Jump to content

AI in Physiotherapy

Original Editor - Lucinda hampton

Top Contributors - Lucinda hampton, Alexandra Stead and Vidya Acharya  

AI in Physiotherapy: an Overview of Current practices

The Current Landscape

Physical disabilities and demands on physiotherapy services are becoming increasingly common with our aging population, creating unprecedented demand for rehabilitation interventions. However, poor availability and accessibility of traditional rehabilitation services limits clinical impact, making it harder for patients to maintain independence and access the care they need. This, coupled with admin-heavy tasks taking up valuable available time, has sparked interest in AI-supported solutions that could revolutionise how we deliver physiotherapy care.

AI's Promise for Enhanced Patient Care

AI systems are transforming physiotherapy by offering unprecedented precision in note taking, as well as patient assessment and treatment planning. Through multivariable prediction models, AI can analyse complex patient data to predict long-term outcomes, particularly valuable for conditions like hip fractures, though these have not yet been found to be superior models to more traditional methods.[1] Machine learning algorithms can create tailored rehabilitation programs by analysing medical records and treatment responses, leading to improved patient compliance and more effective clinical decision-making.[2] The continuous feedback loops built into these systems allow treatment plans to adapt based on patient progress, strengthening the therapeutic relationship and enhancing satisfaction.[3]

Current AI Applications in Practice

Several AI tools are already making their mark in physiotherapy. Gait analysis systems detect subtle movement patterns that might indicate neurological conditions like Parkinson's or cerebral palsy.[4] Natural language processing helps analyse patient reports more efficiently,[2] while specialized tools like physiopedia PAI, physioGPT and PostureFix assist with specific assessment and treatment tasks. There is increasing momentum towards healthcare professionals (HCP) using AI documentation assistance, which has been found to help standardise, streamline and improve quality of notes, as well as reduce HCP burnout and improve job satisfaction.[5]However, caution is still recommended with utilisation of these tools, due to reported limitations such as accuracy of interpretation, misinterpretation, generation of false assessment positives or irrelevant documentation, suggesting human input is still necessary.[5] Nonetheless, AI is becoming more embedded into healthcare practice, with the hope to improve efficiency and quality of care for patients.

Current AI tools span five main categories:

  • App-based systems - deliver exercise programs, patient education and admin-assistance.
  • Robotic assistive devices - like powered wheelchairs and prosthetics compensate for lost function.
  • Robotic rehabilitation - devices including gait trainers and upper limb therapy robots facilitate recovery through active training.
  • Virtual reality gaming - creates engaging environments for motor learning and functional training.
  • Wearable sensors - provide real-time movement analysis and biofeedback for exercise correction and progress monitoring.

Global AI Research in Physiotherapy: What You Need to Know

A comprehensive analysis of 340 studies from 2006-2024 reveals that AI research in physiotherapy has exploded, particularly between 2019-2022, driven by advances in machine learning, wearable technology, and robotic rehabilitation. While North America, Europe, and Asia lead this research revolution, significant geographic disparities exist, highlighting opportunities for global collaboration. The evidence shows AI's clear potential to enhance rehabilitation outcomes, improve operational efficiency, and deliver innovative treatment solutions. However, as physiotherapists embrace these technologies, critical gaps remain in pediatric rehabilitation and AI-based decision-making systems. Moving forward, the profession must prioritize ethical considerations including data privacy, algorithmic bias, and system explainability while focusing on long-term clinical impact studies. This research boom signals that AI integration in physiotherapy is no longer a future possibility but a current reality requiring immediate attention to training, implementation, and equitable access across all regions and patient populations.[6]

Despite the growing interest, clinical evidence remains limited. A systematic review of 9,054 articles identified only 5 randomized controlled trials testing AI-supported rehabilitation technology. While these studies showed improvements in physical function, activity levels, pain management, and quality of life, clinical effects were inconsistent across different applications. This highlights a significant research gap requiring more robust real-world clinical evaluations.[7]

Physiotherapy Receptiveness and Readiness for AI Integration

AI in Physiotherapy Education

Physiotherapy education is struggling to keep up with AI advancements and integration, particularly compared to other specialisms. A recent literature search revealed only one article specifically examining AI in physiotherapy education, while medical education produced over 6,000 relevant articles and dental education yielded 241 results.[8] However, this paper, published in early 2024, shows how quickly research within this field is changing, with many more articles published within the physiotherapy field since .[9] [10][11] Whilst understandably cautious, recent research findings show that HCPs are becoming increasingly favourable to inclusion of AI within education, suggesting either greater willingness to use and include AI, or general acceptance of the inevitable integration into practice.

A recent randomised controlled trial found that learning and digital self-efficacy can be enhanced by use of generative AI through optimising active reflection and self directed learning, without compromising behaviour risks in students.[11] Moreover, a qualitative study found students were receptive to generative AI use to facilitate learning and clinical reasoning, though were mindful of the potential risks this could cause, such as knowledge gaps and impact on learning quality.[9] Similarly, whilst surveyed Canadian healthcare students in 2022 had some scepticism towards AI use, 74.5% felt positively towards AI in healthcare and agreed it needed to be included in educational curriculum.[12]

Whilst student acceptance is important, it is equally essential for academics to be aware of implications of AI use, understanding the potential benefits and risks in healthcare and need to be prepared to adopt educational strategies for how AI can be used in the future of healthcare.[10] The gap in AI education preparation may have serious implications for future physiotherapists who will need to integrate AI tools for diagnosis, treatment planning, and patient monitoring in future roles. Furthermore, to bridge any divide, physiotherapy schools should embed AI into the curriculum to enhance student awareness of AI and encourage innovative practices for how it can shape healthcare.[13] This may be achieved through AI specialised courses, collaboration with AI experts, and faculty development opportunities. Students also need hands-on experience with AI applications in patient data analysis, treatment recommendations, and clinical simulations to prepare for an AI-integrated healthcare future.

Global Physiotherapists Readiness

Acceptance and readiness of integration of AI varies around the world. In Saudi Arabia, physiotherapists appear favourably towards AI integration into practice, claiming potential boost to productivity, improved patient outcomes, and reduced workload.[14] However, the physiotherapists reported barriers to AI use including insufficient training, limited experience, and knowledge gaps, with solutions to these involving integrating AI into curricula, organizing practical workshops, and developing institutional support frameworks.[14]

AI Implementation Barriers

Barriers to overcome:

  • Technology literacy challenges as users struggle with complex interfaces
  • Reliability issues where technical failures disrupt treatment continuity
  • User fatigue as patients lose engagement over time

Enablers supporting adoption:

  • Improved access allowing patients to access rehabilitation programs remotely
  • Remote monitoring enabling continuous progress tracking without clinic visits
  • Reduced staffing needs requiring less manpower for basic interventions
  • Cost savings through lower overall treatment costs[15]

Ethical Considerations

The integration of AI brings important ethical responsibilities that physiotherapists must navigate carefully. Patient data confidentiality remains paramount, as AI systems require extensive datasets for training and operation. There's also the crucial question of maintaining human empathy and emotional support in therapeutic relationships – areas where AI cannot replace the human touch that defines quality physiotherapy care. Clear regulatory frameworks and guidelines are essential to protect patient rights while leveraging AI's benefits for improved clinical outcomes.[7]

Moving Forward Responsibly

The incorporation of AI into physiotherapy represents more than just technological advancement – it's a cultural shift toward data-driven, personalized care. While current research shows promise in improving treatment pathways and diagnostic accuracy, this emerging field is still in early stages. Success lies not in replacing human expertise but in combining AI's analytical power with physiotherapists' clinical reasoning and compassionate care.

Developers must strive to conduct robust clinical evaluations in real-world settings and appraise post-implementation experiences. For physiotherapists, the focus should be on addressing user barriers while leveraging enablers to expand rehabilitation access and reduce costs. This balanced approach ensures that technological progress enhances rather than diminishes the foundational values of physiotherapy practice.

The future potential is promising, but evidence-based implementation remains key to realizing AI's benefits in physiotherapy practice[7].

References

  1. ↑ Lex JR, Di Michele J, Koucheki R, Pincus D, Whyne C, Ravi B. Artificial Intelligence for Hip Fracture Detection and Outcome Prediction: A Systematic Review and Meta-analysis. JAMA Netw Open. 2023 Mar 1;6(3):e233391. doi: 10.1001/jamanetworkopen.2023.3391. PMID: 36930153; PMCID: PMC10024206.
  2. ↑ 2.0 2.1 Rasa AR. Artificial Intelligence and Its Revolutionary Role in Physical and Mental Rehabilitation: A Review of Recent Advancements. Biomed Res Int. 2024 Dec 17;2024:9554590. doi: 10.1155/bmri/9554590. PMID: 39720127; PMCID: PMC11668540.
  3. ↑ Attoh-Mensah E, Boujut A, Desmons M, Perrochon A. Artificial intelligence in personalized rehabilitation: current applications and a SWOT analysis. Front Digit Health. 2025 Jul 24;7:1606088. doi: 10.3389/fdgth.2025.1606088. PMID: 40778384; PMCID: PMC12328449.
  4. ↑ Galna B, Barry G, Jackson D, Mhiripiri D, Olivier P, Rochester L. Accuracy of the Microsoft Kinect sensor for measuring movement in people with Parkinson's disease. Gait Posture. 2014 Apr;39(4):1062-8. doi: 10.1016/j.gaitpost.2014.01.008. Epub 2014 Jan 22. PMID: 24560691.
  5. ↑ 5.0 5.1 Bongurala AR, Save D, Virmani A, Kashyap R. Transforming Health Care With Artificial Intelligence: Redefining Medical Documentation. Mayo Clin Proc Digit Health. 2024 May 22;2(3):342-347. doi: 10.1016/j.mcpdig.2024.05.006. PMID: 40206119; PMCID: PMC11975979.
  6. ↑ Hanafi B, Hasan A, Ahmad A, Ali M. Advancing public health through artificial intelligence in physiotherapy: a bibliometric analysis. Informatics and Health. 2025 Sep 1;2(2):99-118.Available:https://www.sciencedirect.com/science/article/pii/S2949953425000165#ab0010 (accessed 13.11.2025)
  7. ↑ 7.0 7.1 7.2 Nambi, Gopal1; Alghadier, Mshari2; Mohamed, Shahul Hameed Pakkir3; Aldhafian, Osama Rashed4; Alshahrani, Naif A.5; Albarakati, Alaa Jameel A.6. Clinical Usefulness of Artificial Intelligence in Physiotherapy – A Practice-based Review. SBV Journal of Basic, Clinical and Applied Health Science 7(4):p 184-188, Oct–Dec 2024. | DOI: 10.4103/SBVJ.SBVJ_38_24 Available: https://journals.lww.com/sbvj/fulltext/2024/10000/clinical_usefulness_of_artificial_intelligence_in.8.aspx(accessed 8.11.2025)
  8. ↑ Veras M, Dyer JO, Kairy D. Artificial Intelligence and Digital Divide in Physiotherapy Education. Cureus. 2024 Jan 20;16(1):e52617. doi: 10.7759/cureus.52617. PMID: 38374829; PMCID: PMC10875905.Available:https://pmc.ncbi.nlm.nih.gov/articles/PMC10875905/ (accessed 12.11.2025)
  9. ↑ 9.0 9.1 Lindbäck, Y., Schröder, K., Engström, T. et al. Generative artificial intelligence in physiotherapy education: great potential amidst challenges- a qualitative interview study. BMC Med Educ 25, 603 (2025). https://doi.org/10.1186/s12909-025-07106-w
  10. ↑ 10.0 10.1 Lowe, S.W. The role of artificial intelligence in Physical Therapy education. Bull Fac Phys Ther 29, 13 (2024). https://doi.org/10.1186/s43161-024-00177-8
  11. ↑ 11.0 11.1 Ergezen Sahin, G., Aras Bayram, G., Sanchez Sierra, A. et al. Effects of artificial intelligence based physiotherapy educational approach in developing clinical reasoning skills: a randomized controlled trial. BMC Med Educ 25, 1378 (2025). https://doi.org/10.1186/s12909-025-07926-w
  12. ↑ Teng M, Singla R, Yau O, Lamoureux D, Gupta A, Hu Z, Hu R, Aissiou A, Eaton S, Hamm C, Hu S, Kelly D, MacMillan KM, Malik S, Mazzoli V, Teng YW, Laricheva M, Jarus T, Field TS. Health Care Students' Perspectives on Artificial Intelligence: Countrywide Survey in Canada. JMIR Med Educ. 2022 Jan 31;8(1):e33390. doi: 10.2196/33390. PMID: 35099397; PMCID: PMC8845000.
  13. ↑ Issa, W.B., Shorbagi, A., Al-Sharman, A. et al. Shaping the future: perspectives on the Integration of Artificial Intelligence in health profession education: a multi-country survey. BMC Med Educ 24, 1166 (2024). https://doi.org/10.1186/s12909-024-06076-9
  14. ↑ 14.0 14.1 Aldhahi MI, Alorainy AI, Abuzaid MM, Gareeballah A, Alsubaie NF, Alshamary AS, Hamd ZY. Adoption of Artificial Intelligence in Rehabilitation: Perceptions, Knowledge, and Challenges Among Healthcare Providers. Healthcare (Basel). 2025 Feb 7;13(4):350. doi: 10.3390/healthcare13040350. PMID: 39997225; PMCID: PMC11855079.(accessed 13.11.2025)
  15. ↑ Sumner J, Lim HW, Chong LS, Bundele A, Mukhopadhyay A, Kayambu G. Artificial intelligence in physical rehabilitation: A systematic review. Artificial Intelligence in Medicine. 2023 Dec 1;146:102693.Available: https://www.sciencedirect.com/science/article/pii/S0933365723002075 (accessed 8.11.2025)
Retrieved from "https://www.physio-pedia.com/w/index.php?title=AI_in_Physiotherapy&oldid=373638"