Lesion-Aware Generative Artificial Intelligence for Virtual Contrast-Enhanced Mammography in Breast Cancer

2025-05-05

AI Summary

To address the high radiation dose and safety risks associated with iodinated contrast agents in contrast-enhanced spectral mammography (CESM), this work proposes a contrast-free, low-dose virtual enhancement imaging method. We introduce Seg-CycleGAN—a CycleGAN-based architecture augmented with a U-Net–guided lesion segmentation module—and propose a novel lesion-focused local L1/SSIM loss combined with multi-scale perceptual constraints to significantly improve synthesis fidelity in lesion regions. On the CESM@UCBM dataset, our method achieves superior PSNR and SSIM compared to existing baselines, while maintaining competitive MSE and Visual Information Fidelity (VIF) scores. Qualitative evaluation confirms that the synthesized images closely replicate the lesion enhancement characteristics of real CESM scans. This approach establishes a new paradigm for safe, accurate, and radiation-efficient breast cancer screening.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Contrast-Enhanced Spectral Mammography (CESM) is a dual-energy mammographic technique that improves lesion visibility through the administration of an iodinated contrast agent. It acquires both a low-energy image, comparable to standard mammography, and a high-energy image, which are then combined to produce a dual-energy subtracted image highlighting lesion contrast enhancement. While CESM offers superior diagnostic accuracy compared to standard mammography, its use entails higher radiation exposure and potential side effects associated with the contrast medium. To address these limitations, we propose Seg-CycleGAN, a generative deep learning framework for Virtual Contrast Enhancement in CESM. The model synthesizes high-fidelity dual-energy subtracted images from low-energy images, leveraging lesion segmentation maps to guide the generative process and improve lesion reconstruction. Building upon the standard CycleGAN architecture, Seg-CycleGAN introduces localized loss terms focused on lesion areas, enhancing the synthesis of diagnostically relevant regions. Experiments on the CESM@UCBM dataset demonstrate that Seg-CycleGAN outperforms the baseline in terms of PSNR and SSIM, while maintaining competitive MSE and VIF. Qualitative evaluations further confirm improved lesion fidelity in the generated images. These results suggest that segmentation-aware generative models offer a viable pathway toward contrast-free CESM alternatives.

Problem

• Reducing radiation exposure in Contrast-Enhanced Spectral Mammography
• Eliminating contrast medium side effects in mammography
• Generating virtual contrast-enhanced images without contrast agents

Innovation

• Generative AI synthesizes dual-energy mammography images
• Lesion segmentation guides image reconstruction accuracy
• Localized loss terms enhance diagnostic region synthesis