Innovative new app created to help stroke patients with rehabilitation

A UWE Bristol researcher led a project to develop an innovative new app to help patients recover after having a stroke.

Dr Praveen Kumar worked with partners to build an AI-powered prototype app that generates personalised exercise programmes using a library of rehabilitation videos.

The app – designed for patients to use in community settings – can produce tailored exercise programmes based on individual needs, abilities and recovery goals.

Funded by the National Institute for Health and Care Research (NIHR), the project saw Dr Kumar and other academics collaborate with clinicians, technology experts and stroke survivors to create the app.

Dr Kumar, an Associate Professor in Stroke Rehabilitation, said: “The successful development of this prototype app demonstrates the value of combining patient involvement, clinical expertise and technological innovation to address an important unmet need in stroke rehabilitation.

“Our vision is to improve access to personalised rehabilitation support and empower stroke survivors to continue their recovery journey within their own communities.”

Stroke affects 118,000 people in the UK annually, impacting quality of life and NHS resources. Rehabilitation exercises improve walking, fitness, daily activities and prevent future stroke, with clinical guidelines recommending at least three hours of rehabilitation per day and staying active for at least six hours daily.

The app – for use on tablets and computers – was aimed at helping people choose correct exercises, progress with them and self-manage their rehabilitation at home. The videos are existing ones of real stroke patients undertaking exercises under the guidance of physiotherapists.

The development of the app builds on earlier work carried out by Dr Kumar and fellow researchers to deliver group exercises classes in the community for 10 years, in collaboration with the Bristol After Stroke charity. Based on participant feedback, bitesize exercise videos were developed. People recovering from stroke reported that these videos were engaging, and helped their movement and independence but also noted that they felt stuck not knowing how to progress with the exercises.

Integrating AI into portable devices was suggested as an idea to assist people, by offering real-time input and improved engagement with rehabilitation at home.

The project was awarded a grant as part of NIHR funding of more than £3 million to 34 new cutting-edge medical technologies to tackle brain injury in 2025. The projects were funded through the Invention for Innovation (i4i) Funding At the Speed of Translation (FAST) 4 Awards to support the development of innovative healthcare technologies for acquired brain injury (ABI), including the prevention, diagnosis and management of ABI as well as recovery from ABI as a consequence of traumatic brain injury, stroke, tumours and infections.

Dr Kumar was the chief investigator on the project, working alongside co-investigators Dr Michael Loizou and Dr Ali Al-Nuaimi (University of Plymouth), Professor Kenton O’Hara and Dr James Pope (University of Bristol) and Dr Maria Galvez Trigo (Cardiff University). Co-researchers on the study were Regan Shakya, Anika Pinto, Sarah Koerner and Thrisha Rajkumar.

The team worked with several partners on the development of the app. These were Claire Angell and Sarah Sparks (Patient Partners), Amrik Singh (Clinical expert, the Brightwell Centre), Zehra Haq (Community Ethnic Partner) and Rebecca Sheehy, (CEO, Bristol After Stroke).

Dr Kumar said: “In the app, AI can be used to tailor the types of YouTube rehabilitation exercise videos that patients use. Rather than viewing longer videos (20 minutes in duration) they can try shorter versions of the videos, and AI can be used to guide this process. This saves time for clinicians and therapists, and is also better for patients.

“In future through collaboration with industry, the app has the potential to allow clinicians to monitor their patients’ rehabilitation. Their exercises can be recorded and fed back to the clinician so they can see the range of movement that the patient has been experiencing. This provides a complete loop between patient and clinician.”

Over the next 17 months, with support from the UWE Bristol Strategic Research & Impact Programme (SRIP), Dr Kumar and the team will evaluate and cross-validate the app with people with stroke in real-world community settings. The project will explore whether AI-generated personalised exercise programmes align with clinical judgement and are usable, acceptable and trustworthy, as well as how they can support motivation and adherence to rehabilitation at home.