Bi-directional Neural interface for Prosthetic Control

点击翻译按钮获取中文摘要

Bioelectronic medicine is a growing field that uses electrical devices to provide therapeutic effects. This allows for the treatment of conditions that can be intractable to drugs, or require a therapy with a high level of spatial specificity. For example, up to a third of people with epilepsy will develop refractory epilepsy, where anti-epileptic drugs don't work to sufficiently control seizures. For these people, bioelectronic medicines that use electrical stimulation targeted to the focal region of the epileptic activity have been shown to reduce or terminate seizures at onset, and represent a potential treatment for their condition. As the field of bioelectronic medicine has developed, so has our understanding of the workings of the nervous system. This increase in understanding has made it possible to closely integrate computer systems with the nervous system, providing a path to creating new technologies based on these brain-computer interfaces (BCIs). A key focus for medical BCIs has been in providing prosthetic devices - machines that supplement or replace natural capabilities lost to disease or injury - for those with damage to the peripheral nervous system (PNS). In cases such as spinal cord injury or loss of a limb, a BCI could allow for the complete restoration of that person's physical capacity and quality of life.This PhD project is focussed on creating and analysing a PNS-BCI to be implanted into the arm, capable of full bi-directional communication. In essence, such a device would be able to detect nervous signals sent to the limb by the brain and use those to control another device, such as a prosthetic arm; it could then also use sensor data from the device to stimulate the nerves in turn, allowing the user to experience from the arm the sensations that would come from a natural limb. A major issue faced by users of current hand and arm prosthetics, which are designed only with control in mind and so lack sensory feedback, is "embodying" the artificial limb - the sense of the limb being a part of their body. Current control systems that use skin-level muscle activity recordings are also unable to provide dexterity close to a natural limb, with prosthetics limited to a small number of pre-set hand postures. Between these two problems, many users of hand and arm prosthetics end up desisting in their use, with most citing frustration with the capabilities of the prosthetic as a reason for abandoning it. If the prosthetic could be fully integrated into the nervous system through a BCI, as outlined above, these issues could be solved to allow for a much greater quality of life for amputees.The research questions for this project are:1. Can electrical stimulation produce the sensations of touch and proprioception?2. Can the activity of the motor neurones be decoded to classify what action the user wants to perform?3. Can a non-invasive electrode, such as a cuff electrode, selectively stimulate specific fascicles within the nerve?4. How do different stimulation parameters affect both the acute behaviour and chronic health of nerves implanted with electrodes?These research questions will be answered by using a combination of simulations, ex vivo and in vivo tests to quantify the behaviour of different device designs and operational parameters. No research or available device to date provides full bi-directional communication in a BCI; achieving this would represent a significant step towards fully functional restoration for people with limb loss. This research project aligns strongly with the EPRSC's research interests in Assistive Technology, Biological Informatics and Clinical Technologies.