As part of continued efforts to bolster the integrity of the scholarly record, the Science journals have partnered with Proofig AI image integrity software.
Adoption of Proofig AI positions authors to resolve image-related issues before a paper is published. It also ensures the highest standards of accuracy in work published in the six Science family journals.
“Rigorous data are a cornerstone of our publications,” said Valda Vinson, Executive Editor of Science. “Image manipulation and duplication – whether intentional or mistaken – impact trust in our content, and we are committed to preventing these issues, including by exploring innovative technologies that can assist our editors.”
Research content at the six Science family journals, published by the American Association for the Advancement of Science and spanning multiple disciplines, features a range of images – each of which contains valuable information in support of the related study. Image accuracy is critical.
Before a paper is published in any Science family journal, it goes through various checks, including those related to image integrity. But until recently, efforts to screen images have been largely manual and highly time-requisite.
In summer 2023, the Science journals began trialing Proofig AI on select manuscripts at the revision stage. Proofig AI can detect numerous types of the most common image integrity issues, including image manipulation and duplication, which may represent serious forms of misconduct. But even where duplication has occurred due to honest error, it still damages the integrity of the scholarly record, requiring corrective action.
“The ability to better detect image duplications is particularly helpful to our editorial team as duplications are hard to spot with the naked eye,” said Vinson.
Proofig AI’s filters – equipped with enhanced visualization techniques – flag image concerns for editors to investigate more thoroughly. Editors then determine how to proceed, seeking author clarification where necessary.
So far, some image issues discovered in papers at Science journals have required editors to contact authors to clear up matters such as mistaken duplications.
Authors have been appreciative of issues being caught before publication instead of after, where stakes can include need for a correction or retraction, or even reputational damage.
Proofig AI will be particularly critical at the journals as the Science family grows in volume, increasing the number of images requiring careful screening.
“It’s great to see how automating image integrity checks within publication workflows has already helped the Science family journals streamline their publication process prior to manuscript acceptance, ensuring the highest standards of integrity,” said Dror Kolodkin-Gal, Co-founder of Proofig AI. “Proofig empowers the editorial team to make swift and informed decisions. The utilization of our AI tool by such a prestigious publishing house may pave the way for additional publishers to ensure higher integrity assurance before publication. Altogether, this will benefit all parties involved – and the scientific community."