Tiffany Harmanian, BS
Medical Student
ATSU SOMA
Phoenix, Arizona, United States
Chloe Jensen, BS
Medical Student
ATSU SOMA
Phoenix, Arizona, United States
Derek Higgins, DO
Dr. (DO)
A.T. Still University School of Medicine in Arizona
Mesa, Arizona, United States
The integration of Artificial Intelligence (AI) in healthcare has seen a dramatic rise in recent years. Many medical institutions, electronic medical record (EMR) systems, and administrative practices have adopted AI to enhance efficiency and scientific accuracy. However, resistance remains when it comes to utilizing AI recommendations for clinical treatment. While it is widely acknowledged that a computer cannot replicate the nuanced thinking of a physician, AI offers valuable potential when positioned as a supportive tool rather than a definitive decision-maker. This study aims to explore the accuracy of AI in managing complex Physical Medicine and Rehabilitation (PM&R) cases.
Design:
To do so, we conducted a comparative analysis between AI-suggested treatments and those recommended by physicians. A three-point grading system was employed to evaluate the correlation between AI and physician treatment plans: (1) AI treatment was unrelated to the physician’s recommendation; (2) AI treatment shared some components with the physician’s plan but was not identical; (3) AI treatment fully aligned with the physician’s recommendation.
Results:
Upon review, the data revealed that while 40% of AI-generated treatments bore no relation to physician recommendations, the majority of cases (60%) demonstrated overlap, with AI either matching or partially aligning with the physician's treatment plan.
Conclusions:
The findings of this study suggest that AI holds substantial promise as a supportive tool in clinical decision-making, particularly in the field of PM&R. However, AI continues to face limitations when applied to complex, multidimensional medical cases where human expertise and nuanced decision-making remain critical. While AI can enhance efficiency and support evidence-based recommendations, the study highlights the need for further advancements in AI algorithms to better understand the intricacies of patient care. Ultimately, the research emphasizes AI’s role as a supplementary tool rather than a replacement for human clinicians, suggesting future collaboration between AI and healthcare providers for improved patient outcomes.
