Background

AI adoption in healthcare is dominated by efficiency metrics, which is inadequate for complex clinical practice. The true value of AI lies in its capacity to serve as a Cognitive Lever – multiplying the clinician’s intellectual force rather than simply their throughput.

Aims

Using the “Cognitive Lever” framework, this review evaluates how AI augments physiotherapists’ capabilities in kinematicanalysis, prognostication, and personalised care.

Material and methods

PubMed, Cochrane Library, arXiv, and IEEE Xplore were searched (January 2015–November 2025). Of 112 records identified, 42 met the inclusion criteria. Studies were categorised as automation or augmentation tools; synthesis focused on the latter.

Results

The 42 included studies demonstrated a clear transition from basic automation to clinical augmentation. AI serves as a cognitive lever across multiple applications: markerless motion capture achieves a mean absolute error under 3° for joint angles; deep learning detects prodromal neurological signs with high sensitivity; machine learning predicts functional recovery with 85–90% accuracy; and generative AI converts biomechanical goals into hyper-personalised home exercise programmes.

Conclusions

Viewing AI as a cognitive lever transforms the subjective clinical eye into objective data. These tools do not replace healing hands – they clear cognitive clutter. The measure of this technology is not in minutes saved, but in movement restored.