Jacques Souquet

Jacques SOUQUET

Founder and CEO, SuperSonic Imagine

キーワード :

【Introduction】
The purpose of this work was to determine the reproducibility of ShearWave^TM Elastography （SWE） and to determine whether SWE added to B-mode breast US can improve the specificity of lesion characterization without a loss of sensitivity.
【Materials & Methods】
939 BI-RADS^® 2 to 5 breast lesions from 939 women scheduled for breast US were recruited across 16 sites in the United States and Europe. Histology or a definitive follow-up diagnosis was available for 910 lesions, including 313 （34.4%） malignancies. The reproducibility of the SWE images and measurements was assessed on 3 consecutive acquisitions. The best three-variable model, including the BI-RADS assessment and 2 SWE features, was determined using a multivariate regression analysis and was used for reclassification.
【Results】
The mean patient age was 52.1+/-14.8y （21.2-95.3） and mean lesion size 13.1+/-7.6mm （2.6-51.3）. The lesion size on SWE and elasticity showed almost perfect reproducibility （IOR=0.95 and 0.91 respectively）. The AUC of the BI-RADS assessment alone （0.72） significantly improved with the addition of any SWE features. Adding the SWE homogeneity and mean elasticity value （0.89） or homogeneity and lesion-to-fat ratio （0.88） gave the best models. When reclassifying BI-RADS 3 and 4 lesions showing 2 SWE signs of benignity/malignancy, the specificity increased by at least 20 points, while the sensitivity and NPV decreased by less than 4 points. 151 benign lesions were correctly downgraded from BI-RADS^® 4 to 3 and 8 cancers were upgraded from 3 to 4.
【Conclusion】
SWE is highly reproducible. Adding two SWE features to the BI-RADS test significantly improved the AUC and greatly improved the specificity without compromising the sensitivity.