Neb Duric PhD, Presenter: Officer, Delphinus Medical Technologies

Peter John Littrup MD, Abstract Co-Author: Founder, CryoMedix, LLC Research Grant, Galil Medical Ltd Research Grant, Endo Health Solutions Inc Officer, Delphinus Medical Technologies, Inc

Cuiping Li PhD, Abstract Co-Author: Employee, Delphinus Medical Technologies, Inc

Mark J. Krycia BS, Abstract Co-Author: Nothing to Disclose

We evaluated whether quantitative tissue parameters, obtained from whole-breast ultrasound tomography (UST), could enhance discrimination of breast masses, using automated regions-of-interest (ROI).

This HIPAA compliant, IRB approved trial accrued 100 patients having breast masses identified on palpation, mammography or US, for a UST scan. Pathological correlation was based on biopsy results and standard imaging results (simple cysts). A sequential stack of full-breast, coronal B-mode images as well as quantitative sound speed (SS=m/sec) and attenuation (AT=dB/cm/MHz) images were generated. Identified areas of suspicion were outlined by a radiologist using an ROI ellipse, for which 10 progressive peritumoral and 10 intratumoral ellipses were then automatically generated for evaluation of quantitative trends in relative reflectivity, SS and ATT between the mass and its immediate surrounding peritumoral region.

27 palpable regions had no imaging findings on either standard imaging or UST. Of the remaining 73 subjects, 26 cancers, 16 Fibroadenomas, 20 Cysts and 11 miscellaneous benign histologies were noted. As seen in the figure, the full coronal B-mode image shows a ~1.5 cm hypoechoic mass at 1:00, overlaid by progressive intra-and peritumoral ROI's. Radial trends in the reflectivity, SS and ATT of the ROIs from the tumor epicenter to the furthest peritumoral ellipse show a classic pattern for cysts with both low central reflectivity and attenuation [B], whereas malignant masses showed inverted ATT pattern with the highest values at the tumor center. In combination with a trinary (sharp/indistinct/irregular) tumor margin assessment, cysts and fibroadenomas and cancers showed quantitative separation [C]. This resulted in positive predictive values (PPV) for UST of 93%, compared with standard US BI-RADS of 59%.  

The addition of through transmission provides a substantial increase in the PPV of UST over standard ultrasound BI-RADS criteria. A larger UST mass study of 300 patients is ongoing to validate these results.

Whole breast UST provides quantitative evaluations of a tumor and its immediate surroundings, producing characteristic tissue trends that may aid rapid mass evaluations in larger trials.

Duric, N, Littrup, P, Li, C, Krycia, M, Quantitative Lesion Characterization Using Whole Breast Ultrasound Tomography: Initial Results from an ongoing Clinical Study.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14015889.html