WOO Oknee1, SHIN Hey Seon1, HONG Eun Ae1, CHO Kyu Ran2, SEO Bo Kyoung3
Oknee WOO1, Hey Seon SHIN1, Eun Ae HONG1, Kyu Ran CHO2, Bo Kyoung SEO3
1Department of Radiology, Korea University Guro Hospital, 2Department of Radiology, Korea University Anam Hospital, 3Department of Radiology, Korea University Ansan Hospital
【Objective】 To estimate reproducibility and evaluate diagnostic performance of shear wave elastography （SWE） in breast solid mass. 【Methods】 SWE with conventional grayscale ultrasound were performed in 218 women with 264 solid breast masses （152 benign, 112 malignant）. Two breast radiologists with 5 and 10 years of experience in breast ultrasound independently evaluated each breast lesion. After all the examinations were completed, SWE images were reviewed by the two readers for color overlay pattern （COP） classification and quantitative elasticity value measurement. Reproducibility of COP and quantitative elasticity were analyzed by weighted value and intraclass correlation coefficient （ICC）. Diagnostic performance of COP, maximum elasticity values （Emax）, and Emax combined with COP were calculated by analysis of receiver operating characteristics （ROC） curves. 【Results】 Interobserver agreement of COP was almost perfect （=0.908）. ICC of Emax （0.89） was highest among elasticity values. The area under the ROC curve （AUC value） of COP and Emax was 0.953, 0.966 in reader 1 and 0.948, 0.958 in reader 2 respectively. With estimated optimal cut off value 44.1 for Emax and between pattern 2 and 3 for COP, AUC value of COP （0.954） was significantly higher than that of Emax （0.915） （p=0.002）. AUC value of Emax combined with COP （0.957） was not significantly higher than AUC value of COP （p=0.098）. 【Conclusions】 SWE color overlay pattern and Emax are highly reproducible and COP is more reliable parameter of solid breast mass evaluation, showing better or similar diagnostic performance than that of Emax and Emax combined with COP.