Development of methods in musculoskeletal modelling and applications to sports injury and orthopaedics
Lead Research Organisation:
Imperial College London
Department Name: Bioengineering
Abstract
My proposed PhD research project would aim to build upon my MSc project in the department of Biomechanics. Working under the supervision of Professor Anthony Bull, I explored a novel application for functional electrical stimulation (FES), to manipulate gait for the prevention and mitigation of osteoarthritis (OA) of the knee. As a high prevalence disease that often affects relatively young people, and for which there is a paucity of effective therapies in the early stages, OA imposes heavy socioeconomic costs. Our approach exemplifies how technology may be applied in such instances to ease the disease burden.
Employing advanced in-house musculoskeletal modeling technology, and alongside experts in this relatively new field, I was able to leverage my clinical knowledge, drawing on my experience to suggest optimum anatomical targets for electrical stimulation. Out of this successful collaboration came positive results with genuine potential for benefit in patients. Specifically, the medial knee joint reaction force, the size of which is known to predict the risk of onset and progression of arthritis, was significantly reduced with our novel application of FES. Early discussions have taken place with Imperial Innovations, the university's technology transfer department, and a team there is currently assessing potential means of commercialisation of my work. If the technology is to be harnessed in the clinical domain, further work is required to validate its utility and confirm its safety. I am eager to take this forward, and I feel that the PhD programme is the best way to do so.
Employing advanced in-house musculoskeletal modeling technology, and alongside experts in this relatively new field, I was able to leverage my clinical knowledge, drawing on my experience to suggest optimum anatomical targets for electrical stimulation. Out of this successful collaboration came positive results with genuine potential for benefit in patients. Specifically, the medial knee joint reaction force, the size of which is known to predict the risk of onset and progression of arthritis, was significantly reduced with our novel application of FES. Early discussions have taken place with Imperial Innovations, the university's technology transfer department, and a team there is currently assessing potential means of commercialisation of my work. If the technology is to be harnessed in the clinical domain, further work is required to validate its utility and confirm its safety. I am eager to take this forward, and I feel that the PhD programme is the best way to do so.
Organisations
People |
ORCID iD |
| Anthony Bull (Primary Supervisor) | |
| Lance Rane (Student) |
Publications
Rane L
(2019)
Deep Learning for Musculoskeletal Force Prediction.
in Annals of biomedical engineering
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509486/1 | 01/10/2016 | 31/03/2022 | |||
| 1855334 | Studentship | EP/N509486/1 | 01/10/2016 | 31/03/2020 | Lance Rane |
| Description | 1. Development of a neural network that predicts muscle and joint forces given input data on a patient's movement. This work was performed to address a shortcoming in existing methods allowing inference of musculoskeletal forces during movement - they are too slow to be used in real time. The neural network is fast enough that it can be used in real time, for example in order to provide biofeedback to a patient as they walk. This work has been published in peer-review journal. 2. Development of advanced computational techniques that allow efficient modeling of a patient's motor control in simulation. This work was performed as part of a visiting scholarship at Stanford University. It forms the precursor for work aimed at optimising a patient's movement so as to improve disease outcomes in, for example, osteoarthritis, which is ongoing. Part of this work was presented at the Neurips 2018 conference, and is due to be published in an upcoming book. 3. Adaptation of the techniques developed in (2) to create a computational framework for the optimisation of a lower limb prosthesis in simulation. Computational prototyping of prostheses is attractive because it is relatively cheap and quick. One of the difficulties is in modelling the subject's adaptation to the prosthesis - this framework has been developed to address this. 4. Validation of the work described in (2). Previously developed techniques were used to develop personalised models of motor control for a number of healthy subjects. These models were then interrogated in simulation to derive optimal physical retraining patterns for joint loading. The subjects were then trained in reality to match these optimal targets. Most were able to do so, and achieved reductions in their joint loads. |
| Exploitation Route | Once validated, these methods could form the basis for computational tools that: 1. Provide personalised rehabilitation advice to a patient or athlete describing how to change their movement pattern/ which muscles to strengthen in order to achieve a given movement goal eg reduce their risk of severe osteoarthritis/ run more quickly. 2. Allow the properties of a lower limb prosthesis (mass, stiffness etc) to be personalised to an individual amputee. |
| Sectors | Healthcare,Leisure Activities, including Sports, Recreation and Tourism |
| URL | https://www.ncbi.nlm.nih.gov/pubmed/30599054 |
| Description | Following the publication of work detailed in Key Finding (1) we have had discussions with senior coaches at the English Institute of Sport who are exploring the possibility of the incorporation of these techniques into their athlete rehabilitation programme. |
| First Year Of Impact | 2019 |
| Sector | Leisure Activities, including Sports, Recreation and Tourism |
| Impact Types | Economic |