Augmented Intelligence for Multimodal Virtual Biopsy in Breast Cancer Using Generative Artificial Intelligence

TL;DR

This work tackles the challenge of breast cancer diagnosis when only standard mammography (FFDM) is available by introducing a multimodal, multi-view virtual biopsy framework that also imputes missing CESM data using CycleGAN-generated images. The pipeline performs two binary tasks—lesion detection and lesion malignancy classification—across FFDM and CESM in CC and MLO views, with late fusion to produce robust predictions. Key findings show that incorporating CESM, real or synthetic, improves diagnostic performance and resilience to missing data, and the authors release a public CESM dataset to advance research. This approach has potential to augment clinical workflows by expanding access to high-accuracy, AI-assisted breast cancer diagnostics while reducing the need for additional contrast-enhanced imaging in every case.

Abstract

Full-Field Digital Mammography (FFDM) is the primary imaging modality for routine breast cancer screening; however, its effectiveness is limited in patients with dense breast tissue or fibrocystic conditions. Contrast-Enhanced Spectral Mammography (CESM), a second-level imaging technique, offers enhanced accuracy in tumor detection. Nonetheless, its application is restricted due to higher radiation exposure, the use of contrast agents, and limited accessibility. As a result, CESM is typically reserved for select cases, leaving many patients to rely solely on FFDM despite the superior diagnostic performance of CESM. While biopsy remains the gold standard for definitive diagnosis, it is an invasive procedure that can cause discomfort for patients. We introduce a multimodal, multi-view deep learning approach for virtual biopsy, integrating FFDM and CESM modalities in craniocaudal and mediolateral oblique views to classify lesions as malignant or benign. To address the challenge of missing CESM data, we leverage generative artificial intelligence to impute CESM images from FFDM scans. Experimental results demonstrate that incorporating the CESM modality is crucial to enhance the performance of virtual biopsy. When real CESM data is missing, synthetic CESM images proved effective, outperforming the use of FFDM alone, particularly in multimodal configurations that combine FFDM and CESM modalities. The proposed approach has the potential to improve diagnostic workflows, providing clinicians with augmented intelligence tools to improve diagnostic accuracy and patient care. Additionally, as a contribution to the research community, we publicly release the dataset used in our experiments, facilitating further advancements in this field.

Figures (2)

• Figure 1: Schematic representation of the methodology.
• Figure 2: ttt