OBJECTIVE: The objective of our study was to determine the usefulness of recognizing the continuity of the gonadal veins to the pelvic mass to differentiate ovarian versus uterine origin on MDCT in females with a large pelvic mass. MATERIALS AND METHODS: Two radiologists interpreted the MDCT images obtained on a monitor, using paging methods, in 86 female patients with a large pelvic mass (> 8 cm) and 40 patients without an abdominopelvic mass as control subjects. The following issues were recorded using a 5-point scale: visualization of gonadal veins and origin determination based on anatomic continuity. Receiver operating characteristic (ROC) curve analysis was performed, and the interobserver differences were checked with kappa statistics. The maximum diameters of the gonadal veins were also measured. With consensus interpretations, the sensitivity, specificity, and accuracy of ovarian origin determination were calculated. RESULTS: Gonadal veins were shown in more than 70% of the subjects in both the control group and the patients with a mass (hereafter referred to as the "mass group"). There was no significant difference in the diameter of the gonadal veins between the control and mass groups and between patients with an ovarian mass and those with a uterine mass. The values for the area under the ROC curve (A(z)) of the two observers for ovarian origin determination were 0.90 and 0.92. The kappa value was 0.48. The sensitivity, specificity, and accuracy were 83.3%, 87.5%, and 84.9%, respectively. CONCLUSION: Gonadal veins can be shown on MDCT with high consistency; MDCT provides useful information for determining the origin of relatively large pelvic tumors arising in females.
OBJECTIVE: The objective of our study was to determine the usefulness of recognizing the continuity of the gonadal veins to the pelvic mass to differentiate ovarian versus uterine origin on MDCT in females with a large pelvic mass. MATERIALS AND METHODS: Two radiologists interpreted the MDCT images obtained on a monitor, using paging methods, in 86 female patients with a large pelvic mass (> 8 cm) and 40 patients without an abdominopelvic mass as control subjects. The following issues were recorded using a 5-point scale: visualization of gonadal veins and origin determination based on anatomic continuity. Receiver operating characteristic (ROC) curve analysis was performed, and the interobserver differences were checked with kappa statistics. The maximum diameters of the gonadal veins were also measured. With consensus interpretations, the sensitivity, specificity, and accuracy of ovarian origin determination were calculated. RESULTS: Gonadal veins were shown in more than 70% of the subjects in both the control group and the patients with a mass (hereafter referred to as the "mass group"). There was no significant difference in the diameter of the gonadal veins between the control and mass groups and between patients with an ovarian mass and those with a uterine mass. The values for the area under the ROC curve (A(z)) of the two observers for ovarian origin determination were 0.90 and 0.92. The kappa value was 0.48. The sensitivity, specificity, and accuracy were 83.3%, 87.5%, and 84.9%, respectively. CONCLUSION: Gonadal veins can be shown on MDCT with high consistency; MDCT provides useful information for determining the origin of relatively large pelvic tumors arising in females.