Breast Imaging - Research Summary
Dedicated CT Breast Imaging
One of the fundamental limitations with conventional mammography is the structural overlap problem, thereby making the detection of malignant lesions in mammograms very difficult. One solution to this overlap problem is tomographic imaging. The purpose of this research is to investigate the feasibility of dedicated CT breast imaging for the early stage detection and diagnosis of breast cancer. We have fabricated a bench-top, prototype CT breast imager to better understand this novel technology. Using breast phantoms and fresh surgical mastectomy specimens, optimal operating parameters for such a CT system are being evaluated. We have used this bench-top system to develop computer simulation software, breast object models, and to generally gain an understanding of the many problems and challenges that dedicated CT breast imaging systems must solve before they can become a routine clinical tool. Several aspects of breast CT including radiation dose, imaging technique optimization, and detection studies for lesions and microcalcifications with breast CT have been studied.
We have recently completed a 150 patient clinical trial using an experimental breast CT system. These patients were suspected of having breast cancer (BIRADS 4 and 5 category) and accompanied histopathology data is thus available. This data is currently being analyzed.
Illustration of dedicated breast CT system. Breast hangs prone through a hole in the table and x-ray tube and detector rotate around the breast collecting projections.
Development of a Photon Counting Detector for Breast CT
For over a century now, clinical x-ray imaging devices have operated in energy integrating mode, whereby images are formed by integrating x-ray events over a finite acquisition time. For a number of reasons, the performance of energy integrating detectors are sub-optimal for use in CT imaging of the breast. It is expected that the next generation of x-ray detectors for digital radiography and CT will have the capability of counting individually measured photons and recording their energy.
The overall goal of this proposal is to develop and validate an innovative detector system and imaging methodology for improved performance in dedicated CT imaging of the breast. The CT breast imaging system that we propose herein will use direct conversion, small pixel semiconductor detectors operating in a photon counting mode, acquiring projections in a helical CT geometry. Simulation and modeling will be used to examine technical design specifications for such a photon-counting x-ray CT system, and a prototype will be constructed and validated by imaging fresh surgical mastectomy specimens. The deliverables for this project will be a validated photon-counting CT detector system and imaging protocols that can be implemented for future clinical trials
Illustration of photon counting breast CT detector
Optimization of energy windows for dual-energy imaging with energy weighting.
Three materials were studied, invasive ductal carcinoma, microcalcification, and iodine.
Breast tomosynthesis is a limited-angle tomographic breast imaging method. Tomosynthesis systems use similar patient positioning as with conventional mammography with the breast compressed. Typically, low radiation projection images are acquired over an angular range of 30-60 degrees, followed by image reconstruction. We are studying several aspects of breast tomosynthesis including optimization of system parameters, a new variable dose acquisition scheme, and performance of a penalized maximum likelihood reconstruction approach.
Specimen holder used for CT imaging of mastectomy specimens (Left). These CT reconstructions are used to generate compressed voxelized breast phantoms (A,B, and C – orthogonal slices through one compressed breast phantom. These phantoms are then used to simulate realistic mammograms, or tomosynthesis projections.
Dedicated Breast PET
Over the past several years it has become clear that PET imaging with fluro-deoxyglucose (FDG) or other F18 based radiopharmaceuticals being developed for probing breast cancer can play an important role in the detection and diagnosis of breast cancer. Many encouraging studies using whole-body PET systems to image breast cancer have been reported, however, it is now evident that smaller PET systems dedicated to imaging the breast have substantial advantages over imaging of the breast with whole-body PET.
Our research in this important area includes simulation studies to explore the potential benefits of breast PET scanners with fast detectors, as well as evaluating potential improvements with time-of-flight PET. In addition, we are developing statistical based iterative reconstruction methods using a Monte Carlo based system matrix. This approach allows for breast PET systems to have small ring geometries, thereby maximizing sensitivity.
Results of GATE simulation of breast PET using a ring detector. Activity outside of the feild of view (FOV) is simulated with the illustrated geometry. Shown are reconstructions with and without activity outside FOV, showing the degradation in image quality with activity outside FOV. This is predominantly due to the large effects of randoms, especially in the posterior breast.