The usefulness of computed tomographic colonography for evaluation of deep infiltrating endometriosis: comparison with magnetic resonance imaging.

OBJECTIVE: We wanted to assess the diagnostic value and morphologic feature of deep infiltrating endometriosis (DIE), involving rectosigmoid colon, with computed tomography (CT) colonography in comparison with magnetic resonance (MR). METHODS: Fifty patients with DIE, who had undergone CT colonography and MR imaging (MRI) before surgery, were enrolled. Among these 50 patients who underwent laparoscopic surgery with DIE, 37 patients were diagnosed as rectosigmoid involvement of endometriosis (presence of rectal outer wall involvement, endometrial implantation in the rectouterine space, and complete obliteration of cul-de-sac with histologic confirmation). Image findings at CT colonography and MRI were reviewed by 2 radiologists to determine whether there are radiologic features that can help predict rectosigmoid endometriosis. Computed tomography images were analyzed for the luminal alteration of rectosigmoid colon, mural thickening, and mass formation in the rectosigmoid colon, whereas MR images were analyzed for the mass formation in the rectosigmoid colon, fat plane obliteration, between rectum and uterus or vagina, retroversion of uterus, retraction of uterus toward the colon, and detection of endometrial spot in rectovaginal or rectouterine space. Statistical analysis was performed with Pearson χ and receiver operating characteristic curve. RESULTS: With CT colonography, the luminal alteration of rectosigmoid colon was detected with sensitivity of 96.0% and specificity of 48.0% (P < 0.001) in the overall rectosigmoid endometriosis and with sensitivity of 84.0% and specificity of 80.0% (P = 0.005) in the case of DIE with cul-de-sac obliteration. With MR, the sensitivity and specificity for detection of endometriosis of rectosigmoid or rectouterine space were 94.4% and 37.5% (P = 0.013), respectively. Other variables had no statistical significance. The diagnostic accuracy of CT colonography is higher than that of MRI (area under the curve, 0.786 vs 0.691; P < 0.001), for the overall rectosigmoid endometriosis. In the evaluation of complete cul-de-sac obliteration, morphologic change of rectosigmoid colon is identified more accurately with CT colonography than that of MRI (area under the curve, 0.821 vs 0.686; P < 0.001). CONCLUSIONS: Both CT colonography and MRI are highly sensitive to the detection of rectosigmoid endometriosis, but lack specificity. However, the depiction rate of morphologic change in rectosigmoid colon is greater with CT colonography than that of MRI, in the case of cul-de-sac obliteration. The luminal alteration significantly correlates with morphologic change in rectosigmoid endometriosis.

Optimal cut-off value of perfusion parameters for diagnosing prostate cancer and for assessing aggressiveness associated with Gleason score.

To determine cut-off value of dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) parameters for differentiation of prostate malignant from benign and cancer with high-grade Gleason score (GS) (GS>7) from low-grade GS (GS≤7), 35 patients (24 malignant and 11 benign) who underwent DCE-MRI were included. Difference between malignant and benign was statistically significant for all magnetic resonance parameters except Ve. The cut-off values were K(trans)=0.184min(-1), Kep=0.695min(-1), iAUC=4.219mmol/l/min, and ADC=1340.5mm(2)/s. A significant difference in mean values of K(trans) and Kep between cancer with high-grade GS and low-grade GS was also observed. K(trans) and Kep showed a significant correlation with GS.