【作者】 张刚；

【导师】 成官迅；

【作者基本信息】 南方医科大学 ， 影像医学与核医学， 2010， 硕士

【摘要】 [目的]1、以双源CT研究冠状动脉的解剖,评估双源CT冠状动脉成像检查对冠状动脉变异的诊断价值。2、探讨双源CT检查与超声心动图检查在左心室功能评估中的相关性及一致性。3、探讨双源CT冠状动脉成像检查对临床上怀疑或确诊冠心病患者冠状动脉狭窄病变的诊断价值。[材料和方法]1、研究对象第一部分：对2009年5月-2009年10月期间638例双源CT冠状动脉成像检查资料进行研究,其中男416例,女222例,年龄26-87岁,平均年龄56.2岁。第二部分：对2009年7月-2010年1月期间119例3天内同时做双源CT冠状动脉成像和超声心动图检查的患者的影像资料进行研究,其中男71例,女48例,年龄28-87岁,平均年龄60.9岁。第三部分：对2009年5月-2010年3月62例1周内同时做双源CT和传统的冠状动脉造影(coronary angiography, CAG)检查的患者的影像学资料进行研究,其中男43例,女19例,年龄36-87岁,平均年龄62.0岁。2、双源CT扫描方案所有患者于采用德国SIEMENS公司SOMATOM Definition双源CT扫描仪进行扫描。扫描前3min舌下喷服硝酸甘油0.5mg用以扩张冠状动脉。取仰卧位,头足方向屏气扫描。扫描范围为气管分叉下方10-15mm至膈下。在主动脉根部层面设定感兴趣区,触发阈值80-100HU,延迟5s增强扫描。肘前静脉埋置18G套管针,采用美国MEDRAD双筒高压注射器以5.0-6.5ml/s的流速注入60-80ml非离子型对比剂碘普罗胺(370mgI/ml)和后续25ml生理盐水,造影剂注射与增强扫描同时进行。扫描参数：管电压120kV,管电流380-430mAs,准直64×0.6mm,螺距0.20-0.50,旋转时间0.33s,扫描时间7-11s。心电图数据同时记录在CT扫描过程中。扫描完成后机器自动生成舒张期和收缩期冠状动脉最佳时相序列图像,卷积核B26f,重组层厚为0.75mm,间距为0.5mm。3、图像后处理与分析第一部分：将重建序列图像传输至Syngo后处理工作站,对冠状动脉进行多平面重建(multiplanar reconstruction, MPR)、曲面重建(curved planar reformation,CPR)、容积再现(volume rendering, VR)、最大密度投影(maximum intensity projection, MIP)等多种图像后处理,观察冠状动脉的起源、走行、终止情况及与周围组织的解剖关系。所有冠状动脉图像均由2名在心血管诊断方面有丰富经验的影像诊断医师在工作站进行综合评估,意见不一致时通过协商确定诊断。第二部分：将重建完成后的0%-100%全时相图像数据装入左心室功能分析软件。软件自动勾画左心室舒张末期和收缩末期心室腔内表面和外表面。进而软件自动计算出患者左心室舒张末期容积(end-diastolic volume, EDV)、收缩末期容积(end-systolic volume, ESV)、每搏输出量(stroke volume, SV)、射血分数(ejection fraction, EF)、室壁搏动幅度(wall motion)及左心室心肌质量(myocardial mass, MM)等指标。所有图像数据分析由一名有丰富操作经验的影像诊断医师独立完成。第三部分：将重建的数据传输至Syngo后处理工作站,对冠状动脉进行容积再现、多平面重建、最大密度投影等多种图像后处理,观察冠状动脉的狭窄及变异情况。采用美国心脏病协会(American Heart Association, AHA)建议的15段冠状动脉树状结构模型,评价所有直径≥1.5mm的冠状动脉节段。冠状动脉狭窄程度分为5级：①无狭窄；②轻度狭窄(<50%)；③中度狭窄(50%-75%)；④重度狭窄(76%-99%)；⑤完全闭塞(100%)。所有冠状动脉图像后处理及诊断均由2名对心血管方面有丰富经验的影像诊断医师在不知道CAG检查结果的情况下综合评估,意见不一致时通过协商确定诊断。4、统计学分析：第一部分：根据冠状动脉的起源、走行与终止位置异常将冠状动脉的解剖异常分为3大类型。冠状动脉的起源异常包括高位起源、多个开口、单一冠状动脉、冠状动脉异位起源于肺动脉、冠状动脉或其分支起自对侧或非冠状窦。冠状动脉走行异常包括心肌桥、动脉瘤和重复冠状动脉。冠状动脉终止位置异常包括冠状动脉瘘、冠状动脉弓等。第二部分：两种检查由两位医师在互不知道其他检查结果的情况下,独立测量左心室功能各项指标。采用SPSS 13.0统计软件包进行数据统计学处理,测量值以x±s表示。采用配对样本t检验,对两种方法的差异性进行评定；采用Pearson积矩相关,对两种检查方法的相关性趋势进行评定；采用Bland-Altman分析,对两种检查结果的一致性进行评定。P<0.05为差异有统计学意义。第三部分：应用SPSS 13.0软件包,以CAG检查结果作为“金标准”。采用Wilcoxon符号秩检验,评价双源CT与CAG对冠状动脉狭窄程度分级诊断有无差异,P<0.05为差异有统计学意义,并计算出双源CT对冠状动脉狭窄诊断的总体灵敏度、特异度、阳性预测值、阴性预测值及正确率。采用一致性Kappa检验,评价双源CT与CAG冠状动脉狭窄程度分级诊断的一致性。[结果]第一部分：638例中,左优势型占2.8%(18例),右优势型占89.0%(568例),均衡型占8.2%(52例)。冠状动脉变异占9.4%(60例),其中,右冠状动脉高位开口3例,左冠状动脉高位开口8例,副冠状动脉8例,左主干缺如3例,单冠状动脉2例,右冠状动脉冠状动脉瘤1例,无冠窦左房瘘1例,左回旋支起源于右冠状动脉1例,右冠状动脉起源于左冠窦1例,心肌桥32例。第二部分：两种检查方法的相关分析：Pearson积矩相关系数均小于0.05,说明两种方法各测量指标均存在显著正相关关系,其中,EDV、ESV、MM呈较强的相关(r>0.75)。采用配对t检验：除了EF两种检查方法无显著差异(t=0.34,P>0.05)外,两种方法的其余各测量指标均存在显著差异(P<0.05),双源CT的EDV、ESV、SV指标值均显著高于超声心动图检查,MM、LVPW搏幅指标值显著低于超声心动图检查。两种检查方法的一致性分析：两种检查方法还不具有较好的一致性,两种检查方法在临床上还不可以互相代替使用。第三部分：双源CT显示的冠状动脉直径≥1.5mm的节段共有830段,其中781段冠脉节段图像显示良好,可以进行评估。Wilcoxon符号秩检验Z=-4.486,P<0.001,说明两种检查方法所测的冠状动脉狭窄程度分级诊断有显著差异,双源CT所测得冠状动脉狭窄程度准确性低于CAG检查。双源CT对冠状动脉狭窄诊断的总体灵敏度为80.2%(167/208),特异度为95.6%(548/573),阳性预测值为87.0%(167/192),阴性预测值为93.0%(548/559),正确率为91.5%(715/781)。双源CT与CAG对冠状动脉狭窄程度的诊断一致性Kappa检验结果为κ=0.655,P<0.001,说明两种检查方法对冠状动脉狭窄程度的诊断具有中度一致性。双源CT对冠状动脉钙化、小分支病变诊断准确性欠佳。[结论]第一部分：双源CT冠状动脉成像凭借强大的时间和空间分辨及灵活的后处理技术,能准确直观地显示复杂的冠状动脉变异,可作为冠状动脉变异筛查的首选检查方法。第二部分：双源CT左室心功能分析与超声检查具有较好的相关性,但两种检查方法各测量指标还不具有较好的一致性。两种方法在临床上还不可以互相代替使用。双源CT左心室功能各测量指标值在临床上可作为一个参考。第三部分：双源CT冠状动脉成像能准确诊断临床上有意义的冠状动脉狭窄病变。由于双源CT对于冠状动脉钙化、分叉处及细小分支病变诊断的准确性欠佳,使双源CT对冠状动脉病变诊断的准确性仍低于CAG检查。双源CT冠状动脉成像可作临床上怀疑冠心病患者的CAG术前筛查及支架术后随访的首选方法。更多还原

【Abstract】 [Objective]1、To study the coronary artery anatomy and assess the diagnostic value of dual source CT(DSCT) angiography in detecting coronary anomalies.2、To discuss the correlation and concordance between DSCT and echocardiogram in left ventricular function analysis.3、To discuss the diagnostic value of DSCT coronary angiography in detecting clinical suspicion or confirmed diagnosis of coronary artery stenosis.[Meterials and Methods]1、Study populationPart one:The coronary artery imaging data of 638 patients (222 women,416 men; mean age 56.2 years; age range 26-87 years) who underwent DSCT angiography were studyed from May 2009 to October 2009.Part two:The image data of 119 patients (48 women,71 men; mean age 60.9 years; age range 28-87 years) who underwent DSCT and echocardiogram examinations within 3 days were studyed from July 2009 to January 2010.Part three:The coronary artery imaging data of 62 patients(19 women,43 men; mean age 62.0 years; age range 36-87 years) who underwent DSCT and coronary angiography(CAG) examinations within a week were studyed from May 2009 to February 2010.2、DSCT scan protocol All the patients were performed by DSCT scanner (SOMATOM Definition, Siemens Medical Solutions, Germany). Before the DSCT scan all patients were sprayed 0.5mg nitroglycerin sublingually for the expansion of coronary artery. After 3 min, the DSCT coronary angiography scan was started by continuously injecting a bolus of 80ml of ultravist (concentration:370mgI/mL) followed by 25ml saline solution into an antecubital vein via a 18-gauge catheter (injection rate 5.0-6.5ml/s) with high-pressure double-barrel syringe(MEDRAD, America). Contrast agent application was controlled by bolus tracking. Region of interest (ROI) was placed into the aortic root, and image acquisition started 5s after the signal attenuation reached the predefined threshold of 80-100 HU.Data acquisition was performed from 1.0-1.5cm below the level of the tracheal bifurcation to the diaphragm in a cranio-caudal direction with a detector collimation of 64 X 0.6mm, gantry rotation time 0.33s, pitch of 0.2-0.5 adapted to the heart rate, scan time 7-11s, tube current 380-430mAs per rotation, and tube voltage 120kV. Electrocardiogram (ECG) pulsing for radiation dose reduction was applied in all patients. ECG was digitally recorded during data acquisition and was stored with the unprocessed CT dataset. After the scan was completed, the machine automatically generated the best diastolic and systolic phase sequence of coronary artery images, reconstruction kernel B26f, slice width 0.75mm and increment 0.5mm.3、Image post-processing and analysisPart one:The reconstruction of image sequences were transferred to Syngo post-processing workstation. Multiplanar reconstruction(MPR), curved planar reconstruction(CPR), volume rendering(VR), maximum intensity projection (MIP) and many other image post-processing techniques were applied to observe the coronary artery origin, course, termination and the relationship of the surrounding tissue. All images were comprehensive assessed by the two radiologists who had experienced in cardiovascular diagnostic imaging at the workstation. Disagreements were solved through the consultation.Part two:The reconstruction image data of 0%-100% phases was load in left ventricular function analysis software. Software automatically outlined left ventricular end-diastolic and end-systolic ventricular cavity surface and the outer surface. Then the software automatically calculates the patients with left ventricular end-diastolic volume(EDV), end-systolic volume(ESV), stroke volume (SV), ejection fraction(EF), wall motion, left ventricular myocardial mass(MM) and other indicators. All image data analysis was down by a radiologist who had experienced in cardiovascular diagnostic imaging.Part three:The reconstruction data was send to the Syngo post-processing workstation, VR, MPR, MIP and other image processing techniques were applied to observed and the variation of coronary artery stenosis. Coronary artery stenosis was divided into five types:no stenosis; mild stenosis (<50%); moderate stenosis (50%-75%); severe stenosis (76%-99%); complete occlusion (100%). All images were comprehensive assessed by the two radiologists who were blinded to invasive coronary angiography results, using a 15-segment classification(based on American Heart Association, AHA), evaluated of all coronary artery diameter≥1.5mm segments. The consensus interpretation was compared with results of conventional invasive coronary angiography.4、Statistics analysisPart one:Based the anomalies of origin, anomalies of course, and anomalies of termination, we had classified the coronary artery anomalies into three types. These anomalies of origin included either the LCA or the RCA high takeoff, multiple ostia, single coronary artery, either the LCA or the RCA arising from the opposite coronary artery or branch, from the opposite or noncoronary sinus of Valsalva, from the pulmonary artery. These anomalies of course included myocardial bridge, coronary aneurysm and duplication of arteries. These anomalies of termination mainly included congenital coronary artery fistula and coronary arcade, et al.Part two:All DSCT and echocardiogram results were analyzed by an experienced cardiologist and radiologist respectively. Both observers were blinded to each other. Data was processed with SPSS 13.0 statistical package. Measurements were expressed as mean values±standard deviation. The correlation was tested with Pearson test. The difference was tested with paired sample t test. The concordance was tested with Bland-Altman analysis. P< 0.05 was considered statistically significant.Part three:The results of DSCT coronary angiography were compared with CAG which was as a gold standard in diagnosing coronary disease. Wilcoxon signed-rank test was used to compare the difference of DSCT and CAG in detecting coronary artery stenosis, P<0.05 was considered statistically significant. The overall sensitivity, specificity, accuracy, positive predictive value and negative predictive value of DSCT were tested. The concordance was tested with Kappa coefficient to compare the stenosis grading.[Results]Part one:Of 638 cases, left dominance accounted for 2.8%(18 cases), right dominance 89.0%(568 cases), balanced 8.2%(52 cases). The total of coronary anomalies accounted for 9.4%(60 cases), which included RCA high takeoff 3 cases, LCA high takeoff 8 cases, accesssory coronary artery 8 cases, absent LM trunk 2 cases, single coronary artery 2 cases, coronary aneurysm of RCA 1 case, coronary fistula between noncoronary sinus and left atrium 1 case, LCx from RCA 1 case, RCA from left sinus 1 case and myocardial bridge 32 cases.Part two:Correlation analysis of two methods:Pearson correlation coefficient was less than 0.05, indicating the two methods were significantly positive correlation and EDV, ESV, MM showed strong correlation (r>0.75). Paired t test:In addition to EF was no significant difference (t=0.34, P>0.05), the others were significantly different (P<0.05). The EDV, ESV, and SV of DSCT were higher than echocardiogram, MM lower than the echocardiogram. Consistency analysis:Two kinds of examination methods did not have a good consistency and didn’t replace the use of each other in clinical practice.Part three:781 coronary segments (diameter≥1.5mm) were well evaluated in 830 coronary segments of 62 cases by DSCT coronary angiography. Wilcoxon signed-rank test result(Z=-4.486, P<0.001) showed two kinds of examination methods had a significant difference in the diagnosis of coronary artery stenosis. DSCT had a lower accuracy in detecting coronary stenosis compared with CAG. The overall sensitivity, specificity, positive predictive value, negative predictive value and accuracy of DSCT coronary angiography for identifying coronary stenosis were 80.2%,95.6%,87.0%,93.0%,91.5%, respectively. Kappa test results(κ=0.655, P< 0.001) showed two kinds of examination methods a moderate consistency. DSCT have a lower accuracy in diagnoseing coronary artery calcification and small branch disease compared with CAG.[Conclusions]Part one:The DSCT angiography could provide optimal visualization of coronary anomalies accurately, because of the improved temporal resolution, spatial resolution and flexible post-processing tool. The DSCT angiography could be considered as a first selective screening method for coronary anomalies.Part two:DSCT and echocardiogram showed a good correlation and concordance in left ventricular function analysis. DSCT examination had great clinical significance in prediction of acute myocardial infarction and assessment of left ventricular function.DSCT left ventricular function analysis and echocardiogram had a good correlation, but these two examination methods didn’t have a good consistency. Two methods couldn’t completely replace each other in the clinical practice. DSCT left ventricular function analysis could be used as a reference in clinical practice.Part three:DSCT coronary angiography could accurately and noninvasively detect significant coronary stenosis. Compared with CAG, DSCT have a lower diagnostic accuracy because of calcification, bifurcation and small branch disease. However, DSCT can still be considered as a preferred method in screening coronary artery disease before CAG examination and stenting follow up.更多还原