KITE Research Spotlight: Investigating the effectiveness of AI driven virtual rehab at home

The results of this paper will ultimately help improve accessibility to care for those living with the effects of disability, injury and aging.

TORONTO–New research from the KITE Research Institute at UHN determined that further investigation is needed to assess the effectiveness of artificial intelligence (AI) driven virtual rehabilitation in patient’s homes. 

A scoping review led by KITE scientist Dr. Shehroz Khan found that while there is a growing body of literature examining the use of AI driven virtual rehab there aren’t a lot of studies that discuss how effective it is when faced with the various technological and other unique challenges of domestic settings. 

Dr. Khan’s research focuses on the development of machine learning and deep learning algorithms within the realms of aging, rehabilitation, and intelligent assistive living.

The results of this review, which were published in Nature’s npj Digital Medicine, will spur further research into this area and ultimately help increase accessibility to care. 

The KITE Research Institute connected with Dr. Khan to learn more about his findings.

 Which patient groups are most affected by this? 

Our scoping review identified that AI-driven virtual rehabilitation is mostly used by stroke, cardiac, neuro, and physiotherapy rehabilitation patients.

What did you find?

The reviewed studies used a range of AI algorithms, including fuzzy rule-based methods and deep neural networks, to analyze patient data collected from different sensors and build predictive models. These models would then make inferences about patients' rehabilitation outcomes, especially assessing the quality of exercises and providing feedback to both patients and clinicians. Despite increasing research on AI in virtual rehabilitation, its application in real-world settings for people living in the community remains minimally explored.

Why does this matter?

Virtual rehabilitation provides similar health outcomes to in-person rehabilitation and can overcome barriers to participating in and completing rehabilitation programs, including transportation, financial issues and the shortage of staff. However, to include AI-driven virtual rehabilitation programs in regular clinical practice, it is important to evaluate and compare their effectiveness, while taking into account their specific challenges and using standardized metrics.

What is the potential impact? 

This review accentuates the need for further research on AI-driven virtual rehabilitation platforms for people living in the community. The specific focus areas include overcoming infrastructure barriers, co-designing with stakeholders, standardizing usability and acceptability metrics, addressing privacy ethical and legal issues, and emphasizing personalization to improve quality of care.



Research Spotlight: 
 AI scoping review

Assistant Professor at the Institute of Biomedical Engineering University of Toronto

Name of Publication:
Artificial Intelligence-driven Virtual Rehabilitation for People Living in the Community: a Scoping Review

Name of Journal:
npj Digital Medicine