Predictive detection areas for identifying additional MRI-detected breast lesions on second-look ultrasonography.

PURPOSE: Identifying an additional MRI-detected breast lesion on second-look ultrasonography (US) is technically challenging because of lesion displacement with the patient's position change. The aim of this study is to help identify MRI-detected lesions on second-look US by developing a probing area, called "the predictive detection area" (PDA), and by assessing the PDA.

METHODS: We measured the nipple-to-lesion distances (NLDs) for 16 breast lesions on prone- and supine-position MRI sets and calculated the difference and angle between the two NLD vectors, representing the lesion displacement. The minimum and maximum differences and angles were chosen to form the PDA. Another 22 breast lesions, detected in the prone MRI, were identified on US by probing the PDA to evaluate the probability of existence.

RESULTS: The width between the minimum and maximum differences in two NLDs and the angle to form the PDA for the upper-inner, upper-outer, and lower-outer quadrants were 23.0 mm and 95.0°, 29.0 mm and 41.0°, and 18.0 mm and 17.0°, respectively. The respective probabilities of existence were 100, 80, and 100%.

CONCLUSIONS: The PDA had a high probability of existence and was acceptably accurate; therefore, the PDA in a second-look US has the potential to help operators to quickly identify additional MRI-detected lesions.