One hundred thirty patients with 170 epithelial ovarian tumors were prospectively studied with computed tomography (CT) before surgery. Ultrasound (US) was performed in 108 patients with 138 tumors. At pathologic examination, 78 tumors (46%) were benign, 14 (8%) borderline, and 78 (46%) malignant. CT results were compared with surgical and pathologic findings in all patients. CT enabled detection of 148 of 170 tumors (87%), and US enabled detection of 118 of 138 tumors (86%). Benign serous cystadenomas (n = 42) were correctly characterized with a sensitivity of 69% at CT and 70% at US. Benign mucinous cystadenomas (n = 21) were correctly characterized with a sensitivity of 62% at CT and 50% at US. Malignancy was suggested in nine of 14 patients (64%) with borderline tumors at CT and in five of 14 (36%) at US. The overall accuracy of characterization of benign versus malignant tumors (including borderline tumors) was 94% with CT and 80% with US. In the 108 patients studied with both CT and US, the sensitivity of CT was significantly superior to that of US (P less than .03), whereas there was no significant difference in specificity (P = .125).

No abstract

Computed tomography (CT) was performed in 38 patients with 41 benign cystic teratomas of the ovary and two patients with malignant transformation. CT depicted all tumors. The presence of fat in 40 of 43 cases (93%), tooth or calcification in 24 of 43 (56%), Rokitansky protuberance in 35 of 43 (81%), tufts of hair in 28 of 43 (65%), and a fat-fluid level in five of 43 (12%) allowed a definite diagnosis of ovarian cystic teratoma in 42 of 43 cases (98%). In the two cases of malignancy, single large (greater than 10 cm) plugs (with uptake of contrast medium in one) with a cauliflower appearance and an irregular border forming an obtuse angle with the inner wall of the cyst suggested malignant transformation. In three cases of benign cystic teratoma, a mucinous tumor (one benign, one borderline, one malignant) arising in the same ovary was seen at pathologic examination but was only diagnosed with the help of CT in two of three cases. Thickening of the tube was noted in two cases of torsion of the adnexa. CT findings were compared with findings at radiography of the abdomen and hysterosalpingography in 30 cases, ultrasound in 31, and magnetic resonance imaging in three. This study demonstrated that CT was the best procedure for imaging cystic teratomas of the ovary.

Forty patients with 50 ovarian epithelial tumors of the ovary were retrospectively studied. They underwent computed tomography and magnetic resonance (MR) imaging within 1 week of surgery. MR examinations were performed with a superconducting magnet (0.5 T) and predominantly T1- and T2-weighted MR imaging. T1-weighted MR images were obtained in eight patients after contrast material was administered. Signal intensity of tumors was compared with that of urine, muscle, and fat. Morphologic features that were evaluated included size, vegetations, septations, wall thickness, fluid or solid components, and vascularity. On the basis of signal intensity and morphologic characteristics, MR imaging helped in the correct diagnosis of benign serous cystadenoma in 10 of 13 tumors, benign mucinous cystadenoma in seven of 10 tumors, and tumors of low malignant potential and malignant tumors in 23 of 27 tumors. Accuracy for overall characterization of benign versus malignant tumors was 86% with MR imaging and 92% with computed tomography. There was no difference in sensitivity (P = 1) or specificity (P = .5).

No abstract