Enlarge / Colorized picture image of a Western blot check outcome.
Scientific publishers such because the American Association for Cancer Research (AACR) and Taylor & Francis have begun trying to detect fraud in academic paper submissions with an AI image-checking program referred to as Proofig, experiences The Register. Proofig, a product of an Israeli agency of the identical identify, goals to assist use “artificial intelligence, computer vision and image processing to review image integrity in scientific publications,” in line with the corporate’s web site.
During a trial that ran from January 2021 to May 2022, AACR used Proofig to display screen 1,367 papers accepted for publication, in line with The Register. Of these, 208 papers required writer contact to clear up points akin to mistaken duplications, and 4 papers had been withdrawn.
In specific, many journals need assistance detecting image duplication fraud in Western blots, that are a selected type of protein-detection imagery consisting of line segments of varied widths. Subtle variations in a blot’s look can translate to dramatically totally different conclusions about check outcomes, and lots of circumstances of academic fraud have seen unscrupulous researchers duplicate, crop, stretch, and rotate Western blots to make it seem like they’ve extra (or totally different) information than they actually do. Detecting duplicate photographs will be tedious work for human eyes, which is why some companies like Proofig and ImageTwin, a German agency, are trying to automate the method.
But each Proofig’s and ImageTwin’s options presently have vital limitations, in line with The Register. First, human experience continues to be essential to interpret detection outcomes and cut back false positives. Second, Proofig is presently costly resulting from its computationally intensive course of, costing $99 to investigate 120 photographs for a person (the journals have negotiated cheaper charges). Currently, each excessive price and the requirement for handbook oversight is maintaining the journals from analyzing each paper on the submission stage. Instead, they’ve been reserving its use for later in the publication course of.
Academic fraud, whereas unusual, can nonetheless have a devastating impact on a publication’s repute. Between the large quantity of academic papers being printed as we speak and up to date revelations about image fraud in extensively cited Alzheimer’s analysis, the sector does appear ripe for pc imaginative and prescient instruments that may help people with fraud detection. Their total effectiveness—and the way extensively they turn out to be adopted—continues to be a growing story.