The Value Of Monoenergetic Imaging For Optimal Display And Prediction Of Muscle-invasive Bladder Cancer With Dual-energy CT

Part Ⅰ Usefulness of bladder cancer by monoenergetic imaging with dual-energy CTObjective: To objectively and subjectively evaluate the ability of dual-energy CT monoenergetic images and mixed-energy images to display bladder cancer lesions.Methods: This retrospective study evaluated the images and data of 65 patients bladder cancer who had undergone dual-phase contrast-enhanced DECT.120 k Vp polyenergetic images(PEI)and four Mono E in 10-kiloelectron volt(ke V)intervals(range,40~70 ke V),iodine map(ID),iodine overlay map(IO)and effective atomic number map(Zeff)were reconstructed.The quality of the reconstructed image and the clarity of the lesions were scored subjectively and objectively.The objective indexes included image noise,signalto-noise ratio(SNR),contrast-to-noise ratio(CNR)and CTBC-BW value.We also subjectively assessed clarity of BC in PEI and Mono E using a 4-grade Likert scale.Friedman test was used for inter-group comparison.Results: The noise,CNR and CTBC-BW values in arterial and venous phase decreased gradually with the increase of ke V level.The CNR and CTBC-BW values of 40 ke V Mono E were significantly higher than those of 60~70ke V Mono E and 120 k Vp PEI(P <0.05).The subjective score of 40 ke V Mono E in arteriovenous phase was the highest,which was higher than that of 60~70 ke V Mono E,120 k Vp PEI,IO,ID and Zeff(P <0.05).In addition,the objective and subjective scores of the venous phase of bladder cancer was higher than those of the arterial phase,and the difference was statistically significant(P <0.05).Conclusions: 40 ke V Mono E can improve the display ability of bladder cancer,and further improve the accuracy and confidence of radiologists in the diagnosis of bladder cancer.Part Ⅱ The value of dual-energy CT parameters in distinguishing muscle-invasive bladder cancerObjective: To explore the application value of dual-energy CT parameters in the differential diagnosis of muscular invasive bladder cancer and non-muscular invasive bladder cancer.To evaluate the application value of texture features based on 40 ke V Mono E images in distinguishing MIBC and NMIBC.Methods: According to the inclusion and exclusion criteria,33 cases were in the nonmuscle invasion group and 20 cases were in the muscle invasion group.Clinical risk factors for bladder cancer lesions included age,sex,single/multiple lesions,maximum focal diameter,and length of contact between the base of the tumor and the bladder wall(TCL).Iodine concentration(IC),standardized iodine concentration(NIC),effective atomic number(Zeff),standardized effective atomic number(NZeff)and energy spectrum slope(λHU)were measured and calculated.The volume of focus was delineated on the 40 ke V Mono E image,and the texture features of the area of interest were extracted.The optimal texture features were selected by single factor and multi-factor analysis.A joint model was constructed for clinical risk factors,spectrum parameters and texture features with statistical differences,and receiver operating characteristic(ROC)curve was drawn.The area under the curve(AUC),sensitivity and specificity were calculated to test the diagnostic efficacy of each model.Results: Among the clinical risk factors,TCL of MIBC was significantly higher than NMIBC.The optimal threshold was 2.91,the AUC value was 0.817,the sensitivity was 75%,and the specificity was 81.8%.In terms of spectrum parameters,IC,Zeff and λHU of MIBC in the venous phase were higher than NMIBC,the differences were statistically significant(P <0.05),and IC in the venous phase was an independent predictor with the best diagnostic effect,AUC was 0.764,the optimal threshold was 2.425,the sensitivity was 75.0% and the specificity was 69.7%.Among the texture features,90 Percentile,kurtosis and deficit normalization were the best texture parameters to identify MIBC,and the AUC was 0.767,0.730 and 0.700,respectively.The AUC of binary logistic regression model constructed by combining the three texture features was 0.856,the sensitivity was 95%,and the specificity was 66.7%.Combined with clinical risk factor(TCL),spectrum parameters(IC in venous phase)and three texture features,the AUC was increased to 0.979,the sensitivity was 95%,and the specificity was 100%.Conclusion: The quantitative parameters and texture features of dual-energy CT have high diagnostic value in the identification of MIBC and NMIBC,which is helpful for the preoperative diagnosis of bladder cancer.