Towards Trustworthy Diagnostic AI: Mitigating Hallucination in Deep Learning-Based Medical Image Restoration

Deep Learning (DL) has shown impressive promise for medical image restoration, offering improved image quality for precise diagnosis. However, these models—especially generative ones—are vulnerable to a critical failure mode called hallucination, in which they erase subtle pathologies or create plausible but nonexistent anatomical structures. Due to the possibility of false positives and false negatives, this phenomenon seriously jeopardizes patient safety and undermines clinical confidence. In this work, we suggest a thorough framework to reduce hallucinations and develop reliable diagnostic AI. In order to limit the solution space to data-consistent outputs, we first formulate the restoration task within a physics-informed architecture that explicitly incorporates the imaging forward model. Additionally, we present a brand-new uncertainty quantification module that creates a pixel-by-pixel confidence map, enabling medical professionals to see possible hallucination regions. Additionally, we support a hybrid loss function that strikes a balance between strict fidelity to the input data and perceptual quality. Our framework is tested on a variety of clinical datasets, such as fast MRI and low-dose CT. We show that, when compared to state-of-the-art baselines, our method dramatically reduces hallucination artifacts as measured by both conventional metrics (PSNR, SSIM) and a recently proposed Faithfulness Score. Importantly, a reader study involving three board-certified radiologists verifies that images restored using our technique increase interpretative confidence while maintaining diagnostic accuracy. This work demonstrates that creating dependable and clinically useful AI-based restoration tools requires a comprehensive approach that combines uncertainty-aware visualization with model-centric constraints.

Medical Image Restoration, Hallucination Mitigation, Physics-Informed Deep Learning, Uncertainty Quantification, Trustworthy AI.

1. Wang, G., Ye, J. C., & De Man, B. (2020). Deep learning for tomographic image reconstruction. Nature Machine Intelligence, 2(12), 737-748. [Google Scholar] [Crossref]

2. Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., & Sánchez, C. I. (2017). A survey on deep learning in medical image analysis. Medical Image Analysis, 42, 60-88. [Google Scholar] [Crossref]

3. Yang, Q., Yan, P., Zhang, Y., Yu, H., Shi, Y., Mou, X., ... & Wang, G. (2018). Low-dose CT image denoising using a generative adversarial network with Wasserstein distance and perceptual loss. IEEE Transactions on Medical Imaging, 37(6), 1348-1357. [Google Scholar] [Crossref]

4. Antun, V., Renna, F., Poon, C., Adcock, B., & Hansen, A. C. (2020). On instabilities of deep learning in image reconstruction and the potential costs of AI. Proceedings of the National Academy of Sciences, 117(48), 30088-30095. [Google Scholar] [Crossref]

5. Cohen, J. P., Brooks, R., En, S., Zucker, E., Pareek, A., Lungren, M. P., & Bertrand, H. (2021). Gifs planation via latent shift: A simple autoencoder approach to counterfactual generation for chest xrays. In Proceedings of the Machine Learning for Health (pp. 74-95). PMLR. [Google Scholar] [Crossref]

6. Finlayson, S. G., Bowers, J. D., Ito, J., Zittrain, J. L., Beam, A. L., & Kohane, I. S. (2019). Adversarial attacks on medical machine learning. Science, 363(6433), 1287-1289. [Google Scholar] [Crossref]

7. Kendall, A., & Gal, Y. (2017). What uncertainties do we need in Bayesian deep learning for computer vision?. Advances in Neural Information Processing Systems, 30. [Google Scholar] [Crossref]

8. Leibig, C., Allken, V., Ayhan, M. S., Berens, P., & Wahl, S. (2017). Leveraging uncertainty information from deep neural networks for disease detection. Scientific Reports, 7(1), 17816. [Google Scholar] [Crossref]

9. Roy, A. G., Conjeti, S., Navab, N., & Wachinger, C. (2019). Bayesian quicknat: Model uncertainty in deep whole-brain segmentation for structure-wise quality control. NeuroImage, 195, 11-22. [Google Scholar] [Crossref]

10. Hammernik, K., Klatzer, T., Kobler, E., Recht, M. P., Sodickson, D. K., Pock, T., & Knoll, F. (2018). Learning a variational network for reconstruction of accelerated MRI data. Magnetic Resonance in Medicine, 79(6), 3055-3071. [Google Scholar] [Crossref]

11. Zhu, B., Liu, J. Z., Cauley, S. F., Rosen, B. R., & Rosen, M. S. (2018). Image reconstruction by domain-transform manifold learning. Nature, 555(7697), 487-492. [Google Scholar] [Crossref]

12. Sung, M., Kim, H. Z., Hally, D., & Hwang, S. J. (2021). How to trust unlabeled data? instance credibility inference for few-shot learning. Advances in Neural Information Processing Systems, 34. [Google Scholar] [Crossref]

13. Saharia, C., Ho, J., Chan, W., Salimans, T., Fleet, D. J., & Norouzi, M. (2022). Image super resolution via iterative refinement. IEEE Transactions on Pattern Analysis and Machine Intelligence. [Google Scholar] [Crossref]

14. Wolterink, J. M., Leiner, T., Viergever, M. A., & Išgum, I. (2017). Generative adversarial networks for noise reduction in low-dose CT. IEEE Transactions on Medical Imaging, 36(12), 2536-2545. [Google Scholar] [Crossref]

15. Lugmayr, A., Danelljan, M., Romero, A., Yu, F., Timofte, R., & Van Gool, L. (2022). Repaint: Inpainting using denoising diffusion probabilistic models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 11461-11471). [Google Scholar] [Crossref]

16. Maier-Hein, L., Eisenmann, M., Reinke, A., Onogur, S., Stankovic, M., Scholz, P., ... & Full, P. M. (2018). Why rankings of biomedical image analysis competitions should be interpreted with care. Nature Communications, 9(1), 5217. [Google Scholar] [Crossref]

17. Varoquaux, G., & Cheplygina, V. (2022). Machine learning for medical imaging: methodological failures and recommendations for the future. NPJ Digital Medicine, 5(1), 48. [Google Scholar] [Crossref]

• Measuring Waste of Patient Time in Health Care at Non-Digitized Hospital: An Observational Study in Bangabandhu Sheikh Mujib Medical University, Bangladesh
• Reaffirming Clinical Confidence in Atorvastatin Therapy: A Digital Outreach Case Study from Tamil Nadu, India
• Clinical Manifestations and Therapeutic Response in a Patient with Hypothyroidism: A Case Report
• Eranda (Ricinus Communis) In Gridhrasi (Sciatica): Classical Rationale, Pharmacology and Clinical Evidence- A Narrative Literature Review
• Magnetotherapy in Pain Management: Mechanisms, Clinical Applications, and Future Perspectives – A Review