Aim: This study explores the effectiveness of artificial intelligence (AI) versus human expertise in generating scientific article titles within the context of plastic surgery. Acknowledging the pivotal role titles play in attracting readership and conveying a study's significance, this research investigates whether AI can match or surpass human capability in crafting compelling titles for academic articles.
Material and Methods: Utilizing a sample of original articles from the Plastic and Reconstructive Surgery Journal, published in January and February 2023, the study employs OpenAI's ChatGPT to generate alternative titles based on article abstracts. A survey among plastic, reconstructive, and aesthetic surgery experts, with a blinded setup regarding the origin of the titles (AI-generated or original), facilitates the comparison. Statistical analysis, including t-tests and the Mann-Whitney U test, assesses preferences across general and specific subject areas within plastic surgery.
Results: Findings indicate a predominant preference for original titles, yet in certain subjects like hand/peripheral nerve surgery, AI-generated titles were more favored. On the other hand, in experimental and cosmetic subjects original titles were more favored.
Conclusion: The study highlights the potential and limitations of AI in academic creativity, suggesting a nuanced view where AI's effectiveness varies by subject matter, such as experimental and cosmetic subjects. It concludes that while AI exhibits competence in title generation, human oversight remains crucial, especially in areas requiring deep expertise and nuanced understanding. This investigation contributes to the discourse on AI's role in academic publishing, emphasizing the need for balanced integration of AI tools in the scientific communication process.
We would like to express our infinite gratitude to the plastic surgeons who participated in this survey, which took a long time to respond to, and dedicated their valuable time to this study.
| Primary Language | English |
|---|---|
| Subjects | Clinical Sciences (Other) |
| Journal Section | Original Article |
| Authors | |
| Early Pub Date | April 15, 2025 |
| Publication Date | April 30, 2025 |
| Submission Date | March 31, 2024 |
| Acceptance Date | October 10, 2024 |
| Published in Issue | Year 2025 Volume: 35 Issue: 2 |
The Journal of General Medicine is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY NC).