AI chatbot matches orthopedic surgeons on lateral malleolus fracture surgery decisions

An AI chatbot showed diagnostic accuracy comparable to three fellowship-trained orthopedic trauma surgeons for determining surgery in isolated lateral malleolar fractures. AUC values were similar for the AI and all surgeons (0.63–0.67). All raters, including the AI, recommended…

• Inter-rater agreement between the AI and surgeons was only slight, while inter-surgeon agreement was moderate (κ = 0.457–0.589), indicating the AI used a systematically different measurement strategy.
• The AI's radiographic measurements of medial clear space, tibiofibular clear space, and tibiofibular overlap differed systematically from all surgeons across five of six parameters.
• This was a single-center retrospective study of 50 patients with AO/OTA 44B1 fractures, without external or prospective validation.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.