The Application Of Monoexponential And Kurtosis Diffusion Weighted MR Imaging In Evaluation Of Breast Lesions

Recently,breast disease has been the most common disease which effected female's health.Breast cancer is a serious worldwide disease which threat to women's health,especially in Europe and the United States.In China,the rate of breast cancer is slightly lower than Europe and the United States,but in recent years its incidence increase year by year and the age of it is more and more young,so it is more and more important to detect,diagnosis and treatment early in order to improve patient survival and quality of life.In recent years,as we all know magnetic resonance has a high soft tissue resolution,so it has been widely used in the diagnosis of breast disease,especially diffusion-weighted imaging which can reflect the degree of tissue water molecules and the complexity of the organization,it becomes a hot topic of differential diagnosis of benign and malignant breast lesions.With the development of technology,the DTI model,DKI model which are based on DWI are the continuous development and continue to be used in the system of the various systems.The monoexponential model and DTI model are suppose that the diffusion of water molecules are Gaussian distribution which suppose the biological tissue water molecules move in a unrestricted form.However,due to the complexity of the organizational structure,water molecules meet the Gaussian distribution no longer with the increase of b value.In 2005,DKI model was first proposed by Professor Jenson of New York University.Based on DWI,the non-Gaussian distribution of water molecules was detected,which reflected the change of microstructure.MK is the most important parameter of DKI model,which represents the average diffusion peak value of each gradient direction.It is a measure of the complexity of organizational structure,which is proportional to the complexity of organization.The more complex structure is,the more obvious the diffusion of water molecules is.AK and RK are the average diffusion kurtosis values parallel ?vertical direction to the diffusion tensor,and size the quantifies degree of diffusion of water molecules in this direction.MD value of the average diffusion rate,can be reflected in the tensor ball within the three axes of water molecules on the average diffusion strength.In this study,we analyzed the clinical,pathological and magnetic resonance data of 85 patients with breast lesions,and analyzed the effects of the monoexponential model and the diffusion kurtosis model on the differential diagnosis of benign and malignant breast lesions.The relationship between the parameters.The diagnostic efficacy of the parameters in differential of benign and malignant breast lesions.Materials and Methods: A total of 98 patients with clinical palpation,ultrasound or molybdenum target examination were collected.The patients were scanned with GE 750 3.0 T magnetic resonance imaging.ALL patients scan the routine axial T1 WI,axial compression T2 WI,single b Imaging,multiple b-values DKI and dynamic enhancement,finally a total of 85 patients(90 lesions)were included in the study.On the GE workstation(Advantage Windows 4.5),the ADC software and DKI processing software in the Function Tools toolkit were used to post-process the single-b DWI images and the multi-b-valued DKI images to obtain the parameters,ADC?MK?AK?RK?MD.Statistical analysis was performed using SPSS statistical software.Statistical analysis was performed using the Kruskal-Wallis test.The statistical differences between the parameters in benign lesions,malignant lesions and normal breast fibrous glands were compared using the Mann-Whitney U test.The correlation between the parameters were analyzed by Spearman correlated analysis.Using the operating operating curve(ROC curve),the area under the curve was calculated to evaluate the diagnostic efficacy of each parameter in the differential diagnosis of benign and malignant breast lesions.The optimal thresholds of the parameters and the corresponding Jorden index,sensitivity and specificity were calculated.Results: The ADC,MK,AK,RK and MD values were significantly different among the normal breast tissues,benign tumors and malignant tumors(P<0.05).The ADC values and MD of normal glandular tissues and benign lesions were significantly higher than those of malignant lesions.The MK,AK and RK of those of normal gland tissues and benign lesions were significantly lower than malignant lesions.There was a significant positive correlation between ADC value and MD value(r=0.451,P<0.05).MK,AK,RK values were negatively correlated with ADC values(r1=-0.503,P<0.05;r2=-0.485,P<0.05;r3=-0.524,P<0.05).The area under the curve(AUC)of ADC,MK,AK,RK and MD was 0.680,0.741,0.728,0.782 and 0.688 respectively.The optimal cutoff values of each parameter in the identification of benign and malignant breast lesions were 1.145 ×10-3mm2/s,0.595,0.5145,0.5465,1.615×10-3mm2/s.The sensitivity of each parameter was 65.9%,80.6%,80.6%,75% and 81.8% respectively.The specificity was 51.7%,63.6%,65.9%,56.8%,60.1%.Conclusion: The monoexponential model and diffusion kurtosis imaging model has important diagnostic value in the identification of benign and malignant breast lesions.Compared with the monoexponential model,the diffusion kurtosis imaging model can reflect the diffusion of water molecules in the organization more truly,and can reflect the change of tissue microenvironment.The parameters derived from the DKI model have higher sensitivity and specificity than ADC derived from the monoexponential model.