WANG Xue,LU Yi,SONG Jiao, et al. Diagnostic value and correlation analysis of parameters derived from two-compartment exchange model and IVIM and models for uterine cervical carcinoma[J]. JOURNAL OF WEZHOU MEDICAL UNIVERSITY, 2020, 50(10): 820-824.
Abstract：Objective: To analyze the diagnostic value of parameters as well as the correlation of perfusion parameters derived from CC (two-compartment exchange model) and IVIM (intravoxel incoherent motion)models for uterine cervical carcinoma. Methods: Thirty-two patients diagnosed with uterine cervical carcinoma pathologically and 14 healthy subjects were included in the Second Affiliated Hospital & Yuying Children’s Hospital of Wenzhou Medical University from April 2016 to December 2018. MRI examinations including dynamic contrasted-enhanced MRI (DCE-MRI) and IVIM sequences were performed for all the subjects before treatment. Parameters derived from CC model were listed as follows: tissue blood flow (Fp), tissue blood volume (Vp), interstitial volume (Ve), vascular permeability (PS). Parameters derived from IVIM model were listed: true diffusion coefficient (D), pseudo diffusion coefficient (D*), fraction of fast apparent diffusion coefficient (f) and fD*. ROC test was used to compare the corresponding AUC (the area under the curve) value subsequent to the comparsion of the differences of those parameters from CC and IVIM models between cervix carcinoma and normal cervix tissue. Spearman’s correlation was applied to analyze the perfusion parameters between CC and IVIM models from cervical carcinoma and normal cervix tissue. Results: The differences between the parameters of CC model (Vp, Ve, PS) in cervical carcinoma mass and normal cervix tissue were statistically significant(P<0.05). The AUC of the parameters of CC model was 0.546-0.852 while the AUC of CC-PS and CC-Ve werere spectively >0.8 and >0.7. The differencess of IVIM parameters (D, D*, f, fD*) between cervical carcinoma mass and normal cervix tissue were statistically significant (P<0.05). The AUC of each parameter of IVIM was0.642-0.953, in which the AUC of IVIM-D, IVIM-f and IVIM-fD* were >0.8. In cervical carcinoma, IVIM-f and IVIM-fD* were slightly correlated with CC-Ve (0.3< r≤0.5). Conclusion: Evaluating and identifying cervical carcinoma with the quantitative parameters of CC and IVIM models were found to be valuable predictors. The perfusion parameters from IVIM model of cervical carcinoma have the potential to be used as a non-invasive method to evaluate the microcirculation status of the tumor.
