the nervous system at a scale and rate necessary to control many degree-of-freedom machines, such as high-performance prosthetic limbs.
Research and development of neural prostheses based on stimulation, such as cochlear implants, has led to clinically reliable, commonplace, and publicly accepted products that have restored lost function to a large number of patients. Although prostheses based on recording neural activity hold great promise and have high relevance to the Department of Defense (DoD), there are two fundamental and well-known obstacles that are preventing their successful transition to clinical use. Both obstacles deal with reliability. First, miniature and portable neural-machine interfaces cannot reliably obtain accurate information from neural tissue over a period of decades. Second, prosthesis systems cannot reliably use measured signals to control the prostheses with high speed and resolution.
DARPA is interested in addressing the specific fundamental challenges preventing clinical deployment of Reliable Neural-Interface Technology (RE-NET), facilitating its potential to enhance the recovery of injured Service Members and assist them in returning to active duty. Program developments will impact the broad community of patients with medical amputations, spinal cord injuries, and neurological diseases.
The RE-NET program will first build a foundation of understanding why these neural interfaces do not remain operational over multiple years. The Histology for Interface Stability over Time team will delve into the interactions between biotic and abiotic systems and what mechanisms lead to interface failure. The direct interface between tissue and electronics is not the only factor that affects reliability. The Reliable Peripheral Interfaces (RPI) and the Reliable Central-Nervous-System Interfaces (RCI) teams will build complete systems that develop a deeper understanding of how motor-control information is conveyed from neural tissue through implanted interfaces and electronics to efficient and robust decoding algorithms. RPI will interface with the peripheral motor system, and RCI will interface directly with the brain and spinal cord.