A Health Revolution? AI and Musculoskeletal Practice

In this week’s Clinical Insight, we hear some reflections from one of the team who recently attended IFOMPT 2024. This team member is not an artificial Intelligence (AI) expert but after hearing a number of presentations, they felt it really worthwhile to reflect on the learning achieved. We thought we would share it.

AI was a key feature of a number of presentations at the very recent IFOMPT 2024 conference in Basel, Switzerland. AI is so rapid in its advancement that by the time it is researched, the findings are seemingly redundant, which is such a tough concept to appreciate.

Benefits of AI in MSK

AI has been cited as offering good outcomes for MSK diagnosis, self-management, signposting and even exercise prescription and the added value here is that it does not require a workforce. It never gets tired or cancels an appointment, it does not need annual leave and is available 24/7. This technology means the care given is the most contemporary available, and offers up to date, evidence- based practice.

This is quite a sobering thought and one that I am sure many of the clinicians in the lecture theatres had to really consider their bias when taking this in.

Opportunity, not threat

The key here is opportunity and not threat. It is easy to look at where AI might not be able to offer what say a physiotherapist does but there is much it can. Much of diagnosis in many pathways and reasoning is about cognitive algorithms, probability, safety indexes and pattern recognition. This intelligence was presented as having this capability and therefore offering a dramatic change in workforce, patient experience and service design.

So, in the future what physiotherapy might see is less of a need for the basics in some of our philosophical constructs for diagnosis for say structure and pathology.

Self management could be well developed for the vast majority of MSK patients and beyond and this likely will be more intuitive and far further reaching than the current Apps for rehab that are utilised widely.

So, might we ask how does that leave some of the standard work in physiotherapy? Well, would this lead to an end to the orthopaedic triage model, which is assess, diagnose, investigate and sign post?

Perhaps yes. Will primary care be affected in a similar way? Possibly.

Where does this leave rehab?

Where might rehabilitation then be left, as the role of the historical “diagnosis” physiotherapist may become less needed?

This may very well lead us into areas that have been relatively less focussed on with advancing practice in say orthopaedics, which is the nuances of tailored, psychologically supported, narrative informed, and emotionally under-pinned collaborative rehabilitation.

Bedrock of Emotional Connection

Maybe AI will be able to do this, but at this stage from the experiences at this conference, I could see the emotional connection built on a humanistic paradigm becoming the bedrock of how as clinicians we serve our community.

Moving away from a patient model, physiotherapy could become a community-based service that is integrated across all ages, outside the hospital environment that in many cases is an algorithm-based model to ensure safe, acute care and signposting.

Redefining our Place in Society

If physiotherapy is not needed here, then to be considered differently means the profession will need to re-define its place in society and not health.

Health being one aspect of how we will need to impact, but it then will emphasise the need to supporting working, social environments, social care and relationships. These are skilled areas to navigate and are so complex that algorithms (for now) might (I may be wrong) not be able to work through as effectively. Again, we might find this out at a later date.

Essentially, the core elements of rehabilitation, and how that offers a societal benefit across all walks of life, may be the place where physiotherapy steps next as we embrace AI leading to health revolution in all aspects, and in physiotherapy one we have yet to have ever experienced.

References

Griefahn, A., Zalpour, C. and Luedtke, K., 2024. Identifying the risk of exercises, recommended by an artificial intelligence for patients with musculoskeletal disorders. Scientific Reports, 14(1), p.14472.

Tack, C., 2019. Artificial intelligence and machine learning| applications in musculoskeletal physiotherapy. Musculoskeletal Science and Practice, 39, pp.164-169.