Pre-submission image integrity screening with Proofig AI surfaces suspected duplications, manipulations, and other figure problems before they reach peer review. By catching potential issues early, researchers and publishers can reduce the risk of post-publication corrections and retractions, protecting both scientific credibility and individual careers.
Why Are Retractions So Closely Tied to Image Issues?
Retractions carry significant financial and reputational consequences for researchers, institutions, and journals alike. Image-related integrity problems are among the most common triggers. A 2025 study published in Accountability in Research found that self-plagiarism accounted for 25.4% of papers retracted between 2001 and 2022 in the Retraction Watch database, with image duplication identified as the most common reason. These are not rare cases. Estimates suggest that roughly 35,000 articles published between 2009 and 2016 may be prone to post-publication retraction due to faulty figures.
The problem is not limited to a few bad actors. Image integrity analyst Jana Christopher, in an expert interview published by UKRIO, reported that across the journals where she screened manuscripts at acceptance, 20 to 35 percent were flagged for image-related problems. In the same interview she noted that acceptance was ultimately rescinded in one to eight percent of cases. That volume makes manual review of every figure impractical, especially as submission rates continue to climb.
Why Do Even Careful Researchers Face This Risk?
Many image integrity issues are not deliberate. They enter manuscripts through complex, multi-person workflows. A principal investigator may uphold the highest ethical standards and still unknowingly submit a manuscript containing a problematic figure, because the issue was introduced by a collaborator, a student, or someone else involved in data handling and figure preparation.
A peer-reviewed study at the University of Split tested medical students and researchers on their ability to spot image duplications. Students correctly identified a median of only 29.4% of duplications, and researchers identified just 32.4%, with no statistically significant difference between the two groups. This demonstrates that even trained eyes miss the majority of problems without tool-assisted support.
Pre-submission screening is not an accusation. It is a professional safeguard, comparable to running a reference check or a text-similarity scan before submitting a manuscript.
What Does Proofig AI Check, and Why Does It Matter Before Submission?
Proofig AI’s detection capabilities cover the five main categories of image integrity issues that lead to retractions, corrections, and editorial desk rejections.
| Capability | What It Detects | Why It Matters Pre-Submission |
|---|---|---|
| Image Manipulation Detection | Cloning, splicing, deletion, content-aware edits within a single image | Catches alterations that may have been introduced during figure preparation |
| Duplication & Reuse Analysis | Scaling, rotation, flipping, full and partial overlap within a manuscript | Identifies accidental reuse of panels or regions across figures |
| AI-Generated Image Identification | Synthetic images across microscopy, histology, Western blots, medical scans, and more | Flags figures that may not represent real experimental data |
| Image Plagiarism Checker | Reuse of sub-images checked against more than 155 million images in the PubMed Open Source Database | Surfaces potential overlap with previously published work |
| Self-Plagiarism Control (My Database) | Comparison against a researcher’s own prior publications | Prevents unintentional reuse of one’s own previously published figures |
Proofig covers all major scientific image types, including confocal, light, and electron microscopy, Western blots and gels, FACS, histology, cell plates, animal imaging, and medical scans. The workflow is straightforward: upload a manuscript or image files, Proofig scans all figures automatically, review flagged items using built-in forensic tools, and export a report for submission or internal records.
Does Proofig Make Final Judgments About Fraud?
No. Proofig surfaces suspected issues and flags potential integrity problems for review. It does not determine guilt, assign intent, or make final judgments about fraud or misconduct. The tool supports researchers and editorial teams in identifying high-risk visuals before submission, so that a human can investigate further.
This distinction matters for both audiences. Researchers need to know the tool is a safeguard, not a judge. Publishers need to know it integrates into, rather than replaces, editorial judgment. Proofig is built to minimize false positives, using type-specific detection that accounts for the distinct visual characteristics of each image modality. That means researchers spend their time reviewing genuine concerns, not chasing false alarms.
How Does Proofig Fit into Existing Workflows?
For researchers, Proofig functions as a final integrity check before submission. Just as you would run a spell-checker on your text, running an image integrity check on your figures is a practical step that can prevent a rejection, correction, or retraction.
For publishers, Proofig integrates with editorial submission systems. The Aries Systems integration enables automated image integrity screening within existing editorial workflows, making it possible to screen at scale without adding manual burden.
For institutions seeking a broader quality assurance solution, Proofig also offers PubShield, a manuscript quality assurance hub that combines image integrity screening with other checks, including text similarity, reference integrity, AI-generated text detection, and data compliance, all in one dashboard.
Who Has Adopted Proofig AI?
Third-party adoption provides concrete validation. The Science family of journals adopted Proofig in 2024 to detect altered images, after a multi-month pilot demonstrated that problematic figures could be identified before publication. Science Editor-in-Chief H. Holden Thorp described the adoption as analogous to the journals’ earlier adoption of iThenticate for text-plagiarism detection.
MDPI is implementing Proofig AI across selected journals. As Milos Cuculovic, Head of Technology Innovation at MDPI, stated: “From the experts, it seemed Proofig AI was by far the best choice, and here we are: further ensuring research integrity and guaranteeing the high quality of research output published by MDPI.”
These are operational deployments at major publishers rather than short-term trials.
Why Is Catching Issues Early Better Than Correcting Them Later?
A post-publication correction or retraction is far more damaging than a pre-submission revision. Nature has reported on the intense personal and career stress researchers face when retracting papers, even when the underlying cause is honest error. Pre-submission screening could prevent many of these situations from arising in the first place.
Catching a suspected duplication or manipulation before a paper is published protects the researcher’s name, the lab’s reputation, the journal’s credibility, and the integrity of the scientific record. It is a proactive, professional step, not a concession that something is wrong. For any researcher preparing a manuscript or any editorial team processing submissions, running an image integrity check before publication is one of the most practical ways to reduce retraction risk and uphold the standards the scientific community depends on.
