Marilyn A. Roubidoux MD, Presenter: Nothing to Disclose

Won Mean Lee BS, Abstract Co-Author: Nothing to Disclose

Lindsay M. Zeeb MD, Abstract Co-Author: Nothing to Disclose

Alexis Virginia Nees MD, Abstract Co-Author: Nothing to Disclose

Lubomir M. Hadjiiski PhD, Abstract Co-Author: Nothing to Disclose

Deborah Oehler Jeffries MD, Abstract Co-Author: Nothing to Disclose

Heang-Ping Chan PhD, Abstract Co-Author: Research collaboration, General Electric Company

Colleen H. Neal MD, Abstract Co-Author: Consultant, Medken LLC

Paul L. Carson PhD, Abstract Co-Author: Research collaboration, General Electric Company Research collaboration, Sonetics Ultrasound, Inc Research collaboration, ZONARE Medical Systems, Inc Research collaboration, Light Age, Inc

David Melville MD, Abstract Co-Author: Nothing to Disclose

To study the contribution of breast ultrasound (US) to digital breast tomosynthesis (DBT), pilot study

IRB approved study. Case set was 37 patients with benign (25) or malignant (12) masses, having DBT (GE medical systems prototype) and clinical ultrasound 2D static images. 31 DBT cases had two orthogonal views  6 had one view. Truth was determined by biopsy, FNA or >2 year follow up imaging. Two blinded breast imaging radiologists were readers, one of 10 years experience ( R1) and one a current breast imaging fellow (R2). The proven masses on DBT were prospectively identified from mammogram and clinical records by a separate experienced breast imaging radiologist and were designated on DBT volumes with a digital box. Each reader separately first viewed the DBT volumes (DBT-alone mode), rated likelihood of malignancy (LOM) using a 13 point scale (normal to >94% LOM) and BIRADS features of the designated masses, and then immediately viewed digital clinical ultrasound images (DBT +US mode) repeating LOM ratings. A graphical user interface recorded ratings and times.

20/37 cases were dense or heterogeneously dense on DBT. A threshold of > 3% LOM was used as an indicator for a biopsy recommendation. Of 50 LOM ratings of the 25 benign cases (both readers ), 43/50 of DBT alone mode and 25/50 of DBT + US mode met threshold for a biopsy recommendation, i.e., = a 42% decrease in false positive (FP) biopsy recommendations with DBT + US mode (p=<.001) . Among LOM ratings of the 12 malignancies, reader R1 had no false negatives (FN) by either DBT alone or DBT+US, and R2 had one FN for DBT+US. In summary, in ROC analysis, AUC: R1: DBT = .936, DBT + US = .947; Az, R2: DBT= .862, DBT + US = .939 p=NS;  2/18 (11%) of margins rated as circumscribed on DBT were malignant.  Mean time viewing and rating of each mass DBT + US was 134 sec ( 64-252 sec) for R1.  

Sensitivity was high for DBT alone. DBT + US mode decreased false positive biopsy recommendations; DBT + US mode improved accuracy over DBT alone for the less experienced reader, but was not statistically significant. Mean time for characterization of a mass with DBT and US for an experienced reader was about 2 minutes.

Although DBT reportedly provides better conspicuity of margins, US may decrease FP biopsies; circumscribed margins seen on DBT may not ensure benign etiology, nor exclude usefulness of US.

Roubidoux, M, Lee, W, Zeeb, L, Nees, A, Hadjiiski, L, Jeffries, D, Chan, H, Neal, C, Carson, P, Melville, D, Role of Ultrasound to Evaluate and Characterize Masses Evident by DBT.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12043490.html