The long-term objective of this project is to test and refine the use BPET-DBT as a tool for co-registered molecular and anatomic imaging to direct biopsy and deliver personalized breast cancer treatment to patients. The integrated BPET-DBT system acquires PET data in the same (mildly) compressed position immediately after the DBT data acquisition, thereby providing co-registered molecular and anatomical images while also allowing easy patient workflow. The BPET scanner is a dedicated, high spatial resolution, time-of-flight (TOF) breast PET scanner producing quantitative images. The DBT image, in addition to providing superior diagnostic anatomical information, also provides an accurate structural context to the PET image (necessary for guiding biopsy and/or surgical approach based on the PET image) and the means to perform accurate attenuation correction (AC) of the PET data. In this project we will evaluate clinical applications of this novel device using selected cohorts of patients designed to test efficacy of the combined device for surgical guidance, monitoring response to therapy, tissue sampling, and further enhancing the quantitative imaging capability of the device both in hardware and software. Our aims are to: (1) demonstrate imaging capability and utility of this unique device, with two specific clinical situations as demonstration (tumor characterization and margin definition, and residual tumor measurement after neo-adjuvant therapy), (2) develop new quantitative image generation techniques and to evaluate them both retroactively and prospectively in clinical studies, and (3) test DBT-guided biopsy using the PET portion of the image set co-registered to the DBT image to direct stereotactic biopsy. The clinical studies in Aims 1 and 3 will: (i) characterize tumor phenotype and determine tumor margins using 18F-fluoroestradiol (18F-FES) as the biomarker and compare to surgical pathology results and SUV measurements with WB PET-CT, (ii) measure 18F-fluorodeoxyglucose (18F-FDG) uptake in small residual tumor prior to surgery in patients undergoing neoadjuvant chemotherapy and compare BPET-DBT and whole-body PET-CT SUV to histopathology findings post-surgery, and (iii) perform targeted breast biopsy using co-registered 18F-FES BPET-DBT images. In Aim 2 we will: (i) evaluate the performance of the existing TOF reconstruction algorithm in mitigating the limited angle artifacts present in the BPET images, (ii) develop and evaluate new BPET reconstruction and AC methods using image- based resolution modeling (IRM) methods as well as Deep-Learning (DL) methods for direct PET image reconstruction and generation of robust DBT images for PET AC, and (iii) develop and test a new high performance PET detector as an extension to existing BPET detectors for improved images. At the end of the study we expect to have demonstrated the role a high resolution, BPET-DBT scanner can play in the future for personalized breast cancer treatment.