After a few months, what are the uses of OpenAI's ChatGPT in medicine? A Scopus-based systematic review

被引：1

作者：

Moya-Salazar, Jeel ^{[1
]}

Salazar, Carmen R. ^{[2
]}

Delzo, Stephany S. ^{[3
]}

Goicochea-Palomino, Eliane A. ^{[4
]}

Rojas-Zumaran, Victor ^{[3
]}

机构：

[1] Univ Norbert Wiener, Digital Transformat Ctr, Dept Res, Lima, Peru

[2] Nesh Hubbs, Qualitat Unit, Lima, Peru

[3] Hosp Nacl Docente Madre Nino San Bartolome, Dept Pathol, Lima, Peru

[4] Univ Tecnol Peru, Fac Med Technol, Lima, Peru

来源：

ELECTRONIC JOURNAL OF GENERAL MEDICINE | 2024年 / 21卷 / 02期

关键词：

machine learning; Scopus; healthcare; ChatGPT; medical writing; artificial intelligence; ethics; ARTIFICIAL-INTELLIGENCE; OPPORTUNITIES; SCIENCE;

D O I：

10.29333/ejgm/14354

中图分类号：

R5 [内科学];

学科分类号：

1002 ; 100201 ;

摘要：

Introduction: Acute ChatGPT, an artificial inteeligence-powered chatbot, presents a potential influence on healthcare and research in medicine. In this investigation, we systematically reviewed medical literature regarding the utilization of ChatGPT, utilizing the Scopus database. Methods: This systematic review follows the preferred reporting items for systematic reviews and meta-analyses guidelines, and according to the search in Scopus, 11,219 articles were identified. Original studies, clinical trials, correspondence articles, letters, notes, and research protocols were included, resulting in 14 studies. Results: All studies were published in 2023, among them 35.7% (5/14) were editorials and 14.3% (2/14) were opinion or perspective articles. Only one study (19) conducted a pilot study on the application of ChatGPT in medical diagnosis. Furthermore, 64.3% of the studies were published in quartile 1 journals according to Scimago rankings, and the USA had the highest production of articles on ChatGPT in medicine (35.7%). The included studies highlighted the potential and challenges of ChatGPT in various domains of healthcare. Ethical concerns, such as data privacy and bias, were also analyzed. Various fields of health were evaluated with ChatGPT, including nuclear medicine, gynecology, nuclear medicine, and emergency medicine. ChatGPT was considered a tool to improve patient education. ChatGPT has a promising role in scientific writing, emphasizing the need for human involvement, ethical guidelines, and addressing concerns related to bias, fabrication, and plagiarism. Conclusions: Scopus-retrieved articles underscored the potential advantages and ethical dilemmas associated with the integration of ChatGPT in healthcare. While ChatGPT demonstrated promising applications in various medical domains, caution, further evaluation, and ethical considerations are needed for responsible implementation in real -world clinical settings.

引用

页数：7