The Diagnostic Value Of IVIM-DWI Combined With Time-intensity Curve In Breast Lesions And Its Correlation With Ki-67 Expression

Part one The differential diagnosis value of IVIM-DWI combined with DCE-TIC curve in breast benign and malignant lesionsPurposeTo explore the diagnostic value of single and double exponential diffusion-weighted imaging and dynamic enhancement time-signal intensity curve（TIC）curve in distinguishing benign and malignant breast lesions;combine single and double exponential diffusion-weighted imaging with TIC curve respectively,and to explore the diagnostic efficacy of combined factors for benign and malignant breast lesions,and to provide a more reliable imaging method for preoperative diagnosis.Materials and MethodsProspectively enrolled patients with breast lesions confirmed by pathology at Taizhou People’s Hospital from April 2019 to October 2020.A total of 71 women were included,including 81 lesions（41 malignant lesions and 40 benign lesions）.All patients underwent conventional MRI scan,DWI,multi-b value IVIM-DWI and dynamic enhanced scan.Send DCE-MRI and DWI images to the post-processing workstation respectively,select the images of all phases of dynamic enhancement,generate subtraction maps,use Mean-Curve post-processing software,take the most obvious lesion enhancement as the ROI,and plot the TIC curve;refer to T₂WI turbo inversion recovery magnitude（Tirm）and dynamic enhancement scan image,manually outline ROI on the ADC map,and measure the apparent diffusion coefficient（ADC）;import IVIM sequence into third-party software（Fire Voxel:CAL²R,New York University,NY）,refer to ADC image,Tirm sequence and dynamic enhancement scan image,manually outline ROI,obtain Pure diffusion coefficient（D）,Pseudo diffusion coefficient（D^*）and Perfusion fraction（f）respectively.All the above ROIs are delineated at the maximum cross-sectional area of the lesion,and the size and location are as consistent as possible,including the substantial part of the lesion,and try to avoid the cystic and necrotic area of the lesion.All data are measured 3 times and the average value is taken.The parameter values（D,D^*,f,ADC）of the two groups were statistically tested.Two independent samples t-test was used for normal distribution,while Mann-Whitney U test was used for non-normal distribution.χ²test was used to compare the types of TIC curves between the two groups.By drawing the receiver operating characteristic curve（ROC）curve,comparing the differential diagnosis performance of each parameter for benign and malignant breast lesions,and selecting statistically significant parameters to combine with the TIC curve to compare the diagnostic performance of the combined factors.P<0.05 considered the difference to be statistically significant.Results1.The D value of the benign group[（1.36±0.28）x10^-3mm²/s],ADC value[（1.31±0.25）x10^-3mm²/s]was significantly higher than the D value of the malignant group[（0.90±0.16）x10^-3mm²/s],ADC value[（0.93±0.16）x10^-3mm²/s]（P<0.01）;the difference in f value and D^*value was not statistically significant（P>0.05）.2.The benign group is more common in the inflow curve[52.50%（21/40）];the malignant group is more common in the outflow curve[53.66%（22/41）].There was a statistically significant difference in the type of TIC curve between the two groups（χ2=37.8,P<0.05）.3.ROC curve shows:TIC curve has high sensitivity（95.12%）in diagnosing breast cancer,but relatively low specificity（52.50%）;among the parameters of IVIM and DWI,D value has the highest diagnostic efficiency,with AUC of 0.914,sensitivity of92.70%,specificity of 80.00%,and accuracy of 86.41%;combining the D value and ADC value with the TIC curve respectively,shows that the combined diagnostic efficiency of the D value and the TIC curve is the best.The AUC is 0.951,the sensitivity is increased by 95.10%,the specificity is increased to 82.5%,and the accuracy is increased to 88.89%.Conclusion1.The sensitivity of the DCE-TIC curve to diagnose breast cancer is very high,but the specificity is relatively low.2.Both IVIM and traditional DWI quantitative parameters are helpful for the differential diagnosis of breast cancer,among which the D value has the best diagnostic performance.3.The combination of D value and TIC curve shows higher diagnostic efficiency,which improves the accuracy of diagnosis.Part two Correlation between IVIM-DWI quantitative parameters of breast cancer and Ki-67 expressionPurpose To explore the correlation between single and double exponential diffusion-weighted imaging and the expression index of the prognostic factor Ki-67 in breast cancer.Materials and Methods Prospectively enrolled patients with breast cancer who were pathologically confirmed by Taizhou People’s Hospital from April 2019 to October 2020,a total of 35 female patients（a total of 41 lesions）.All patients underwent conventional plain MRI,DWI,multi-b value IVIM-DWI and dynamic enhancement scan.The pathological types and immunohistochemical index Ki-67 expression index of 41 breast cancers were analyzed,and 20% was used as the cut-off value.Patients were divided into Ki-67 high expression group（n=22）and Ki-67 low expression group（n =19）.Transfer the DWI images to the post-processing workstation,refer to the T2 WI turbo inversion recovery magnitude（Tirm）and the dynamic enhancement scan image,manually outline the ROI on the ADC map,and measure the apparent diffusion coefficient（ADC）;Import the IVIM sequence into a third-party software（Fire Voxel: CAL2 R,New York University,NY）,refer to the ADC image,Tirm sequence and dynamic enhanced scan image,manually outline the ROI,and obtain the Pure diffusion coefficient（D）and Pseudo diffusion coefficient（D*）And Perfusion fraction（f）.All the above ROIs are delineated at the maximum cross-sectional area of the lesion,and the size and location are as consistent as possible,including the substantial part of the lesion,and try to avoid the cystic and necrotic area of the lesion.All data are measured 3 times and the average value is taken.The two groups of parameter values（D,D*,f,ADC）were statistically tested.If they conform to the normal distribution,the t test is used,otherwise,the non-parametric test is used.The differences between the parameters in the Ki-67 high and low expression groups were compared,and use Pearson correlation analysis to calculate the correlation coefficient to evaluate the relationship between each parameter and Ki-67 index.P<0.05 considered the difference to be statistically significant.Results1.The D value of Ki-67 high expression group [（0.81±0.13）x10-3mm2/s],ADC value [（0.87±0.15）x10-3mm2/s] were significantly lower than the D value of low expression group [（1.00±0.13）x10-3mm2/s],ADC value [（1.00±0.15）x10-3mm2/s]（P<0.05）;the difference in f value and D* value was not statistically significant（P>0.05）.2.The D value,ADC value and breast cancer Ki-67 expression index are negatively correlated（r value is-0.598,-0.378,P<0.05）,among which the correlation coefficient of D value is the largest;f value,D* value and breast The expression of Ki-67 in cancer has no significant correlation（P>0.05）.Conclusion The D value of IVIM and the ADC value of traditional DWI are both correlated with the Ki-67 expression index of breast cancer,and the correlation between the D value and the Ki-67 index is better than the ADC value,indicating that the D value is expected to become the breast Ki-67 index The predictive indicators of,provide an important reference for guiding individualized treatment and assessing prognosis.