Utilizing generative AI in ophthalmic medical paper writing: Applications, limitations, and practical tools

Generative Artificial Intelligence (GenAI) is increasingly impacting medical research and academic writing, including ophthalmology. While GenAI’s influence is expanding, there remains a lack of comprehensive literature that thoroughly addresses its applications, limitations, and ethical challenges in medical writing including ophthalmology. Generally, GenAI’s applications in this field cover ten key considerations: idea generation, literature review, institutional review board (IRB) preparation, data collection, data analysis, image generation, manuscript drafting, writing refinement, peer review, and the limitations and ethical issues that accompany these processes ( Fig. 1 ). GenAI has made significant contributions to academic writing; however, its potential risks also warrant careful examination and further investigation.

Fig. 1
Summary of the considerations and applications of generative artificial intelligence in medical paper writing. There is a total of 10 possible roles, including (1) idea generation, (2) literature review, (3) institutional review board preparation, (4) data collection, (5) data analysis, (6) image generation, (7) manuscript drafting, (8) writing refinement, (9) peer review, (10) limitations and ethical issues. The common publicly available specialized AI tools designed for medical paper writing were also demonstrated, including ChatGPT, Semantic Scholar, Dimensions, Scite, Elicit, Research Rabbit, Julius, IBM Wastonx, DeepL, Grammarly, QuillBot, DALL-E 3, Midjourney and MyLens. Created with Flaticon.com and BioRender.com.

Given that GenAI is a relatively new technology, the medical community is still in the early stages of understanding both its benefits and limitations. The issue of the “digital divide”—where some individuals lack access to or knowledge of computers, the internet, or AI tools available to others—also presents a significant barrier for the medical community to overcome. In terms of medical manuscript submissions, the guidelines for GenAI usage in ophthalmology journals listed in the Journal Citation Reports (JCR) are currently inconsistent and lack standardized directives. We manually reviewed the guidelines in JCR-listed ophthalmology journals up to September 2024 according to the methods used in previous literature and found that, among the 57 ophthalmology journals listed in the 2022 Journal Citation Reports (JCR) and published in English, 42/57 (74 %) journals provided guidance on GenAI use for authors. Most journals require disclosure of GenAI use in manuscripts and prohibit listing GenAI as an author. Among the 42 journals, 32 (76 %) required the author to take responsibility for the AI-generated content. As for specific GenAI tools, ChatGPT was the most frequently mentioned (n = 20, 48 %), with 27 journals mentioning large language models, and only 3 journals mentioning specific AI tools other than ChatGPT. This issue mirrors similar challenges in journals across other medical fields.

On a positive note, various AI tools have proven highly beneficial for medical paper writing. Non-GenAI tools like Semantic Scholar, Elicit, and Research Rabbit enhance literature review, search efficiency, contextual analysis, generating search terms, summarizing articles, extracting key information, and assisting in systematic review screening. GenAI tools, including Grammarly, and QuillBot, improve writing quality. Midjourney supports visual creation. ChatGPT aids in idea generation, literature review, IRB preparation, data collection, and manuscript drafting ( Supplementary Table 1 ). While its accuracy and reference reliability are still lower than that of human researchers, it serves as an effective tool for initial screening and drafting. When used appropriately, these tools, along with new emerging AI tools, can greatly assist in the writing of medical articles.

Despite its benefits, GenAI has notable limitations, particularly in data privacy, academic ethics, content authenticity, and the risk of “hallucination”, where it generates meaningless or incorrect information. These challenges emphasize the need for rigorous review and cross-checking by authors and reviewers to ensure accuracy and prevent misinformation. In medical writing, GenAI also raises ethical concerns, such as the potential for fabricated data or unintentional copyright violations. Thus, GenAI should be regarded as a complementary tool to human expertise, not a replacement, to maintain research integrity and originality. GenAI’s ethical and technical limitations are especially significant for medical writing in ophthalmology, where sensitive patient data and image-dependent analyses increase the need for data privacy and ethical transparency. Responsible use of GenAI requires both disclosure and adherence to best practices to avoid over-reliance on AI-generated content. Without consistent guidelines, authors may inadvertently misuse GenAI, affecting the quality and integrity of research. Nowadays, many specialized tools are being used to detect fraudulent articles. Authors and reviewers need to familiarize themselves with these tools to prevent the misuse of GenAI in academic writing.

In summary, while GenAI offers transformative benefits for ophthalmic medical writing, unified guidelines are crucial for effective management. Consistent policies will help regulate its use, allowing researchers to leverage GenAI’s advantages while addressing ethical concerns, ensuring research quality and reliability.

Supplementary material

  Supplementary material

Supplementary material

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Apr 20, 2025 | Posted by in OPHTHALMOLOGY | Comments Off on Utilizing generative AI in ophthalmic medical paper writing: Applications, limitations, and practical tools

Full access? Get Clinical Tree

Get Clinical Tree app for offline access