Purpose: The purpose of the study was to analyze sonographic criteria for the differentiation of benign and malignant breast lesions using real-time spatial compound imaging (CT) in combination with adaptive image processing (XRES).
Materials and methods: In a retrospective analysis of 460 patients, the sonographic criteria: shape, orientation, margin, echo pattern and posterior acoustic features were determined using CT and XRES. All investigations were performed using a 12 MHz linear transducer. The findings were classified according to the DEGUM criteria analogous to BIRADS and were histologically confirmed by core needle or vacuum biopsy. Statistical analysis was performed using a Chi-square test, logistic uni- and multivariate regression analysis and an ROC-curve analysis to detect the false-positive rate.
Results: All investigated diagnostic criteria were significant in the descriptive analysis (Chi-squared). The multivariate analysis showed that the criteria of irregular shape versus round, and not circumscribed margin versus circumscribed margin as well as the posterior acoustic features of enhanced versus reduced have a significant influence on the prediction of a malignant finding. The univariate analysis also showed a statistical significance using the indifferent and not parallel orientation versus parallel. To attain a detection rate of > 95 %, a false-positive rate of 60 % must be expected.
Conclusion: This study shows that when using CT in combination with XRES, the analyzed sonographic criteria for the differentiation of benign and malignant breast lesions are still of diagnostic value. In particular, the shape, margin and posterior sonographic features are important.
© Georg Thieme Verlag KG Stuttgart · New York.