The traditional methods for breast cancer screening (physical exam and mammography) suffer from poor specificity and a difficulty in detecting lesions in mammographically dense breasts. Contrast enhanced magnetic resonance imaging of the breast has been suggested as a method to improve the specificity of breast imaging by differentiating benign from malignant lesions within the breast. The presence of contrast enhancement alone is not specific for breast cancer. Techniques to perform contrast enhanced MR images of the breast with 300 x 600 micron in plane resolution have been developed by this group. The correlation between high resolution contrast enhanced breast MR images and pathology has been performed in order to identify those imaging features that correlate with pathologic diagnosis. The hope is to ultimately reduce the number of false positive breast biopsies. An interdisciplinary team of radiologists, pathologists, surgeons, and statistician has successfully identified architectural features that are very predictive of benign or malignant breast pathology. Research described in the current proposal extends the investigation of architectural features of high resolution breast MR images in order to refine and test interpretation models. In addition, quantitative time course kinetics will be obtained from dynamically acquired high resolution 3D breast images in order to define optical imaging time and further improve predictive models. The limits of resolution will be investigated by performing in vivo MR microscopic (150 micron in plane resolution) images of the breast and correlating them with pathologic sections. Finally, a cost and outcomes model will be established to assess the clinical impact of high resolution breast MRI in the care of patients suspected of having breast cancer.
Showing the most recent 10 out of 15 publications
