Role of large language models in medical image processing: issues that should be considered

We would like to respond to a comment on the published article entitled “The role of large language models in medical image processing: a narrative review (1)”. Even if the literature search for this report appears thorough and spans a sizable amount of time, there are a few possible flaws that might be fixed to strengthen the research as a whole. First off, limiting publications to only English could lead to bias and reduce how inclusive the results are. There might be insightful studies on the subject that have been published in other languages. Comparing research covering a larger range of languages might be an intriguing way to investigate the effects of linguistic constraints.

Furthermore, while recent articles on the database are mentioned in the review, neither the time period nor the methodology utilized to find these papers are specified. Further information about the selection procedure and the integration of these papers into the study would be helpful. A detailed justification for using the Web of Science and PubMed databases as the main sources for the literature search would also be beneficial. Exist any more reliable databases in the field of medical imaging that weren’t taken into account?

It would be beneficial to look at the restrictions and difficulties related to large language models (LLMs) in medical image processing in future research. Even if the study emphasizes the advantages and possibilities of these models, it’s crucial to talk about any potential drawbacks and areas that could use development. Are there any particular kinds of medical imaging data, for instance, that could not be ideal for LLMs? Do ethical issues need to be taken into account when applying these models in a medical setting?

Evaluation of the findings’ generalizability and reproducibility may also be the main emphasis of future research. Do the improvements and changes seen in medical image processing using LLMs hold true for various datasets and healthcare facilities? Exist any particular elements that affect how well LLMs function and execute in practical situations? Overall, the study offers insightful information about how LLMs affect medical imaging processing, but it could be strengthened by addressing these flaws and presenting pertinent research issues for later investigations. Lastly, it should be emphasized once more that the user of the AI system bears the ultimate responsibility for maintaining moral norms (2).

Appendix 1: Response to “Role of large language models in medical image processing: issues that should be considered”.

Acknowledgments

Funding: None.

Footnote

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-89/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Tian D, Jiang S, Zhang L, Lu X, Xu Y. The role of large language models in medical image processing: a narrative review. Quant Imaging Med Surg 2024;14:1108-21. [Crossref] [PubMed]
2. Kleebayoon A, Wiwanitkit V. ChatGPT, critical thing and ethical practice. Clin Chem Lab Med 2023;61:e221. [Crossref] [PubMed]