A Sensorimotor Prosthesis for the Upper Limb

Overview

The finger and fingertip are the most frequently amputated body parts, due to work-related incidents. Yet because of space, weight and cost constraints, prosthetic fingers and fingertips are heavy and bulky with limited active motion and sensation. Most are basic variations on the hook and claw. Lower limb prostheses have become extremely technologically advanced in their design and materials, and upper limbs lag behind in all of these areas. This is due to the complexity of the anatomy and function of the upper limb compared to the lower. There are no commercially available prostheses that offers direct sensory feedback and as such, rely on visual feedback from the wearer. The original PROLIMB study (PROLIMB I) used a Leap Motion Controller (LMC) to investigate the type of grasp adaptation that have been undertaken by patients during the rehabilitation process following amputation and compared this to similar data from healthy volunteers. PROLIMB I also looked at refining the tactile feedback system by investigating the sensation felt on amputation sites in order to feed this information into the haptic feedback system. The vision of the PROLIMB II project is to build on the work completed in PROLIMB I and develop and combine mechanistic models of hand motion and haptic sensing to deliver novel, affordable body-powered prosthetic fingertip digits with enhanced motion and sensation to address current clinical needs and support the quality of life of amputees. With collaboration from the University of Warwick (UoW) and University College London (UCL), Steeper Group and Naked Prosthetics the PROLIMB II study will aim to model, design, fabricate and validate a body-powered prosthetic fingertip digit with integrated sensory feedback. The University Hospital Coventry \& Warwickshire (UHCW) will provide the clinical facility with which to assess the comfort, usability and acceptance of this prosthetic in the daily lives of patients with digit amputations. This project will be a proof of concept study with verification of the prosthetic in motion capture (gait) laboratories as well as the use of simple validation data collection over a longer period.

Worldwide, there are about 3 million upper limb amputees - 80% of these people live in low-income countries. Only 2% have access to prosthetic care. The World Health Organization and Prosthetics Outreach Foundation identified a number of reasons why patients in low-income countries do not have access to prosthetics: lack of trained personnel, frequent replacement and repair (prosthetic replacement is needed typically every 6-12 months for children and every 3-5 years for adults). This problem is not limited to developing countries: in the UK, the prostheses available from the National Health Service (NHS) have either a greatly simplified choice of actions available or are too expensive to be routinely funded. Upper-limb amputees suffering from traumatic injuries require long-term, frequent intensive medical care (work-related injuries cause the highest number of days away from work) and psychological support. NHS England spends approximately £60m/year on services for patients. Current prostheses are limited to cosmetic devices that fail to recover the original functionality of the hand. Up to 20% of the adult population with upper limb amputations choose not to use a prosthesis (Biddiss \& Chau, 2017) - the prescribed device is discarded as it is often seen as an obstacle. Of those that do accept a prosthesis, 26% of adults and 45% of children and adolescents are dissatisfied with their device and choose not to use them (NHS England, 2015). Hence, the NHS effectively does not provide service to over 50% of upper-limb amputees. Reasons cited include poorly developed fine motor control, absence of sensory function, weight and a lack of warmth and humanness. Currently the NHS is working in collaboration with companies who uitilize 3D printing technology to dramatically reduce the waiting time and costs of custom prosthetic hands. Open Bionics' Brunel Hand is a simple myoelectric (using the electrical signals from the muscle) control hook in the form of a human hand (open bionics, 2018). Although it is low-cost (£1200) and at a price point that the NHS may fund, there are still areas in which it could be improved, such as providing sensory feedback to the user. At present, the only form of feedback that upper limb prostheses provide is visual, rather than tactile. As such, the user must be able to see the prosthesis in order to use it effectively. A body powered prosthesis with more grasps and therefore more functionality would also be preferred. Body powered prosthetic devices are considered to have many advantages over myoelectric devices due to their robust, rugged and predictable nature. Other companies currently developing upper-limb prostheses include Touch Bionics, RSL Steeper, Ottobock and Defence Advanced Research Projects Agency (DARPA). The prosthetic devices created by these companies are much more expensive (approximately £20,000 - £75,000). All require an extrinsic power source (e.g. battery) and with high complexity comes high maintenance and risk of failure. None of the available prosthesis described provide any discrete tactile feedback. Although sensory information can be relayed to patients via direct integration with the central nervous system (Yildiz, et al., 2020), it is safer and simpler to provide this via haptic feedback. A body-powered hydro elastic haptic sensor has been developed and can relay fine sensory information from the tip of a prosthetic finger to an actuator that is in direct contact with the sensate residual limb. Although the resolution of the sensitivity in the residual limb is reduced (for example there is reduced ability to resolve between two discrete points of contact), the actual light touch sensation is retained after injury (e.g. on the residual palm or forearm). The haptic sensor and actuator can also amplify the signal to improve sensitivity. To design a body-powered sensory upper limb prosthesis, the action of the uninjured hand must be fully appreciated. The hand requires function over form, rather than vice versa. Although the visual appearance of a hand may be preferential, this can limit the design and capabilities of the prosthesis. Combining the synergetic expertise in developing parameterised mathematical models of limb motion from the UoW and in creating soft, stiffness-controllable robotic structures, and haptic feedback interfaces from UCL, the aim of this project is to model, design, fabricate and validate a body-powered prosthetic fingertip digit with integrated mechanical haptic feedback. Of key importance for the project's success is its transformative nature through close collaboration with experienced clinical experts from the UHCW and the UHCW Innovation Hub, who will offer consultation and clinical input throughout. The PROLIMB II study is a single centre pilot study that will be testing the feasibility of the trial design when integrating a tactile haptic feedback mechanism into prostheses. The PROLIMB II study will utilise a pre-existing CE marked finger digit prostheses (Naked Prostheses) and a non-CE marked prosthesis that has been created and designed by UCL. The CE marked finger digit prostheses will be custom made and manufactured by Naked Prosthetics (NP) - a commercial company in the USA that distributes their custom-fabricated finger prostheses to the UK through a secondary company called Steeper Group (SG). The non-CE marked prosthesis will be designed and manufactured by PROLIMB II researchers at UCL The PROLIMB research team will modify the prostheses so that they provide participants with sensory feedback when touching objects at the fingertip (full details of the device and modifications can be found in section 1.3). Testing will also be conducted with the haptic feedback mechanism attached to the residual stump, in order to collect sensation data. The study will recruit up to 10 participants. The total number of patients consented into the study may be greater than 10 as, following consent, NP are required to screen participants who will be receiving their prosthesis and so may deem some participants as unideal candidates for their prosthetic. If this occurs, participants will be offered the UCL prosthesis to use instead. Potential participants will have had one or more finger/thumb digits amputated through either accident or surgery within 1 year of consenting into the study. The wound will be completely healed at the point of consent and the prosthesis/haptic feedback system will not come into contact with an injured dermis. Potential participants will be identified by delegated individuals of the PROLIMB research team at University UHCW. If eligible, participants will be sent the study information pack and called to discuss the study further. If the participant is happy, they will be invited to take part in the study. For those who are receiving prosthetic devices from NP, these are custom-made for each participant; therefore consented participants will then need to be screened by experts at NP. This will involve the participants completing a partial hand information form and sending photos and videos of the participant's hands showing their level of amputation and their current range of motion. No personal data will be shared with Naked Prosthetics and all photo's/videos will be pseudo-anonymised (linked to a unique participant ID) and only show the participants hands/arms. If deemed suitable, devices will be custom made for participants and sent to UHCW for initial attachment and testing with the participant. Following this the prosthesis will be sent to collaborators at UCL who will modify the prosthetic to incorporate tactile haptic feedback into the fingertip and then returned to UHCW. The modified prosthetic will then be reattached to the participant and further testing will take place (details of testing procedures can be found in section 7.2). The prosthesis may also be supplied by UCL, further details of which can be found in section 1.3.2. Testing will continue over the next 10 months and take place at bi-monthly (± 3 weeks) intervals at UHCW gait lab. Tests will be performed both with and without the PROLIMB II prosthesis attached. All performance tests will be conducted whilst the participant is recorded using a VICON Motion Capture System (VMCS). During each visit a series of questionnaires and interviews will also take place to collect data related to participant's satisfaction with the PROLIMB II prosthesis. At the end of the study, participants will have the option to keep the PROLIMB II prosthesis if they wish, however they will be made aware that this is without any form of warranty for the devices.