Related Experiment Video
Updated: May 6, 2026
05:01
A Pathway Association Study Tool for GWAS Analyses of Metabolic Pathway Information
Published on: July 1, 2020
3.3K
Using GPT-4 to write a scientific review article: a pilot evaluation study
Zhiping Paul Wang1, Priyanka Bhandary1, Yizhou Wang1
1Department of Computational Biomedicine, Cedars Sinai Medical Center, 700 N. San Vicente Blvd, Pacific Design Center, Suite G-541, West Hollywood, CA, 90069, USA.
Biodata Mining
|June 18, 2024
Summary
This study evaluated GPT-4
Area of Science:
- Artificial Intelligence
- Biomedical Sciences
- Scientific Writing
Background:
- Large language models, such as GPT-4, are increasingly utilized in scientific research.
- The application of advanced AI in generating content for scientific review papers requires thorough evaluation.
Purpose of the Study:
- To assess GPT-4's proficiency in generating text, tables, and diagrams for biomedical reviews.
- To evaluate the consistency and potential plagiarism risks associated with GPT-4 in scientific writing.
Main Methods:
- GPT-4 was tested for its ability to generate various components of biomedical review papers.
- Text generation consistency and plagiarism checks were performed on the AI-generated content.
Main Results:
- GPT-4 demonstrated capabilities in generating text, tables, and diagrams.
- Areas for improvement include handling complex biomedical concepts, precise table generation, and scientific diagram creation.
- Consistency and plagiarism were noted as factors requiring attention.
Conclusions:
- GPT-4 shows potential as a tool for scientific review paper composition.
- Enhancements are needed in document processing, conceptual understanding, table generation, and diagram creation for scientific applications.
- Further development of AI tools like ChatGPT is recommended to better serve the scientific community.
