Anatomy-Correlated Breast Imaging and Visual Grading Analysis Using Quantitative Transmission Ultrasound ™ .

Objectives . This study presents correlations between cross-sectional anatomy of human female breasts and Quantitative Transmission (QT) Ultrasound, does discriminate classifier analysis to validate the speed of sound correlations, and does a visual grading analysis comparing QT Ultrasound with mammography. Materials and Methods . Human cadaver breasts were imaged using QT Ultrasound, sectioned, and photographed. Biopsies confirmed microanatomy and areas were correlated with QT Ultrasound images. Measurements were taken in live subjects from QT Ultrasound images and values of speed of sound for each identified anatomical structure were plotted. Finally, a visual grading analysis was performed on images to determine whether radiologists' confidence in identifying breast structures with mammography (XRM) is comparable to QT Ultrasound. Results . QT Ultrasound identified all major anatomical features of the breast, and speed of sound calculations showed specific values for different breast tissues. Using linear discriminant analysis overall accuracy is 91.4%. Using visual grading analysis readers scored the image quality on QT Ultrasound as better than on XRM in 69%-90% of breasts for specific tissues. Conclusions . QT Ultrasound provides accurate anatomic information and high tissue specificity using speed of sound information. Quantitative Transmission Ultrasound can distinguish different types of breast tissue with high resolution and accuracy.
Competing Interests: This study was supported by QT Ultrasound Labs and by Grant 1RO1CA138536 from the National Cancer Institute. Nancy A. Obuchowski is a consultant for and the other authors are employees of QT Ultrasound Labs.