Project Summary/Abstract The objective of the project is to develop the world-first X-ray dual-energy cone-beam microCT (DECB µCT) dedicated to intraoperative imaging of lumpectomy specimen for improving positive margin identification (PMI) and thus re-call rate for re-surgery in breast conserving surgery (BCS). Breast cancer remains the leading cancer among women in the US, BCS is used as part of the BC treatment for majority of the cases. In BCS, the specimen of negative margin containing tumor is excised with a minimum rim of normal fibroglandular (FG) and adipose tissues, as opposed to removal of the entire breast in mastectomy. Positive margins (PM), i.e., breast cancer (BC) close to, or on, the specimen edge, if missed, can lead to a high re-surgery rate. While the BC-adipose tissue contrast can be high, the BC-FG tissue contrast remains low in standard X-ray imaging. Therefore, it is critically important to develop intraoperative imaging techniques to differentiate BC and FG tissues during the BCS. Our DECB µCT, enabled with innovative algorithms for image reconstruction and analysis, can yield 3D volumetric specimen images with enhanced BC-FG contrast that can directly improve intraoperative PMI and minimize missed PMs. It thus addresses the need to reduce the BCS recall and re- surgery rates. The specific aims of the project are (1) To develop DECB µCT for quantitative lumpectomy specimen imaging; (2a) To acquire highly-sampled DECB µCT of invasive ductal/lobular carcinomas specimens; (2b) To optimize and adapt scan designs of the DECB µCT to BCS clinical workflow; and (3) To evaluate the DECB µCT using patient specimens and to compile a tumor database. The project is built upon our previous success in the development and clinical application of single-energy cone-beam (SECB) µCT, and a key outcome of the project is that we will have established the feasibility of DECB μCT for yielding enhanced BC-FG tissue contrast for improving intraoperative-PMI with AUC of ∼0.95 or higher, thus leading to a reduction of the current BCS re-surgery rate. The first database of quantitative images and their histopathological analysis data of breast malignant tissues will also be created that provides unprecedented amount of detailed information about breast normal and tumor tissues valuable to the development of machine learning (ML)-/deep learning (DL)-based methods for automated intraoperative-PMI in BCS and to the fundamental understanding of BC characteristics for advancing breast cancer research and applications.