Muscles and microchips: AI-enabled smart prosthetics
Patients who have had life-changing traumatic injuries that require amputation can be given prosthetics to help them regain some use of their limbs. However, as current prostheses don't provide sensations, the small adjustments needed for everyday tasks such as gripping a pen without dropping it or holding an egg without crushing it are very difficult, requiring huge concentration and brain power to complete.
This project aims to assist these patients to regain an element of normality within their everyday lives. Focusing on patients who have lost their upper limbs, researchers from the University of Edinburgh are using AI to develop a prosthesis that reduces some of the brain power required to conduct these everyday tasks while enhancing the user’s control of the limb.
The project uses a range of artificial sensors that replicate the lost sensory capability of the natural hand. The AI within the prosthesis interprets sensory data and allows the artificial hand to react to the environment and make some of the movements required on behalf of the user, mimicking the automatic movements of natural limbs. For example, using the sensor data, the AI can recognise the shape and firmness of a cup and grip the handle appropriately. By taking some of the cognitive load required to conduct these tasks away from the user, this smart prosthesis makes everyday physical tasks easier for the user and allows them to focus instead on higher order movement tasks such as controlling multiple individual finger and wrist movements simultaneously.
A key challenge in making prosthetic limbs more agile is to give the limb the ability to perform quick movement adjustments using simple sensors instead of heavy instrumentation. The team is currently building and refining a prototype and AI for Good funding is helping to procure new sensors and electronics. The project team is also looking to engage with more patients to help with the development of the next stage of prototype.
Aidan Roche, MBBS PhD BEng, Clinical Lecturer at the University of Edinburgh and Principal Investigator on the project, says the implications of the technology are significant. "The AI techniques we are using can be translated to other applications such as lower limb prosthetics or exoskeletons for patients with intact but non-functioning limbs or spinal cord injuries. AI has enormous potential across other areas of healthcare, from aiding quicker diagnosis of diseases to discovering new antibiotics in an era of microbiological resistance. AI makes all this possible by decoding large volumes of data in a fraction of the time humans alone would be able to achieve this."
This project is led by Aidan Roche of the University of Edinburgh in partnership with Robust Autonomy & Decisions Group and the Alan Turing Institute. The project has received funding from Nesta in Scotland’s AI for Good programme.