Deep Learning based Radiological Breast Image Reconstruction, Synthesis and Diagnosis

Half-time seminar 

Abstract: Breast cancer is the second leading cause of cancer-related death among women worldwide. Accurate diagnosis is essential for effective treatment, and breast imaging plays a critical role in this process. Various imaging modalities—such as Mammography, Digital Breast Tomosynthesis (DBT), Magnetic Resonance Imaging (MRI), Ultrasound, and Positron Emission Tomography (PET)—provide complementary perspectives for evaluating tumor characteristics. However, interpreting these images requires substantial clinical expertise and can be time-consuming in practice. With the rapid advancement of artificial intelligence, especially deep learning, neural network-based approaches have been widely adopted in many fields. In medical imaging, deep learning has shown great promise in tasks such as image reconstruction, synthesis, and computer-aided diagnosis, supporting clinicians in improving diagnostic accuracy and efficiency. In this half-time report, we explore the application of deep learning techniques in breast imaging, focusing on image reconstruction, synthesis, and computer-aided diagnosis. We explore using deep learning based method on different breast image modalities and evaluate in different datasets.

Opponent: Prof. Robin Strand (Uppsala University)