Breast cancer screening using mammography is current practice, however other imaging and non-imaging procedures are being studied for earlier detection.  X-ray CT has been largely dismissed for breast cancer screening, due to misconceptions about radiation dose.   Using computer simulation and physical experiments, we have shown that the radiation dose levels for high quality breast CT are equivalent to or lower than mammography.  Because of the higher contrast resolution that is the hallmark of CT, and due to the dramatic reduction (100 fold) of background anatomical “structure” noise that tomography offers, breast CT will almost certainly lead to earlier detection of breast cancer than mammography.


We believe that breast CT will be capable of far earlier detection than x-ray mammography (digital or screen-film), using average glandular doses comparable to or less than mammography.  The hypothesis, expressed as the null hypothesis, is that breast CT and mammography have identical detection performance.


The overall objectives of this project are to study the potential of dedicated breast CT for the early detection of breast cancer.  A breast CT scanner will be designed (Aim #1) and fabricated (Aim #2) in the physics research laboratory.  Breast cancer advocates will participate in the design aspects of the scanner pertaining to patient positioning and in the assessment of breast dimensions pertinent to scanner design.  A 3D viewing workstation will be developed (Aim #3), and the breast CT scanner will be thoroughly evaluated for radiation dose, safety, and imaging quality (Aim #4).  After assessment on advocate volunteers (Aim #5), the breast CT scanner will be moved to the breast screening clinic and a cohort (~50) of women with suspicious mammographic findings (BIRADS 4 and 5) will be imaged (Aim #6) in a phase II trial.  In addition to subjective image comparison, quantitative observer performance testing will be used to accept or reject the null hypothesis (Aim #7). 


The breast CT system will scan the pendulant breast, hanging from a hole in a shielded patient table.  A low power x-ray tube and flat panel detector system will be mounted onto a commercially available gantry ring that will rotate in the horizontal plane below the patient.  Cone beam geometry will be used, which will require one ~375º rotation of the gantry for all CT images.  No compression will be necessary, but some mechanical elongation of the breast will be used to separate tissues and for immobilization.  In addition to the development of scanner hardware and reconstruction software, an image review workstation will be developed for viewing the volume data set.  Once built and tested, the scanner will be used to scan approximately 50 women with suspicious mammographic findings.  This initial data will provide the necessary initial data for designing a larger clinical trial, beyond the 3 year scope of this proposal.

Impact on breast cancer:

Based on preliminary studies (CT images of cadaver breasts), it is likely that breast CT will detect much smaller breast cancers than mammography using the same or less radiation doses.  Detection of cancer at 5 mm instead of the median lesion size of 11 mm with mammography represents a 0.93 year earlier detection.  Detection of 3 mm lesions represents a 1.5 year advance in detection.  The inherently 3D nature of breast CT will allow the breast scanner to be used to accurately guide cancer excision, resulting in breast cancer cure for the majority of women.  Breast CT has the potential to dramatically reduce breast cancer mortality and morbidity by detecting cancer long before it becomes an invasive disease.