AI software helps stop image fraud in academic papers

AI software helps stop image fraud in academic papers

AI software helps stop image fraud in academic papers

Negative image of a Western blot test result.
Magnify / Colored image of a Western blot test result.

Scientific publishers such as the American Association for Cancer Research (AACR) and Taylor & Francis have begun trying to detect fraud in academic articles with an AI image-checking program called Proofig, The Register reports. Proofig, a product of an Israeli firm of the same name, aims to help use “artificial intelligence, computer vision and image processing to assess image integrity in scientific publications,” according to the company’s website.

During a trial that ran from January 2021 to May 2022, the AACR used Proofig to screen 1,367 papers accepted for publication, according to The Register. Of these, 208 articles required author contact to clarify issues such as erroneous duplications, and four articles were retracted.

In particular, many journals need help detecting image duplication fraud in Western blots, which are a specific style of protein detection images consisting of line segments of different widths. Subtle differences in a blot’s appearance can translate into dramatically different conclusions about test results, and many cases of academic fraud have seen unscrupulous researchers duplicate, crop, stretch, and rotate Western blots to make it appear as if they have multiple (or different ) data than they really do. Detecting duplicate images can be tedious work for human eyes, which is why some companies like Proofig and ImageTwin, a German company, are trying to automate the process.

But both Proofig’s and ImageTwin’s solutions currently have significant limitations, according to The Register. First, human expertise is still needed to interpret detection results and reduce false positives. Second, Proofig is currently expensive due to its computationally intensive process, costing $99 to analyze 120 images for an individual (the journals have negotiated cheaper prices). Currently, both high costs and the requirement for manual review prevent journals from analyzing every article at the submission stage. Instead, they have reserved its use for later in the publishing process.

Academic fraud, although uncommon, can still have a devastating effect on a publication’s reputation. Between the sheer volume of academic papers being published today and recent revelations about image fraud in widely cited Alzheimer’s research, the field seems ripe for computer vision tools that can help people with fraud detection. Their overall effectiveness—and how widely they are adopted—is still a developing story.

Leave a Reply

Your email address will not be published.