Large language models in physical therapy: time to adapt and adept

Area of Science:

• Healthcare technology
• Physical therapy innovation
• Artificial intelligence in medicine

Background:

• Healthcare is undergoing significant transformation driven by AI and ML.
• Physical therapists (PTs) are poised for a paradigm shift in their field.
• AI presents opportunities rather than threats to PT education, practice, and research.

Purpose of the Study:

• To examine the role of large language models (LLMs) in physical therapy.
• To explore the potential benefits and challenges of AI integration for PTs.
• To advocate for PTs' active engagement in shaping AI development and ethical use.

Main Methods:

• Review of current AI and ML capabilities, specifically LLMs like ChatGPT and BioMedLM.
• Analysis of the application of LLMs in physical therapy and rehabilitation contexts.
• Discussion of challenges related to data accuracy and bias in AI models for PT.

Main Results:

• LLMs can offer human-like performance but face accuracy challenges due to vast datasets.
• Potential benefits for PTs include streamlined administrative tasks, global connectivity, and customized treatments.
• Human touch and creativity remain indispensable alongside AI integration.

Conclusions:

• PTs should actively learn about and shape AI models for ethical use and human supervision.
• Integrating AI requires addressing data accuracy and model-feeding challenges.
• A collaborative future where AI enriches the PT field is achievable with careful consideration and proactive engagement.