Multiaxis Prosthesis Control Through an Osseointegrated Neural Interface
Abstract
This research addresses the RESTORE Focus Area of Optimization of Warfighter Performance Following Limb Loss. Persons who experience amputation lose, not just the structure of a limb, but also control over its many functions. In lower-limb amputation, this loss of control leads to poorer balance, greater incidence of falls, and lower activity levels. The limitations are threefold: currently available devices have limited functionality; the current socket technology does not provide a good interface; and there are no widely used means of exchanging information between the person and the prosthesis for control and sensation. This research will contribute to solving all three of these problems. First, we will refine a recently developed two-axis prosthetic ankle intended to help improve balance and reduce discomfort for current prosthesis users. The prosthesis will be programmed to automatically adapt to the terrain and to different movement behaviors. We will test this prosthesis to determine its benefits in activities, including sloped walking and walking on curved paths or turns. We expect these benefits will include reduced forces on the residual limb and higher ground clearance to avoid tripping. In everyday life, these improvements could lead to higher activity levels; greater participation in work, social life, and exercise; and potentially a return to combat. The risk to using this new device is low, similar to using existing prostheses. The main risk is of device malfunction, either components breaking or a controller error. If the benefits to the users are as anticipated, we will pursue transfer of this prosthetic ankle technology to the marketplace through a commercial partner. This process could take 2-5 years of further development. Second, we will work to advance the science of osseointegration and neural interfaces through surgical and technological advancement. We will use sheep as a model of human amputation and implant both a mechanical anchor to attach a prosthesis to the bone and a two-way neural recording and stimulation system to exchange information. The mechanical anchor will be implanted in the bone and will protrude through the skin, providing a firm mechanical attachment of the prosthesis to the skeleton. We will build a special version of the new prosthesis suitable for use by the sheep. We will implant electrodes around nerves and bones in the upper leg to record neural signals the sheep produces. These signals will be transmitted out of the body to control the prosthetic foot. We will also implant both the damaged nerve endings and more electrodes from the neural interface inside the bone so the neural interface can stimulate the sensory neurons after they heal. These electrodes will stimulate the nerves with information about the forces sensed by the prosthesis. Together, this system will provide an advanced multiaxial prosthetic foot with neural control and sensory feedback. We will test the ability of the sheep to control the prosthesis and to sense ground forces using this two-way neural interface. The sheep will stand, walk on level and sloped ground, and walk over uneven ground with the control and the sensing functions turned on or off. We will test whether specific scenarios lead to repeatable movements of the prosthesis and whether sensory feedback changes these movement patterns. The risks to this approach are that the bone interface may not heal or may not be stable; the prosthetic anchor protruding through the skin could become a site of infection; and the neural interface could cause pain or undesired responses to stimulation. But, if the sheep adapt well and are able to use the neural interface to control the prosthesis, this result will establish that this whole combined system is a viable approach to pursue for establishing control and sensation with prostheses. We will then pursue further studies to make the osseointegration and the neural inter
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 10, 2021
- Source ID
- W81XWH2010884
Entities
People
- Peter G. Adamczyk
Organizations
- United States Army
- University of Wisconsin–Madison