Development of aneuralinterface with spinal motoneurons for prostheticsand orthotics

The current impact ofneuralinterfaces is hindered by the need to design and calibrate solutions for a given user to achieve precise control. Success in operating a device with a neural interfacevaries with the neurophysiological states of the user,skill level, and information transfer rate of the interface. An intuitive and adaptive system, with a fast user training period, is needed to sustain reliable operation under non-identical setups and environments. This in turn would increase the number of cases where neural interfacingprovidesa viableapproach to improve mobility and quality of lifeof patients. Numerous attempts have been made at the interface level targeting different signal modalities. Among the proposed interfacing paradigms, non-invasive solutions (e.g. electroencephalogram, surface electromyogram) arepractical(since they do not require surgery) butlimited in information transfer(since they usually do not target individual neural cells). On the other hand, invasive approaches (e.g.,intra-cortical brain recordings)offer significantly higher information transfer rates(since they decode individual neurons) but are limited in clinical translation because of the risks associated to surgery and of technical challenges in transferring power and information to/from the interface. While transcutaneous links exposethe user to risk of infections, wireless neural interfaces arealsolimited by the power-bandwidth bottleneck.
We aim todevelop new optimized, self-learning and unsupervised decomposition and classification algorithmsfor HDsEMG. Novel algorithms and architectures based on deep learning approaches will be designedto automatically extract motoneuron spiking activity in realtime from HDsEMGrecordings without any user calibration. For this purpose, weenvision a framework based on autoencodersthat achieve dimensionality reduction by minimising the reconstruction error of the input recordings. This architecture is ambitious and highly challengingand will require extensive iterative simulation work and experimental data analysis.These developments will be followed by clinical trials that will representativelyfocus on upper limb prosthetic control(although the developed technology will have potential applications in a number of other applied areas).

The CDT students will help create solutions for amputees and people with debilitating conditions such as stroke and diabetes, reducing mortality and enabling them to live more satisfying, productive and fulfilling lives. These solutions, co-created with industry and people living with disabilities, will have direct economic and societal benefits. The principal beneficiaries are industry, P&O service delivery, people who need P&O devices, and society in general.
Industry
The novel methods, devices and processes co-created with users and industry will have a direct economic value to our industry partners (by the creation of IP, new products, and improved industry and academic links). Our CDT graduates will be the natural potential employees of our industry partners and for companies in the wider healthcare technology sector. This will help address the identified critical skills need and shortage leading to improvement in the UK's competitiveness in this rapidly developing and growing global market. The CDT outcomes will help UK businesses spread risk (because new developments are well founded) and more confidently enter new markets with highly skilled employees (CDT graduates).

P&O service delivery
Doctoral engineering graduates with clinical knowledge are needed to improve the deployment of advanced technologies in practice. Our main UK industry partner, Blatchford, stated: "As technology develops it will become easier for the end-user (the patient), but the providers (the clinicians) are going to need to have a higher level of engineering training, ideally to PhD level". The British Association of Prosthetists and Orthotists estimates that no more than ten practising P&O clinicians have a PhD in the UK. Long-term P&O clinical academic leadership will be substantially improved by the CDT supporting a select number of clinically qualified P&O professionals to gain doctorates.

Users
The innovation of devices, use of device and patient monitoring, and innovation approaches in LMIC should not only lead to improved care but also lower healthcare costs. Diabetes UK estimates that the total healthcare expenditure related to foot ulceration and amputation in diabetes was £1billion (2014-15), with 2/3 of this related to foot ulceration. Small innovations could lead to large cost savings if targeted at the right aspects of care (e.g. earlier adoption, and reducing device abandonment).
An ability to work is fundamental to a person's place in society and their sense of purpose and has a significant societal impact in all territories. This is perhaps greatest in LMIC where attitudes towards disability may still be maturing, and appropriate social care infrastructure is not always in place. In these cases, an ability to work is essential for survival.
Improved design approaches will impact on all users regardless of context, since the device solutions will better match local and individual user needs. Addressing issues related to prosthetic/orthotic device abandonment (e.g. cosmesis) and improved adherence should also lead to greater social participation. Improved device solutions will shift focus from what users "cannot do" to what they now "can do", and help progress attitudes towards acceptance of disability.
Societal
The majority of the global P&O users are of working age, and a key economic impact will be keeping users in work. The average age at amputation due to diabetes is just 52 in the USA but much younger in countries with less well-developed health care and trauma services (e.g. 38 in Iran). Diabetes UK reports that 35-50% of people are of working age at diagnosis and that there are around 70,000 foot ulcers in the UK, precursors to amputation. There is a similar concern for stroke survivors around a quarter of whom are of working age and are 2-3 times more likely to be out of work after eight years.