Blood Res.  2024;59:45. 10.1007/s44313-024-00045-3.

Strategies for integrating ChatGPT and generative AI into clinical studies

• Lee J

• KMID: 2564320
• DOI: http://doi.org/10.1007/s44313-024-00045-3

Abstract

Large language models, specifically ChatGPT, are revolutionizing clinical research by improving content creation and providing specific useful features. These technologies can transform clinical research, including data collection, analysis, interpretation, and results sharing. However, integrating these technologies into the academic writing work‑ flow poses significant challenges. In this review, I investigated the integration of large-language model-based AI tools into clinical research, focusing on practical implementation strategies and addressing the ethical considerations asso‑ ciated with their use. Additionally, I provide examples of the safe and sound use of generative AI in clinical research and emphasize the need to ensure that AI-generated outputs are reliable and valid in scholarly writing settings. In conclusion, large language models are a powerful tool for organizing and expressing ideas efficiently; however, they have limitations. Writing an academic paper requires critical analysis and intellectual input from the authors. Moreo‑ ver, AI-generated text must be carefully reviewed to reflect the authors’ insights. These AI tools significantly enhance the efficiency of repetitive research tasks, although challenges related to plagiarism detection and ethical use persist.

Keyword

Generative AI; Clinical Research, ChatGPT; Research Methodology; Academic Writing

Figure

• 

  Fig. 1 Informed consent form generated using GPT for a clinical study on brentuximab vedotin in patients with CD30-positive lymphoma

• 

  Fig. 2 Case report form (CRF) generated using GPT for a clinical study on brentuximab vedotin in patients with CD30-positive lymphoma

• 

  Fig. 3 Database form Excel file generated using GPT for a clinical study on brentuximab vedotin in CD30-positive lymphoma

• 

  Fig. 4 Webpage of E-CRF Excel file generated using Claude 3.5 sonnet for a clinical study on brentuximab vedotin in patients with CD30-positive lymphoma

• 

  Fig. 5 Data preprocessing (pathologic report parsing) process generated using GPT4o. The pathology data, including tumor location, size, histologic grade, and underlying liver disease, are organized into a clear, easily interpretable format. This facilitates time-saving for researchers and clinicians by efficiently converting complex medical data into a standardized table format

• 

  Fig. 6 Python code generated using GPT4o for data analysis (Machine learning model)

• 

  Fig. 7 Example of three-group comparison table generation using GPT4o in the risk factor of reintubation after liver transplantation study

• 

  Fig. 8 Example of figure generation using GPT4o in the study to identify the factors affecting recurrence after liver transplantation in patients with hepatocellular carcinoma, analyzing each variable in the form of a heatmap

Reference

1. Haque MA, Li S. 2024; Exploring ChatGPT and its impact on society. AI Ethics. https://doi.org/10.1007/s43681-024-00435-4. DOI: 10.1007/s43681-024-00435-4.

2. Mehta N, Devarakonda MV. 2021; Chatbots for healthcare: a systematic literature review. Stud Health Technol Inform. 281:998–1002.

3. Open AI. Introducing ChatGPT. Published November 30, 2022. https://openai.com/blog/chatgpt. Accessed 13 May 2024.

4. Lee P, Bubeck S, Petro J. 2023; Benefits, limits, and risks of GPT-4 as an AI chatbot for medicine. N Engl J Med. 388(13):1233–9. https://doi.org/10.1056/NEJMsr2214184. DOI: 10.1056/NEJMsr2214184. PMID: 36988602.

5. Microsoft. Microsoft and OpenAI extend partnership. Published January 23, 2023. https://blogs.microsoft.com/blog/2023/01/23/microsoft-andopenai-extend-partn ership/. Accessed 29 July 2024.

6. Anthropic. Introducing Claude. Published March 14, 2023. https://www.anthropic.com/index/introducing-claude. Accessed 29 July 2024.

7. Perplexity AI. About Pperplexity AI. https://www.perplexity.ai/about. Accessed 29 July 2024.

8. Dave T, Athaluri SA, Singh S. 2023; ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations. Front Artif Intell. 6:1169595. https://doi.org/10.3389/frai.2023.1169595. DOI: 10.3389/frai.2023.1169595. PMID: 37215063. PMCID: PMC10192861.

9. Schwalbe N, Wahl B. 2020; Artificial intelligence and the future of global health. Lancet. 395(10236):1579–86. https://doi.org/10.1016/S0140-6736(20)30226-9. DOI: 10.1016/S0140-6736(20)30226-9. PMID: 32416782.

10. Bader R, Imam A, Alnees M, et al. 2024; REMOVED: Successful management of an iatrogenic portal vein and hepatic artery injury in a 4-month-old female patient: a case report and literature review. Radiol Case Rep. 19(6):2106–11. https://doi.org/10.1016/j.radcr.2024.02.037. DOI: 10.1016/j.radcr.2024.02.037. PMID: 38645539. PMCID: PMC11026926.

11. Else H. 2023; Abstracts written by ChatGPT fool scientists. Nature. 613(7945):423. https://doi.org/10.1038/d41586-023-00056-1. DOI: 10.1038/d41586-023-00056-7. PMID: 36635510.

12. International Committee of Medical Journal Editors. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. http://www.icmje.org/recommendations/. Accessed 29 July 2024.

13. Lee JY. 2023; Can an artificial intelligence chatbot be the author of a scholarly article? J Educ Eval Health Prof. 20:6. https://doi.org/10.3352/jeehp.2023.20.6. DOI: 10.3352/jeehp.2023.20.6. PMID: 36842449. PMCID: PMC10033224.

14. The Lancet. Information for authors: the use of AI and AI-assisted technologies in scientific writing. https://www.thelancet.com/preparing-yourmanuscript. Accessed 29 July 2024.

15. Flanagin A, Kendall-Taylor J, Bibbins-Domingo K. 2023; Guidance for authors, peer reviewers, and editors on use of AI, language models, and chatbots. JAMA. 330(8):702–3. https://doi.org/10.1001/jama.2023.12500. DOI: 10.1001/jama.2023.12500. PMID: 37498593.

16. Ganjavi C, Eppler MB, Pekcan A, et al. 2024; Publishers' and journals' instructions to authors on use of generative artificial intelligence in academic and scientific publishing: bibliometric analysis. BMJ. 384:e077192. https://doi.org/10.1136/bmj-2023-077192. DOI: 10.1136/bmj-2023-077192. PMID: 38296328. PMCID: PMC10828852.

17. Hosseini M, Rasmussen LM, Resnik DB. 2024; Using AI to write scholarly publications. Account Res. 31(7):715–23. DOI: 10.1080/08989621.2023.2168535. PMID: 36697395.

18. Open AI. GPT-4o system card. Published August 8, 2024. https://openai.com/index/gpt-4o-system-card/. Accessed 10 Aug 2024.

19. Kim SJ, Do YR, Lee HS, et al. 2023; A multi-center and non-interventional registry of brentuximab vedotin in patients with relapsed or refractory CD30-positive lymphoma: the CISL1803/BRAVO study. Blood Res. 58(4):194–200. https://doi.org/10.5045/br.2023.2023206. DOI: 10.5045/br.2023.2023206. PMID: 38031473. PMCID: PMC10758628.

20. Open AI. Prompt engineering: ask the model to adopt a persona. https://platform.openai.com/docs/guides/prompt-engin eering/tactic-ask-themodel-to-adopt-a-persona. Accessed 29 July 2024.

21. Masukume G. The impact of AI on scientific literature: a surge in AI-associated words in academic and biomedical writing. medRxiv. Preprint posted online May 31, 2024. https://doi.org/10.1101/2024.05.31.24308296. DOI: 10.1101/2024.05.31.24308296. PMCID: PMC10906258.

22. Habibzadeh F. 2023; GPTZero Performance in Identifying Artificial Intelligence-Generated Medical Texts: A Preliminary Study. J Korean Med Sci. 38(38):e319. https://doi.org/10.3346/jkms.2023.38.e319. DOI: 10.3346/jkms.2023.38.e319. PMID: 37750374. PMCID: PMC10519776.