Parametric mapping of contrasted ovarian transvaginal sonography

Abstract

The purpose of this study was to assess the accuracy of parametric analysis of transvaginal contrast-enhanced ultrasound (TV-CEUS) for distinguishing benign versus malignant ovarian masses. A total of 48 ovarian masses (37 benign and 11 borderline/malignant) were examined with TV-CEUS (Definity; Lantheus, North Billerica, MA; Philips iU22; Philips Medical Systems, Bothell, WA). Parametric images were created offline with a quantification software (Bracco Suisse SA, Geneva, Switzerland) with map color scales adjusted such that abnormal hemodynamics were represented by the color red and the presence of any red color could be used to differentiate benign and malignant tumors. Using these map color scales, low values of the perfusion parameter were coded in blue, and intermediate values of the perfusion parameter were coded in yellow. Additionally, for each individual color (red, blue, or yellow), a darker shade of that color indicated a higher intensity value. Our study found that the parametric mapping method was considerably more sensitive than standard region of interest (ROI) analysis for the detection of malignant tumors but was also less specific than standard ROI analysis. Parametric mapping allows for stricter cutoff criteria, as hemodynamics are visualized on a finer scale than ROI analyses, and as such, parametric maps are a useful addition to TV-CEUS analysis by allowing ROIs to be limited to areas of the highest malignant potential.

Figure 1

Parametric map of a corpus luteum cyst during peak enhancement (a), time to peak (b), wiAUC (c), woAUC (d), wiwoAUC (e), and woR (f). Map color scales were constructed to show possible tumor neovascularity as shades of red. The majority of the region of interest is depicted in blue colors with the exception of time to peak, which was shown to be non-specific.

Figure 2

Parametric map of a fibrothecoma during peak enhancement (a), time to peak (b), wiAUC (c), woAUC (d), wiwoAUC (e), and woR (f). Consistent with the rest of our analysis, the time to peak was the least specific for detecting malignancy out of all the variables analyzed, as dark red areas are supposed to correlate with abnormal hemodynamics.

Figure 3

Parametric map of a serous borderline tumor during peak enhancement (a), time to peak (b), wiAUC (c), woAUC (d), wiwoAUC (e), and woR (f). All six enhancement kinetic parameters studied revealed marked areas of dark red indicating malignancy, and this was later correlated with histopathology.

Figure 4

Parametric map of a serous adenocarcinoma during peak enhancement (a), time to peak (b), wiAUC (c), woAUC (d), wiwoAUC (e), and woR (f). Areas of tumor neovascularity in red contrast sharply with the cystic portion of the tumor shown in blue.

Figure 5

Parametric map of metastatic breast cancer during peak enhancement (a), time to peak (b), wiAUC (c), woAUC (d), wiwoAUC (e), and woR (f). Multiple areas of tumor neovascularity, as shown as shades of red, are present.

Figure 6

a direct, side-by-side comparison of parametric maps of the above benign fibrothecoma (from Figure 2) and above borderline serous adenocarcinoma (from Figure 3).