|
KMID : 1114620110080020115
|
|
Journal of the Korean Society for Breast Screening
2011 Volume.8 No. 2 p.115 ~ p.120
|
|
|
Differentiation of Breast Mass Using Automated Breast US: Application of US BI-RADS Lexicon
|
|
Chang Jung-Min
Yi Ann
Koo Hye-Ryoung
|
|
|
Abstract
|
|
|
Purpose: To assess the agreement of US findings between ABUS and HHUS using American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) lexicon and final assessment category, and to evaluate the diagnostic performance of automated breast US (ABUS) and hand held breast US(HHUS) for the differentiation of breast masses.
Materials and Methods: In January 2008, among patients who are scheduled to perform breast biopsy or surgery, 92 breast masses in 89 patients who had both HHUS and ABUS images of breast masses were included in our study population. Two breast radiologists retrospectively evaluated breast masses using HHUS and ABUS images and recorded the findings and final assessment using BI-RADS lexicon and assessment category, and they recorded possibility of malignancy on a five point scale. Agreements of each BI-RADS descriptors and final assessment between HHUS and ABUS were evaluated. In addition, Receiver operating characteristic (ROC) curve analysis for the HHUS and ABUS were performed to evaluate
diagnostic performance.
Results: Ninety- two breast masses in 89 patients were identified, and among them, there were 31 malignancies and 61 benign lesions. Moderate agreement was obtained for lesion shape, margin, and fair agreement was obtained for orientation, echo pattern, posterior acoustic features and surrounding tissue. Slight agreement was noted for lesion boundary. However, substantial agreement was noted for final assessment. The area under ROC values were 0.853 (95% confidence interval: 0.764, 0.918) for ABUS and 0.838 (95% confidence interval: 0.746, 0.907) for HHUS, which was not significantly different (p = 0.769).
Conclusion: Even though variable degrees of agreements were noted for lesion descriptors, substantial agreement was noted for BI-RADS final assessment. There were no significant differences for diagnostic performance between ABUS and HHUS.
|
|
|
KEYWORD
|
|
|
Hand held US, Automated US
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|
|