XDeepScan
AI and cybersecurity project for detecting fraudulent alterations and deepfakes in diagnostic medical images.
Industrial Research
XDeepScan
AI and cybersecurity project for detecting fraudulent alterations and deepfakes in diagnostic medical images.
Overview
XDeepScan targets the integrity of medical imaging data. The project develops AI methods to detect fraudulent alterations and deepfakes in diagnostic images, protecting clinical data reliability and supporting cybersecurity in healthcare.
The development plan combines discriminative models, generative models, latent diffusion techniques, and explainable AI to identify and interpret image tampering across CT, MR, and other diagnostic imaging scenarios.
Research Directions
- Detection of manipulated CT, MR, and diagnostic images.
- Discriminative and generative AI models for medical image tamper detection.
- Explainable AI methods for interpreting detected alterations in clinical images.