Veritas is a forensic image analysis tool, specifically tailored for scientific research.

Veritas is a dedicated analytical tool for the forensic analysis of scientific images, specially designed for the needs of research institutions, scientific publishers, and ethics committees. Veritas enables reviewers to establish a high-confidence link between a physical recording device (e.g., a microscope camera) and the image data used or published in a scientific study, providing powerful evidence to verify scientific integrity.

A Modern Approach to Integrity Assurance

At the heart of Veritas is a two-step process that empowers reviewers to establish a fast, direct, and verifiable link between a physical recording device and the image data used in a scientific study. This workflow is designed for integration into existing laboratory and publication pipelines, shifting forensic capability from a centralized bottleneck to an accessible, in-house decision support tool.

  1. Lab Integrity Certification (“Device Enrollment”)

To certify a device in a laboratory as a reliable source, the examiner initiates a Lab Integrity Certification process. This involves taking several reference images with the device’s camera (e.g., a Western blot scanner or a fluorescence microscope camera). Veritas uses these reference images to calculate the camera’s stable PRNU fingerprint and stores it securely. The device is then certified and registered in the system as an authentic source of scientific image data. This process creates a foundational, tamper-evident record of a lab’s imaging equipment, enhancing its long-term credibility.

  1. Analysis and Confidence Scoring of Image Data

The examiner can then upload any images from a scientific publication or data set. Veritas compares each file with the previously created fingerprints of the registered laboratory devices. The result is not a simple binary value, but a “confidence score and similarity index” that quantifies the likelihood that the image originated from a specific device. This nuanced output provides a robust scientific basis for further investigation, flagging images that deviate from expected patterns for expert review. This process not only enables the verification of research data but also reveals complex manipulation strategies that might otherwise go unnoticed.

Core Applications: Empowering Scientific Integrity

  1. Validation of Publications and Detection of Manipulation
  • Scenario: A scientific journal receives a submission with impressive results. The illustrations show Western blots and microscopy images.
  • Use of Veritas: The reviewer registers the cameras of the devices used by the author’s laboratory. Veritas analyzes all submitted images and provides a confidence score for each. Images with high scores matching the registered device are considered authentic. Images with low scores or inconsistent PRNU patterns are flagged as anomalies, indicating potential manipulation – e.g., copying and pasting bands from other experiments or using stock photos. This provides clear, actionable evidence for the editorial team.
  1. Forensic Detection of “Paper Mills” and Author Networks
  • Scenario: There is suspicion that several publications by different authors from different institutes are based on the same falsified data sets.
  • Use of Veritas: This is where Veritas excels. Reviewers analyze images from the various publications. Even without the original device, Veritas can determine that images in Paper A (Author X) and Paper B (Author Y) share an identical underlying PRNU fingerprint, traceable to a single, unregistered source. This is powerful forensic evidence of a common data source and proves the activity of a “paper mill.”
  1. Retrospective Forensic Assessment
  • Scenario: A university receives a tip-off about potential data fraud in a doctoral thesis published years ago. The original devices may no longer be available.
  • Use of Veritas: The images contained in the thesis are analyzed for internal PRNU consistency. Veritas can determine whether all images that supposedly originate from the same series of experiments share a common fingerprint. Inconsistencies in the fingerprint are a strong indicator of subsequent image manipulation and can form the basis for a formal investigation into scientific misrepresentation.

Decentralized Integrity Assurance at the Source

Veritas is designed as a “decision support tool” for decentralized use, enabling forensic image analysis to be carried out directly, quickly, and easily at research institutions or publishing houses. Unlike costly external expert opinions, Veritas provides immediate analysis, resulting in significant time savings. It serves as an efficient filter for routine checks, empowering integrity officers, editors, and reviewers with powerful evidence while leaving the final conclusion to human expert judgment. This scalable model makes quality assurance more agile and effective, integrating seamlessly into modern scientific workflows.

Designed for Modern Research Ecosystems

Veritas is built to fit into the way science is actually done. We provide tools and APIs for integration with existing systems:

  • Image Analysis Plugins: Work within the environments you already use, with plugins for popular scientific image software.
  • LIMS Connectivity: Link your Laboratory Information Management System to maintain a live, certified inventory of your imaging equipment.
  • Submission System Integration: Publishers can integrate Veritas analysis directly into their manuscript submission platforms for automated preliminary checks.

You want to learn more about Veritas?

CONTACT US!