Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/61008
Title: A synergistic physio-neuro rehabilitation platform to accelerate recovery of hand function after stroke
Authors: Banerji Subhasis
Keywords: DRNTU::Science::Medicine::Biomedical engineering
DRNTU::Engineering::Mechanical engineering::Assistive technology
DRNTU::Engineering::Mechanical engineering::Bio-mechatronics
Issue Date: 2014
Source: Banerji Subhasis. (2014). A synergistic physio-neuro rehabilitation platform to accelerate recovery of hand function after stroke. Doctoral thesis, Nanyang Technological University, Singapore.
Abstract: More people are surviving stroke than ever before due to advances in emergency care. However, the numbers of stroke patients that have not fully regained their ability to perform activities of daily living (ADL) have also increased. As the stroke survivor population grows, healthcare institutions and families are finding it increasingly difficult to restore disabled patients to independent living and provide adequate treatment and rehabilitation time. Recovery of functional use of the hand is a key factor in regaining independence post-stroke. Engineering and robotics solutions proposed for regaining hand function have so far not had the widespread success that was expected. The ability to reach out to the majority of stroke patients using robotic rehabilitation solutions is also not viable due to high cost and infrastructure requirements. In addition, a major reason, as discovered in this work, has been the traditional pre-occupation with the disability and augmentation of lost function and insufficient research into utilizing the patient’s residual abilities and innate, natural learning mechanisms during the course to recovery. This PhD research proposes a paradigm shift in rehabilitation strategy. It envisions a learning model and technology solution which would harness and leverage the intuitive learning abilities that exist in human beings as an integral part of therapy. Bio-signals from the muscles and brain are used as a universal language to assess and train patients to re-learn key mind-body strategies which enable accelerated improvement in functional abilities. Such strategies include self-regulating the musculoskeletal and brain responses to the demands of motor activities, which is a must before proceeding to extended repetitive practice. The human learning model which drives the strategy of SYNergistic PHysio-NEuro rehabilitation platform or “SynPhNe”, is explained in this report along with innovative hardware and software architectures. The assessment and rehabilitation capabilities of this platform are described, with a focus on hand function recovery. Technical feasibility studies with healthy and stroke patients were conducted as an initial stage proof-of-concept, which included observing patient responses to a video imitation and biofeedback routine over a single session, using a Phase I prototype. Short term improvements in function and ability to activate and inhibit muscles were seen. Subsequently, a clinical feasibility trial was conducted with a more wearable, user friendly Phase II clinical prototype. This was clinically supervised at the largest public rehabilitation centre in Singapore. The subjects were long term, “plateaued” chronic patients who underwent therapy over four weeks, along with pre and post clinical assessment. The data for five subjects is presented and specific case studies discussed to illustrate how patients with different challenges may adopt different pathways to function recovery. Three hypotheses were tested: • Biofeedback integrated with mirror image video-based imitation (as a feedforward) supports accelerated re-learning of hand function after stroke. • It is technologically feasible to deliver this form of training via a safe, low cost, wearable, automated device, particularly to stroke patients whose recovery had “plateaued” with little hope for further improvement. • It is feasible to track obvious and non-obvious (sub-clinical) improvements using novel, bio-signal based real-time brain and muscle metrics which can help to personalize therapy and make it more meaningful to patients. The data showed that the ability to successfully perform repetitive actions was as much associated with the ability to volitionally relax muscles as it was with the ability to contract muscles. Those who could achieve such relaxation-activation balance for both the muscles and the brain achieved the best results. The SynPhNe training resulted in functional improvement as per clinical scales, better manipulation of objects, lower self-reported pain, and better intra-limb co-ordination. More severely affected subjects seemed to progress proportionately faster as compared to less impaired subjects. Subjects with lower impairments, however, improved functional status faster over 12 sessions as expected although their recovery had reached a “plateau” prior to this trial. Incremental changes in performance that could not be captured by typical clinical scales were successfully displayed by the bio-signal parameters being tracked. This study showed that patients unconsciously train hitherto unseen, maladaptive brain-muscle reactions repetitively along with desired actions and tasks, thus making brain plasticity a double edged sword. If patients are made aware of these unconscious, non-obvious reactions, it is possible for them to self-correct and inhibit these reactions to overcome recovery “plateaus”. This is an important adjunct to standard care and potentially applicable to sub-acute and chronic stroke patients, including those with co-morbidities such as ataxia, sensory deficit, attention difficulties, high muscle tone, intra-limb muscle coupling and pain.
URI: http://hdl.handle.net/10356/61008
Schools: School of Mechanical and Aerospace Engineering 
Research Centres: Robotics Research Centre 
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:MAE Theses

Files in This Item:
File Description SizeFormat 
Banerji+Subhasis+Consolidated+thesis+manuscript+(Online+ver).pdf
  Restricted Access
6.43 MBAdobe PDFView/Open

Page view(s) 50

600
Updated on Mar 28, 2024

Download(s)

8
Updated on Mar 28, 2024

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.