Chinese Medical Academy of new crown virus mouse model image is alleged to be a multi-use, suspected image fraud?

On May 12th Elisabeth Bik, an academic counterfeiter, reported that in a recent paper published in Nature, images from two different groups "may overlap."
This tweet has been retweeted thousands of times, and the team and Nature have yet to respond publicly.
image fraud has become a major problem in the biomedical field.
recently, a number of academic publishers have set up a working group to explore how to establish a mechanism for image re-examination and cooperation, and plans to more clearly classify and define image issues. On May 7,
, the Chinese research team published a paper in Nature online in the form of an accelerated review entitled "The pathogenicity of SARS-CoV-2 in hACE2 genetically modified mice" (The pathogenicity of SARS-CoV-2 in hACE2 transgenic mice) in mice, and transferred the human agiotensin conversion enzyme 2 (hACE2) gene into mice to build a model of mice that can infect the new coronavirus to promote the development of relevant drugs.
the study was originally published on February 28 in preprinted bioRxiv. The
research was carried out by the Institute of Experimental Animals of the Chinese Academy of Medical Sciences, the Key Laboratory of Animal Models of New Infectious Diseases in Beijing, the Comparative Medical Center of Beijing Concord Medical College and the National Institute for Viral Disease Control and Prevention, and the co-authors were Professor Qin Chuan, Director of the Institute of Medical Experimental Animals of the Chinese Academy of Medical Sciences, and Wu Guizhen, a researcher at the China Center for Disease Control and Prevention's Viral Disease Prevention and Control Institute. On May 12,
, academic fake celebrity Elisabeth Bik pointed out that in fig.2, images from two different control groups appeared to coincide in the region.
in the hedding images of lung tissue in wild strain mice (WB-HB-01) and hACE2 blank control mice (ACE2-Mock), the texture of some areas appears to be highly consistent (see blue box).
pubPeer users conducted an analysis of image processing, the two picture blue frame area of the intensity and edge height consistent.
asked Bik what might be the possible motivation for such reused images, And Bik replied: "I noticed that the coincident part was more likely to appear in two different control groups, as in this case."
this may mean that the researchers set up only one control group to save the experimental animals, or to meet the requirements of the reviewers.
": "This is an 'unedited manuscript' that 'will be reviewed before it is finally published'."
this is likely to be considered a 'small mistake in the certification process'.
", but it was clear that a picture of the problem was enough to cause a strong distrust among peers.
many people listed other doubts in the paper below Bik's twitter feed.
repeatedly banned image fraud now the paper text weight-checking technology has been relatively mature, in contrast, the recognition of image fraud technology has many shortcomings.
2016, the Bik team conducted a manual analysis of more than 20,000 papers in the biomedical field, and found that as many as 4% of the papers contained problematic images.
in recent years, the Bik team has also uncovered a number of incidents of image fraud by Chinese scholars.
November 2019, the Bik team pointed out that Cao Xuetao, president of Nankai University and a member of the Chinese Academy of Engineering, had suspected data on several papers, and that the experimental images had PS traces.
, more than 60 papers were eventually identified as questionable on PubPeer, and 13 corrections were made as of May this year, with most corrections stating that the findings were not affected, Nature reported.
On February 21 this year, the Bik team announced that it had found that more than 400 papers may have come from the same "paper workshop".
these papers are all from dozens of hospitals in China, including a number of triple-A hospitals, involving different authors, different research areas.
many of these papers also reveal their feet because of the high lysimilarity of the images.
another report in Nature, many journals are not currently hiring someone to review the images in their papers, citing the cost of doing so or taking too much time.
existing technology cannot scan paper pictures in large numbers.
companies that have started offering small-scale image review services such as LPIXEL in Japan, Proofig in Israel and Resis in Italy.
researchers are also developing software tools that can cross-contrast images from multiple papers.
, however, what big publishers need is software that can handle a large number of images, and that it can be integrated into the peer review process, preferably by cross-contrasting images from a large number of papers.
images also need to "check the database" in order to solve this problem, the academic publishing community has taken action.
April this year, the International Association of Scientific, Technical and Medical Publishers (STM) Standards and Technology Committee (STEC) has set up a working group to explore how to automatically identify images in papers. The
Working Group is headed by IJsbrand Jan Aalbersberg, Senior Vice President of Academic Integrity at Airwell, and is represented by academic publishers such as Springer Nature, Veoli, and T.F.
task is to sort out the minimum requirements for image scanning software and explore how publishers can use this technology to process hundreds of thousands or even millions of papers.
the team also wants to categorize "the type and severity of image problems" and "develop guidelines on what circumstances image adjustments are permissible." Catriona Fennell, director of publishing services at
, who is also a member of the working group, said Eswell was highly concerned about the "industrial fraud" of a few papers. This is the case
the paper workshop mentioned above.
noted that similarities in different papers are difficult to detect in the peer review process, not only because most reviewers do not pay attention to such issues, but also because papers produced in this way are submitted to different journals at the same time and reviewed confidentially.
therefore, the academic publishing community needed a shared database to identify images that were reused between papers, just as crossCheck did for text checking. "We also need the same collaborative model when it comes to image review," said
Fennell.
" Source: Research Circle.