Sci-Fri PM Imaging-06: Registered Digital Stereotactic Mammography and 3D-Ultrasound for Breast Biopsy Guidance

Abstract

Large core needle biopsy is a common procedure used to obtain histological samples when a suspicious lesion is detected in diagnostic breast images. The procedure is typically performed using image guidance, with ultrasound (US) and stereotactic mammography (SM) being the most common modalities used. Each of these imaging methods, however, has limitations that impact their clinical utility. For example, some breast structures are not visible in ultrasound. Mammography provides better visualization of features such as microcalcifications, but does not support real-time imaging. A prototype device combining the advantages of digital SM and 3D-US with computer-aided needle guidance was developed at our centre. The objective of this work was to determine the position of biopsy targets with an error of less than 0.5 mm. A methodology was first developed to calibrate the SM system. Then, by imaging a set of the same physical points identifiable in both SM and 3DUS images, the two modalities were registered. In both cases, the target registration error (TRE) was calculated to quantify the error in determining the position of points imaged within the breast. For locating points in the SM images alone, the TRE was found to be 0.35 mm. The TRE found in registering the two modalities was 0.37 mm. The TRE value is clinically relevant as it indicates the position error associated with selecting an arbitrary image point for biopsy. When compared to the typical breast biopsy needle diameter of 2.1 mm, the calculated TRE from the two imaging modalities was sufficiently small.