大脑中动脉CTA-MIP后处理参数的优化及初步研究

Preliminary study of optimizing post-processing parameters of CTA-MIP for visualizing middle cerebral artery

目的应用CT血管成像(CTA)最大密度投影(MIP)的多平面重组(MPR)后处理技术分析大脑中动脉(MCA)分段显示与所选层面厚度及角度间的关系,确定MCA分段显示的优化参数。方法回顾性分析2014年6月—2016年6月因疑似MCA缺血行头颅CTA及数字减影血管造影(DSA)检查的100例病人(764处节段血管)资料,其中男60例,女40例,平均(64.3±10.7)岁。利用CTA容积再现(VR)平台,分别测量所有病例M1段、M2段、M2-5段的空间跨度及走行角度参数值并进行分组,随后在CTA-MIP-MPR平台,对MPR厚度和偏移角度的不同测试值进行验证,并对各个重组影像进行质量评分,确定各分段最优参数,并将优化CTA-MIP影像与DSA影像进行比较分析,评价其优化效果。多组间比较采用单因素方差分析,应用受试者操作特征(ROC)曲线评价优化CTA-MIP诊断血管狭窄的价值, CTA与DSA检查结果的一致性分析采用Kappa检验。结果在CTA-MIP中,MCA M1段最优化参数为冠状面、层厚15 mm、角度偏移0°,其影像质量评分为3.85±0.51,均高于其他组(均P<0.05);MCA M2段最优化参数为矢状面、层厚25 mm、角度偏移0°,其影像质量评分为3.36±0.53,均高于其他组(均P<0.05);M2-5段最优化参数为(1)低头冠状面、层厚35 mm、角度偏移35°;(2)外旋矢状面、层厚35 mm、轴转25°、矢转30°;2组影像质量评分分别为3.41±0.40、3.42±0.33,均高于其他组(均P<0.05)。优化CTA-MIP检查对血管狭窄的诊断敏感度、特异度、阳性预测值、阴性预测值、ROC曲线下面积分别为90.5%、97.7%、87.5%、98.3%、0.941。对于4种不同狭窄程度血管的诊断,优化CTA-MIP检查与DSA检查结果一致率为96.5%(κ=0.783,P<0.05)。结论基于优化的选层厚度及偏移角度,CTA-MIP分段显示能够清晰展示MCA各段管腔大小、走行及分支细节,与DSA有较高的一致性。

Objective Multiplanar reconstruction(MPR)post-processing technique of CT angiography(CTA)-max-imum intensity projection(MIP)was used to analyze the relationship between the visibility of middle cerebral artery(MCA)segments and the thickness and project angle of the selected layers,and to determine the optimal parameters for displaying MCA segments.Methods The data of 100 patients(764 segments)with suspected MCA ischemia examined by CTA and digital subtraction angiography(DSA)from June 2014 to June 2016 were retrospectively analyzed.There were 60 males and 40 females,with an average age of 64.3±10.7 years.Using the CTA-volume representation(VR)platform,the spatial spans and traveling angles of the M1,M2 and M2-5 segments were measured and classified for all cases.Then,on the CTA-MIPMPR platform,the quality of each group of recombinant images were verified and scored using different MPR thicknesses and projection angles,the optimal parameters of each segment were determined,and the optimized CTA-MIP images were com-pared with DSA images to evaluate the accuracy.One-way ANOVA analysis was used for multi-group comparisons,ROC curve was used to evaluate the efficiencies of optimized CTA-MIP for identifying vascular stenosis.The consistency between the CTA and the DSA was tested by Kappa.Results In CTA-MIP,the optimal parameters of presenting MCA M1 segment were coronal plane,slice thickness 15 mm and project angle offset 0 degree,and the image quality score of MCA M1 segment was 3.85±0.51,which was higher than those of other combinations(all P<0.05).The optimal parameters of presenting MCA M2 segment were sagittal plane,slice thickness 25 mm and project angle offset 0 degree,and the image quality score of MCA M2 segment was 3.36±0.53,which was higher than those of other combinations(all P<0.05).The optimal parameters of M2-5 segment were:①low head coronal plane,slice thickness 35 mm and project angle offset 35 degrees;②external rotation sagit-tal plane,slice thickness 35 mm and axial rotation 25 degrees,sagittal rotation 30 degrees,and the image quality scores of the two groups were 3.41±0.40 and 3.42±0.33 respectively,which were higher than those of the other combinations(all P<0.05).The sensitivity,specificity,positive predictive value,negative predictive value,and the area under the ROC curve of CTAMIP examination to identify stenosis were 90.5%,97.7%,87.5%,98.3%,0.941,respectively.After all blood vessels were di-vided into 4 types according to the degree of stenosis,the agreement rate between the optimized CTA-MIP and the DSA was 96.5%(κ=0.783,P<0.05),which was statistically significant.Conclusions With the optimized slice thickness and project angle,CTA-MIP segmentation can clearly show the lumen size,walking and branch details of MCA,and has high consistency with DSA.