Breast Radiography Image Dimension Generation and Enhancement

ongoing

BRIDGE (Breast Radiography Image Dimension Generation and Enhancement) is an innovative research project aimed at revolutionising breast cancer screening in low-resource settings. Despite being the most frequently diagnosed cancer in women worldwide, with alarming mortality rates, access to advanced 3D imaging technologies remains limited in many regions. BRIDGE proposes to bridge this gap by developing a novel artificial intelligence (AI) framework to generate synthetic 3D mammography images and/or magnetic resonance images (MRI) from standard 2D mammograms. Further, it aims to detect the presence of anomalies in the segmented 3D images. Leveraging state-of-the-art deep learning techniques, including Swin Transformers and Generative Adversarial Networks, BRIDGE will create a robust system capable of producing clinically actionable 3D breast images. This approach will enhance diagnostic capabilities in resource-limited settings without the need for expensive 3D imaging equipment, potentially improving early detection rates and patient outcomes. Additionally, the GAN method is utilized to detect anomalies in 3D images. The project's interdisciplinary methodology integrates expertise from AI, medical imaging, clinical practice, and ethics. It focuses on four key areas: cross-modality synthesis, clinical applicability, resource optimisation, and ethical AI development. By collaborating with radiologists and oncologists of low-income countries, BRIDGE ensures that the generated images not only are visually accurate but also enhance clinical decision-making. BRIDGE aims to validate its approach through rigorous clinical trials and cost-effectiveness analyses. The project's success could significantly narrow the global disparity in breast cancer diagnostics, contributing to improved healthcare equity worldwide. Moreover, this research will advance the field of synthetic medical imaging, potentially opening new avenues for AI applications in healthcare.