Related Experiment Video
Updated: Jul 23, 2025
07:45
The Use of Mixed Reality in Custom-Made Revision Hip Arthroplasty: A First Case Report
Published on: August 4, 2022
3.4K
Artificial Intelligence in Orthopaedic Surgery: Can a Large Language Model "Write" a Believable Orthopaedic Journal
Devon T Brameier1, Ahmad A Alnasser2, Jonathan M Carnino3
1Department of Orthopaedic Surgery, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts.
The Journal of Bone and Joint Surgery. American Volume
|July 12, 2023
Summary
Large language models (LLMs) in orthopaedic surgery can generate publishable text but risk AI hallucinations and misinformation. Current editorial processes need updates to detect LLM use and ensure safe integration into scientific publishing.
Area of Science:
- Artificial intelligence (AI)
- Natural language processing (NLP)
- Large language models (LLMs)
Background:
- LLMs, a subset of AI, leverage NLP for text generation, with growing applications in medicine and orthopaedic surgery.
- While LLMs can produce high-quality scientific manuscripts, they are prone to 'AI hallucinations,' presenting inaccuracies confidently.
- This poses risks of research misconduct and the dissemination of misinformation within the clinical literature.
Purpose of the Study:
- To evaluate the capabilities and risks of LLMs in generating scientific text for orthopaedic surgery.
- To assess the adequacy of current editorial processes in identifying LLM-generated content.
- To propose necessary adaptations in academic publishing for the safe utilization of LLMs.
Main Methods:
- Review of LLM capabilities in text generation for scientific manuscripts.
- Analysis of AI hallucination risks and their implications for medical literature.
- Evaluation of current academic publishing editorial workflows.
Main Results:
- LLMs demonstrate potential for generating publishable scientific text in orthopaedics.
- A significant risk exists for LLMs to introduce factual inaccuracies (hallucinations) into manuscripts.
- Existing editorial screening methods are insufficient to detect LLM involvement.
Conclusions:
- Academic publishing requires updated guidelines for responsible LLM use in orthopaedic literature.
- Enhanced editorial screening processes are essential to identify LLM-generated content.
- Proactive adaptation is necessary to mitigate risks associated with LLMs in scientific communication.
