In the UK approximately 100,000 people have strokes each year (stroke strikes once every five minutes on average) and there are an estimated 1.2 million stroke survivors in the UK. Rehabilitation processes to enable motor recovery of the body, including functional rehabilitation therapies and exercises, are an essential part of patient care. These must be combined with well-coordinated multidisciplinary stroke rehabilitation provision, including via early support discharge teams. Efficient communication to evaluate degree of impairment and to enable bespoke individualisation of rehabilitation therapies, with coordination of monitoring and feedback between clinicians, stroke patients and their families, is essential to achieve best post-stroke rehabilitation success.
The aim of this project is to develop an effective post-stroke rehabilitation platform which provides continuous and real-time assistance for both stroke survivors as well as their healthcare providers. This rehabilitation platform will consist of a set of wearable devices using artificial intelligence (AI) algorithms to monitor patient’s rehabilitation process and provide real-time assessment and feedback. Unlike cumbersome traditional video-based activity monitoring systems, our solution uses non-invasive wearable sensing technology in conjunction with state-of-the-art AI algorithms to both recognise and analyse post-stroke patient’s physical activities. It will provide useful feedback to patients and clinicians: an accurate and quantified assessment of the exercises which permits secure monitoring the progress and safeguard patient privacy. The project team has already started the early research work with the nationally top-rated Stroke Unit at Colchester General Hospital. We will continually collaborate with the clinicians, physiotherapists, and engineers for the proposed study.
- 10/2021: Our paper “Wearable sensors and machine learning in post-stroke rehabilitation assessment: A systematic review” has been accepted in Biomedical Signal Processing and Control.
- 09/2021: Our clinical trial request has been accepted and supported by East Suffolk and North Essex NHS Foundation Trust.
- 07/2021: Our paper “A comprehensive evaluation of state-of-the-art time-series deep learning models for activity-recognition in post-stroke rehabilitation assessment” has been accepted in the 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society.
- 12/2020: Our paper “Improving the activity recognition using GMAF and transfer learning in post-stroke rehabilitation assessment” has been accepted in the IEEE 19th World Symposium on Applied Machine Intelligence and Informatics.