Purpose: The objective of our study was to reconsider the usefulness of many characteristics on US for differentiating
benign and malignant masses.
Materials and Methods: From February 2000 through April 2001, we retrospectively evaluated 309 solid
breast nodules (282 patients) on US, which were categorized suspicious malignancy or suggestive malignancy,
prospectively and confirmed by core-needle biopsy or operation. We analyzed the solid nodules,
according to eight known malignant characteristics, which have been described in Stavros et al’s study.
The sensitivity, specificity, positive predictive value, negative predictive value and odds ratio were calculated.
Results: Among 309 lesions diagnosed as suspicious or suggestive malignancy prospectively, 140 lesions (45.31
%) were confirmed as malignant nodules. The mean number of these findings per one nodule was 2.5
(benign nodule: 2.2, malignant nodule: 2.7). Spiculated margin, marked hypoechogenecity, shadowing,
and punctate microcalcifications were statistically significant. Odds ratio of punctate calcifications is the
highest (19.5), followed by marked hypoechogenecity (3.1), spiculated margin (1.9), spiculated, microlobulated
or angular margin (1.7), taller than wide (1.2), additional findings (1), and shadowing (0.6).
Conclusion: Ultrasonographic findings are useful to differentiate malignant from benign solid nodules by that
of Stavros et al’s.