| Objective:The study aims to investigate the potential value of monoexponential,biexponential diffusion models parameters(ADC,D,D*and f values)in prognostic risk factors(histological grade,tumor diameter,immunohistochemical factors,lymph node status and molecular subtypes)of invasive breast cancer;and to evaluate diagnostic efficacy of those parameters in prognostic risk factors.We expect that it can provide radiological references for prognosis prediction and treatment guidance of breast cancer patients.Materials and methods:Our study retrospectively gathered 70patients in The Second Hospital,University of South China with invasive breast cancer from May 2019 to December 2020.The research excluded17 cases.Eventually,a total of 53 patients were enrolled,with a mean age of 52.21±10.95 years.Mammary MR routine plain scan and DCE-MRI,DWI,IVIM sequence examination were performed in all patients before biopsy.All images were transmitted to the GE ADW 4.6 post-processing workstation.The relevant parameters ADC,D,D*and f values were automatically obtained through the DWI single and daul index models.Two-sample t test or Wilcoxon rank sum test was used to analyze the difference of each parameter in histological grade,lymph node state and tumor diameter of breast cancer.Afterwards,evaluating the diagnostic capabilities of each meaningful parameter by ROC curve.They were divided into negative and positive groups according to the expression of four biomarkers(ER,PR,Ki-67 and HER-2).Two-sample t test or Wilcoxon rank sum test was used to analyze the differences among parameters.The diagnostic capabilities of each meaningful parameter was evaluated by ROC curve.Results:1.The ADC and D values of high-grade breast cancer were less than those of low-grade breast cancer.it was meaningful in statistics.Moreover,the diagnostic capabilities of D value was better than ADC value.The difference in D*and f values of lesions were not statistically significant between the high-grade and low-grade breast cancer.2.The ROC curve analysis for ADC and D values showed that the cut-off values of high-grade and low-grade breast cancer were 0.895 and0.687.The sensitivity,specificity and AUC were 1.00,0.655,0.882 and0.958,0.897,0.958,respectively.3.Ki-67 negative group demonstrated significantly lower ADC,D values as compared with positive group;D*and f values has nothing to do with Ki-67 groups;The difference in parameters of DWI single and dual index models was not statistically significant among PR,ER,HER-2 negative and positive groups.4.The cut-off value of ADC and D differentiating high Ki-67 and less Ki-67 breast cancer were 0.958,0.655.Among which the sensitivity,specificity and AUC were 0.650,0.727,0.708 and 0.850,0.606,0.723 respectively;5.ADC and D value of TNBC were less than non-TNBC,and it was meaningful in statistics.D*and f values has no relation with TNBC,non-TNBC;6.The cut-off values of ADC and D differentiating TNBC and non-TNBC are 0.912,0.655;among which sensitivity,specificity and AUC were 0.714,0.818,0.742 and 0.690,0.999,0.810 respectively;7.monoexponential-biexponential models parameters such as ADC,D,D*and f values has no obvious statistically significant in tumor diameter group and lymph node status group;Conclusion:1.The ADC and D value were a useful parameter to identify the high-grade breast cancer;2.It was feasible to use DWI single and dual index models to assess the Ki-67 expression;3.Histological grade of TNBC was more common at the level III.Compared with non-TNBC,TNBC shows lower ADC and D values;4.DWI single and dual index models can noninvasively assess the prognosis of invasive breast cancer patients by parameters such as ADC and D values.Besides,the assess efficacy of D value was better than that of ADC value. |
PDF Full Text Request