Abstract
Introduction
Radial ultrasound (r-US) or ductosonography is usually applied as an adjunct to meander-like ultrasound (m-US) but rarely as the sole scanning method. Here we compare r-US and m-US with regard to breast lesions detected and interpreted as BI-RADS 3 , i.e. probably benign.
Materials and Methods
Eligible patients received a meander-like and a radial breast ultrasound in random order on the same day by two different examiners. The same type of ultrasound equipment was used but with specific probes.
Results
We performed 1984 dual ultrasound examinations. In 121 BI-RADS 3 lesions, a breast biopsy was performed and the histology of two (1.7%) BI-RADS 3 lesions turned out to be malignant. The specificity for m-US was 95.0%, and 96.6% for r-US. One (0.8%) benign lesion was missed by m-US, whereas r-US missed 2 (1.7%) benign lesions. Each missed lesion was identified by the other scanning method. The mean maximal lesion diameter (ICC 0.82), the mean lesion volume (ICC 0.87), the clock-face localization (κ 0.82) and the mean distance to the skin (ICC 0.77) show excellent, and the mean distance from the lesion to the mammilla (ICC 0.65) good agreement between m-US and r-US. The agreement between m-US and r-US in regard to sonomorphologic criteria ranged from excellent to poor. In 71.9% the lesion was classified as BI-RADS 3 by m-US as well as r-US. The examination time for r-US was significantly shorter than for m-US.
Conclusion
For BI-RADS 3 breast lesions, radial breast ultrasound is an alternative to meander-like ultrasound since the diagnostic accuracy of the two scanning methods is comparable. Notably, patients benefit from a significantly shorter examination time.