Combined subjective and quantitative analysis of magnetic resonance images could improve the diagnostic performance of deep myometrial invasion in endometrial cancer

Publication date: Available online 10 February 2017
Source:Clinical Imaging
Author(s): Lei Deng, Qiu-ping Wang, Rui Yan, Nan Yu, Lu Bai, Xiao-yi Duan, You-min Guo
PurposeTo evaluate whether the combination of subjective magnetic resonance imaging (MRI) and quantitative analysis by using the exponential ADC (EADC) value of the peri-endometrial zone can improve the diagnostic performance of deep myometrial invasion in endometrial cancer patients.Materials and methodsWe prospectively evaluated 111 patients with either cervical cancer (normal endometria group) or endometrial cancer (endometrial cancer group). Two radiologists assessed all preoperative MR images with T1, T2, and diffusion-weighted imaging. The EADC value of the peri-endometrial zone was measured. Sensitivity, specificity, positive and negative predictive values, and the area under the receiver operating characteristic curve (Az) were calculated for Subjective MRI, an EADC cutoff value of the peri-endometrial zone and the combination of the two methods in assessing the prediction of deep myometrial invasion.ResultsSpecificity for EADC cutoff of the peri-endometrial zone was higher (0.93) than for Subjective MRI (0.80), as were the positive predictive values (EADC, 0.79; visual, 0.60). Sensitivity for the combined test was higher (0.88) than for Subjective MRI (0.71) and the EADC cutoff value (0.65), as were the negative predictive values (the combined test, 0.94; vs. EADC, 0.79; vs. Subjective MRI, 0.60). There were no differences in Az between the three methods (P>0.05), but the combined test had the highest Az.ConclusionsCombined with conventional Subjective MRI, calculating EADC value of the peri-endometrial zone could improve the accuracy of preoperative assessment of deep myometrial invasion in endometrial cancer patients, and maybe helpful in tailoring a surgical approach for intervention.



http://ift.tt/2kfOzr5