Imagetwin is an AI-based software solution designed to detect integrity issues in figures found in scientific publications, particularly within the life sciences. The software efficiently identifies inappropriate manipulation and duplication across a variety of figure types, including blots (such as western blots), microscopy images, and light photography.

Imagetwin offers a powerful tool for maintaining the integrity of scientific research. The software analyzes submitted PDF documents or image files (JPG, PNG, GIF, and others) to detect potential issues, presenting results rapidly through a user-friendly web interface. All uploaded content is kept private and secure, with no storage of user data and encrypted communication with servers.

Proofig AI is an advanced platform leveraging artificial intelligence and computer vision to uphold the integrity of scientific research images. It specializes in automatically detecting various forms of image manipulation, duplication, and plagiarism within scientific manuscripts, catering primarily to the life sciences. By offering a comprehensive suite of analytical tools, Proofig AI assists researchers, academic institutions, publishers, and ethics committees in maintaining the highest standards of quality and scientific ethics in publications, thereby safeguarding reputations and mitigating the risks associated with post-publication retractions.

The system operates by analyzing uploaded documents rapidly, providing detailed reports on potential image integrity issues in minutes. It can identify alterations such as cloning, editing, and splicing, as well as reuse of images through scaling, rotation, or flipping. Proofig AI also features plagiarism detection by comparing images against PubMed's vast database and can identify AI-generated microscopy images. The service ensures user data confidentiality by processing all analyses on private, secure servers and not using uploaded manuscripts for model training. This proactive approach to image screening helps save valuable time, reduce human error, and ensure the credibility of scientific findings before they reach peer review and publication.