aifollow.news
返回 IT之家 科技新闻
IT之家 科技新闻· · 原发布时间 AI 评分33

鼻咽癌论文实验图片遭质疑,湖南省科技厅同步核查

AI 辅助摘要

网友发现论文实验图片中的尺子每厘米有12个刻度格,怀疑图片由AI生成。中南大学湘雅二医院已介入调查;湖南省科技厅表示将同步核实是否存在学术不端及科研程序问题。

正文 · AI 翻译

IT Home reported on November 18 that a research paper on nasopharyngeal carcinoma published in the Journal of Translational Medicine drew questions from internet users yesterday, and the topic briefly trended on Weibo.

The paper contains several irregularities in its experimental images, data and author affiliation information. For example, a 1-centimeter span on a ruler contains 12 divisions. Some internet users therefore suspect that the images may have been generated by AI.

On November 17, the Supervision and Research Integrity Division of the Hunan Provincial Department of Science and Technology said it had instructed the authors' institution, the Second Xiangya Hospital of Central South University, to investigate. The department will also independently check for possible academic misconduct and problems with research procedures.

The disputed paper, titled “Ferroptosis inhibition via the ROS-GPX4 axis drives microplastic-induced malignant progression of nasopharyngeal carcinoma,” was published online on December 22, 2025. Its corresponding author is listed as “Lin Zhao,” and the first author affiliation listed is the Department of Pathology at the Second Xiangya Hospital of Central South University.

The Beijing News noted that the author affiliation address on the paper's first page says “Changsha, Human” instead of “Hunan,” the English spelling of the province's name. The postal code is given as “41001l,” with a letter “l” rather than the digit “1” at the end.

More prominent points of suspicion appear in the experimental images on pages 10 and 15. The rulers used to indicate sample size have 12 small divisions within a 1-centimeter interval, whereas a conventional ruler has 10 divisions in that interval. This detail, which conflicts with basic physical expectations, has become the main basis for internet users' speculation that the images may have been generated by AI.

In addition, some internet users described the tumor images in the paper as “resembling shrimp paste” and questioned some of the experimental data and results. These claims remain part of online discussion and have not been independently verified.

IT Home notes: The Journal of Translational Medicine was founded in 2003 and is published by Springer Nature's BioMed Central. Its 2025 impact factor was 9.7, and it is indexed in databases including SCIE, Scopus and MEDLINE.

Springer Nature's publishing policy explicitly states that AI-generated images are generally not permitted for publication. They are considered only under limited exceptions, such as when their legal provenance is clear, the content directly concerns AI, or they are based on verifiable scientific data, and they must be clearly labeled “AI-generated.”

The research department of the Second Xiangya Hospital told The Beijing News on November 17 that the hospital had begun an investigation. As of publication, neither Springer Nature's research integrity department nor the paper's corresponding author had replied to the media's verification emails.

Advertising statement: External links in this article, including but not limited to hyperlinks, QR codes and passcodes, are provided to convey more information and save readers time in selecting information. The results are for reference only. All IT Home articles containing external links include this statement.

发现内容有误?提交纠错