Enhancing Privacy: The Utility of Stand-Alone Synthetic CT and MRI for Tumor and Bone Segmentation

2025-06-13

AI Summary

A fundamental tension exists between medical imaging data privacy protection and the data requirements of AI model training—particularly in anatomically complex regions such as the head and neck, where pathological lesions overlap with critical structures, rendering conventional anonymization difficult and synthetic data fidelity and clinical utility insufficiently validated. This work introduces, for the first time, a dual-track evaluation paradigm integrating radiomics analysis and a visual Turing test to systematically assess the substitutability of GAN- and diffusion-based synthetic CT/MRI for head-and-neck cancer (bone segmentation) and glioblastoma (tumor segmentation). Results demonstrate high fidelity of synthetic MRI: tumor Dice similarity coefficient (DSC) = 0.834, bone DSC = 0.841, and strong radiomic feature preservation (Pearson *r* = 0.8784). In contrast, synthetic CT exhibits severely degraded performance (tumor DSC = 0.064, *r* = 0.5461), confirming anatomical complexity as a critical determinant of synthetic data utility.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Privacy
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Security and privacy of machine learning and AI applications
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

AI requires extensive datasets, while medical data is subject to high data protection. Anonymization is essential, but poses a challenge for some regions, such as the head, as identifying structures overlap with regions of clinical interest. Synthetic data offers a potential solution, but studies often lack rigorous evaluation of realism and utility. Therefore, we investigate to what extent synthetic data can replace real data in segmentation tasks. We employed head and neck cancer CT scans and brain glioma MRI scans from two large datasets. Synthetic data were generated using generative adversarial networks and diffusion models. We evaluated the quality of the synthetic data using MAE, MS-SSIM, Radiomics and a Visual Turing Test (VTT) performed by 5 radiologists and their usefulness in segmentation tasks using DSC. Radiomics indicates high fidelity of synthetic MRIs, but fall short in producing highly realistic CT tissue, with correlation coefficient of 0.8784 and 0.5461 for MRI and CT tumors, respectively. DSC results indicate limited utility of synthetic data: tumor segmentation achieved DSC=0.064 on CT and 0.834 on MRI, while bone segmentation a mean DSC=0.841. Relation between DSC and correlation is observed, but is limited by the complexity of the task. VTT results show synthetic CTs' utility, but with limited educational applications. Synthetic data can be used independently for the segmentation task, although limited by the complexity of the structures to segment. Advancing generative models to better tolerate heterogeneous inputs and learn subtle details is essential for enhancing their realism and expanding their application potential.

Problem

• Evaluating synthetic data for medical image segmentation tasks
• Assessing realism and utility of synthetic CT and MRI scans
• Exploring limitations of synthetic data in tumor and bone segmentation

Innovation

• Generative adversarial networks for synthetic data
• Diffusion models enhance data realism
• Visual Turing Test evaluates synthetic quality