【作者】 刘鸿圣；

【导师】 张雪林；

【作者基本信息】 南方医科大学 ， 影像医学与核医学， 2010， 博士

【摘要】 研究背景胆道闭锁(biliary atresia,BA)是一种新生儿及婴儿期的少见畸形,是新生儿阻塞性黄疸最常见的病因,也是造成婴幼儿死亡的主要肝脏疾病。BA的治疗主要是早期手术,重建胆道,恢复胆流,促进预后,手术最佳时间为40～60天,60天内手术生存率可达75%,而90天后接受治疗者生存率仅10%,未接受手术者,患者存活时间一般不超过一年,故早期诊断早期治疗关系到患儿的生存质量,意义重大。如何早期准确诊断BA一直是广大儿科医生普遍关注和研究的课题,目前临床诊断BA的方法虽多,但均存在明显不足。主要包括：(1)B超：其诊断BA特异性尚可,敏感性不足。(2)经皮肝穿刺胆管造影或内镜下逆行胰胆管造影：要求苛刻,成功率低,临床已很少应用。(3)腹腔镜检查：虽能明确诊断但为创伤性手段。(4)肝穿刺病理组织学检查,创伤性检查对BA的早期鉴别诊断有一定意义。(5)动态持续十二指肠液检查,受引流技术限制。(6)单光子发射计算机断层扫描检查,特异性差,假阳性率高。(7)磁共振胆胰管成像显示婴幼儿胆道系能力弱,且婴幼儿难以配合检查,应用受限。(8)多层螺旋CT胰胆管成像包括胆道对比剂的CT胆道造影和非胆道对比剂的阴性法CT胰胆管成像(negative CT cholangiopancreatography,N-CTCP)。随着多层螺旋CT及64层CT的广泛应用,N-CTCP整体显示胆系能力强,图像质量达到甚至超过MRCP水平。迄今为止,国内外尚未见采用N-CTCP诊断婴儿胆道闭锁的报道。本研究采用东芝64层螺旋CT对胆道闭锁婴儿行N-CTCP成像,充分发挥64层CT扫描速度极快、大范围的高分辨率各向同性成像及强大后处理等优势,并采用胆道容积重建(volume render, VR)、多平面重建(multiplanar reconstruction,MPR)及最小密度投影(minimum intension prejection,MinIP)等图像后处理,总结分析胆道闭锁的N-CTCP影像特点,N-CTCP诊断胆道闭锁的敏感性、特异性和准确性,探讨N-CTCP在胆道闭锁诊断中的应用价值。第一部分64层CT阴性法胆胰管成像技术参数研究研究目的优化婴儿采用Toshiba Aquilion 64层螺旋CT行N-CTCP检查,在合理使用低剂量的原则下获取优质图像的合理化技术参数。材料与方法1.研究对象全部对象来自2008年11月至2009年3月来广州市妇女儿童医疗中心门诊或住院就诊者,包括临床确诊为非肝、胆疾病者、胆道扩张及疑诊胆道闭锁者共40例。对象随机分为2组,第1组为高清组,其中男13例,女7例,平均年龄63.84±36.053天；第2组为标准组,其中男9例,女11例,平均年龄64.30±32.916天。2.仪器设备采用Toshiba Aquilion 64层螺旋CT,东芝公司提供的vitrea2.0图像工作站。3.检查方法及参数所有检查者检查前空腹3～4小时,检查前15～30min饮用约250ml清水,扫描前口服水合氯醛(0.5ml/kg体重)镇静。采用自动管电流调节技术(ATCM),第一组采用Sure Exp 3D中"High Quality"(高清组)档次像质曝光成像,第二组采用Sure Exp 3D中"Standard"(标准组)档次像质曝光成像。两组扫描主要技术参数统一,扫描范围肝上缘至髂前上棘。所有对象行腹部CT平扫及动脉期、门脉期、平衡期增强扫描。4.后处理及分析CT平扫及增强扫描各期容积数据传入vitrea2.0工作站后行多平面重建(MPR)。3位资深放射医师根据盲法采用优劣3级评分法分别评价平扫与各期增强容积数据所生成MPR图像的质量。优为图像层次、对比度优良,胆管壁轮廓光滑、胆管假性缺损或狭窄少,1级分支之间辨认清晰；合格为图像层次、对比度合格,胆管壁轮廓光整,胆管假性缺损或狭窄较多,1级分支之间可分辨；劣为缺乏层次和对比度,胆管壁轮廓毛糙,假性缺损、狭窄明显,1级分支之间分辨不清。5.统计学处理：所有统计学处理均采用SPSS 13.0软件完成,所有计量资料用均数±标准差表示,P<0.05认为有统计学意义。运用卡方检验比较高清组及标准组研究对象的性别及疾病分布,运用Kruskal-Wallis H检验比较两组对象平扫及各期增强的图像质量,运用独立样本t检验比较两组对象的年龄及放射剂量。结果1、高清组和标准组病例间样本性别差异χ2=1.616,P>0.05,差异无统计学意义；二者疾病分布比较χ2=3.582,P=0.167>0.05,差异亦无统计学意义。2、高清组扫描及标准扫描图像质量差别Z=-0.548,P=0.584>0.05,无统计学意义。高清组及标准组间各个扫描期图像质量有统计学差异(高清组χ2=61.297,P=0.000；标准组：χ2=40.497,P=0.000),高清组和标准组各期图像质量结果显示,门脉期图像质量优于其他各期。3、高清组和标准组方差齐性条件下二者扫描剂量结果行两独立样本t检验,t=14.310,P=0.000<0.05,差异有统计学意义,标准组剂量低于高清组。结论1.按照Sure Exp 3D中"High Quality"(高清组)档次像质曝光成像组扫描与按照Sure Exp 3D中"Standard"(标准组)档次像质曝光成像组扫描图像质量无差别。2.高清组和标准组各期图像质量结果显示,门脉期图像质量优于平扫、动脉期及平衡期。3.高清组扫描放射剂量高于标准组,二者差异具有统计学意义。4.婴儿N-CTCP检查采用自动管电流调节技术ACTM标准组扫描条件于门脉期扫描即可获取满意质量的图像。第二部分正常婴儿64层CT的N-CTCP表现及正常值研究研究目的：采用合理化参数N-CTCP成像,评价正常婴儿的N-CTCP表现及其胆系显示能力,分别测量胆囊宽径、胆总管及左右肝管直径。材料与方法：1.研究对象全部对象来自2009年4月至2009年12月来广州市妇女儿童医疗中心门诊或住院就诊,行腹部增强CT检查者,临床无阻塞性黄疸、无胆道梗阻、非胆道疾病或体检者共30例。2.仪器设备采用Toshiba Aquilion 64层螺旋CT,东芝公司提供的vitrea2.0图像工作站。3.检查方法及参数对象准备同第一部分,扫描参数采用自动管电流调节技术,Sure Exp3D"Standard"档次象质曝光成像,行CT平扫后于门脉期增强扫描。4.后处理及分析CT平扫及增强扫描各期容积数据传入vitrea2.0工作站后行MPR、VR及MinIP重建,评价正常婴儿的N-CTCP表现,分别测量胆囊宽径、左、右肝管及胆总管内径。5.统计学处理：所有统计学处理均采用SPSS 13.0软件完成,所有计量资料用均数±标准差表示,P<0.05认为有统计学意义。采用VR、MINIP、MPR各种重建方法重建左右肝管及胆总管,采用卡方检验比较比较胆总管、左右肝管的显示率差异；采用单向方差分析比较左、右肝管及胆总管内径的差别。结果：30例正常婴儿全部肝脏大小正常,胆囊可见,宽径约为8.875±2.238mm,肝门区结构清晰,未见异常密度,门脉间隙未见增宽。部分婴儿肝外胆管及左、右肝管显示,主干边缘光整,无一例可见肝内胆管分支。左肝管内径、右肝管内径及胆总管内径分别为1.61±0.40mm、1.59±0.38mm及1.73±0.44mm。胆总管及左右肝管显示率比较,χ2=2.102,v=2,P=0.350,差异无统计学意义。结论：1.正常婴儿肝外胆管可不显示,显示者胆管主干连续,边缘光滑；2.正常婴儿肝内胆管分支在各种(VR、MPR、MINIP)重建方法中均无法显示。3.正常婴儿左、右肝管显示率与胆总管显示率无差别。4.左右肝管、胆总管内径在N-CTCP上差异无统计学意义。第三部分胆道闭锁婴儿64层CT阴性法胰胆管成像研究研究目的：评价胆道闭锁婴儿的N-CTCP影像表现,探讨其诊断价值。材料与方法1.研究对象选择2009年4月至2009年12月来广州市妇女儿童医疗中心门诊或住院就诊,临床疑诊为胆道闭锁行腹部增强CT检查者共52例,其中男30例,女23例,平均年龄63.34±29.18天。其中32例经手术病理证实为胆道闭锁,20例临床确诊为婴儿肝炎综合征。2.仪器设备同第一部分。3.检查方法及参数对象准备及扫描方法、扫描参数同第二部分。4.后处理及分析CT平扫及增强扫描各期容积数据传入vitrea2.0工作站后分别行MinIP、VR、MPR重建。5.统计学分析所有统计学处理均采用SPSS 13.0软件完成,所有计量资料用均数±标准差表示,P<0.05认为有统计学意义。采用One-Way ANOVA及LSD法比较BA组、IHS组及正常婴儿组的胆囊宽径大小,进一步绘制ROC曲线,确定阳性事件(BA)的胆囊宽径的最佳界点。采用卡方检验比较BA组及IHS组N-CTCP各种影像征象阳性率差异。比较胆道闭锁所见各种征象的敏感性、特异性及准确性。采用Mann-Whistney U检验比较胆道闭锁三角征、小胆囊、双边征三种征象在MPR、MinIP及VR三种重建方法中的显示率差异。结果：1.52例临床疑诊BA患儿,病理确诊32例为BA,临床确诊20例为婴儿肝炎综合征(IHS),其N-CTCP诊断结果与病理手术结果比较,κ=0.718,提示N-CTCP诊断结果与病理学结果吻合度好。2.胆道闭锁的N-CTCP影像征象包括：1)无或小胆囊(胆囊宽径小于6.3为小胆囊)；2)肝门区出现三角形低密度区(“三角征”)；3)门静脉主干及分支间隙增宽,可见“双边征”；4)胆总管未见显示,左、右肝管无显示(5)肝、脾肿大。小胆囊(χ2=10.833,P=0.001<0.05)、三角征(χ2=14.444,P=0.000<0.05)及门静脉间隙增宽(χ2=11.123,P=0.001<0.05)在BA及IHS患儿中差别有统计学意义,而左右肝管、胆总管无显示(χ2=0.075,P=0.784>0.05；χ2=0.328,P=0.068>0.05)、肝脾肿大(χ2=0.033,P=0.857>0.05)两者差别无统计学意义。3.“三角征”、小胆囊、双边征在MINIP及VR的显示率比较χ2=7.082,P=0.029<0.05,差异有统计学意义,其中MINIP重建法阳性征象显示率高于VR及MPR。4.“三角征”特异性高达100%,但敏感性低(50%)；小胆囊(包括无胆囊)敏感性(90.3%)、特异性(85.0%)及准确性(86.5%)均较高；双边征特异性(90.0%)亦较高,但敏感性(56.3%)及准确性(69.2%)较低。胆总管、左右肝管未见显示及肝脾大诊断的特异性及准确性均较低；三角征、小胆囊、双边征这三种影像征象对鉴别诊断有意义。结论：1.N-CTCP诊断结果与病理手术结果比较,两种诊断方法的吻合度较强。2.肝门区三角形低密度影(三角征)、小胆囊、门静脉间隙增宽出现“双边征”是诊断BA的重要征象。3.胆道闭锁的三角征、小胆囊、双边征三种阳性征象在MINIP中的显示率高于VR及MPR。更多还原

【Abstract】 BackgroundBiliary atresia (BA) is a rare abnormality in new born babies or infants, but is the most common cause for infant obstructive jaundice, and the major hepatic disease that causes infant death. The treatment of BA is mainly early operation, with biliary reconstruction to reintegrate biliary passage,to challenge better prognosis. The best operative time is 40～60 days after birth. Survival rate can reach 75% by operation within 60 days after birth, however only 10% after 90 days after birth. If haven’t operation, patient can’t surpass one year of life time.Prompt diagnosis and treatment will largely affect the infants’survival qualities, which may have great significance.How to make accurate diagnosis of BA is paid widespread attention by most pediatric doctors on the long way. Although there are many diagnostic methods for BA evaluation, there exist many shortcomings, as like as:(1)Type-B ultrasonography, with good specificity but short of sensitivity. (2) Percutaneous transhepatic cholangiography and endoscopic retrograde cholangiopancreatography require strict manipulation, with low achievement ratio, which are already rarely used clinically.(3) Peritoneoscope can provide accurate diagnosis, but have invasive drawbacks.(4) Needle biopsy of liver also an invasive method, which can help early differential diagnosis.(5)Persistent duodenal secretion examination is impeded by drainage technique.(6)SPECT examination has low specificity and high false positive rate.(7)MRCP, the ability of displaying biliary system is still unsatisfactory, and infants can’t cooperate well, which limit its practicing.(8)multi-slice CT cholangiopancreatography includes positive CT cholangiopancreatography with biliary contrast agent and negative CT cholangiopancreatography(N-CTCP) without biliary contrast agent. With widely used of multi-slice CT and 64-slice CT, N-CTCP have great ability of displaying the total biliary system, and its image quality reaches or even surpasses that of MRCP.Up till now, there are no literatures on diagnosis of biliary atresia by using N-CTCP from domestic and abroad.We exam the suspected biliary atresia patients by using N-CTCP by Toshiba 64-slice spiral CT, and take advantage of its rapid scanning speed, large coverage with high resolution, isotropic imaging and great reconstruction abilities. By biliary system volumn rendering(VR), multiplanar reconstruction(MPR),minimum intension projection (MinIP)and other post-reconstruction methods, summarize the N-CTCP characteristics of biliary atresia, and evaluate its sensitivity, specificity and accuracy, to explore its practicing value in diagnosis of biliary atresia.PART ONE A Study on 64-slice Negative CT Cholangiopancreatography Scanning ConditionObjectivesOptimize reasonable scanning conditions of N-CTCP with good image quality under the conception of ALARA(as low as reasonably achievable)by using Toshiba Aquilion 64-slice spiral CT. Material and Methods1.Subjects:All subjects were from out-patients or in-patients of Guangzhou Women and Children’s Hospital during November 2008 to March 2009, including healthy human beings or chidren suspected abdominal diseases, excluding suspected biliary atresia.40 peoples were recruited in this study, randomly allocated into two groups. The first group was the high quality group, among them there were 13 male patients and 7 female patients, averaging 63.84±36.053 birth days;the second group was the standard group,among them there were 9 male patients and 11 female patients, averaging 63.34±29.18 birth days.2.Equitment:All studies were done using Toshiba Aquilion 64-slice spiral CT, and post-reconstruction work were done on the vitrea2.0 workstation provided by Toshiba.3.Scan methods and Conditions:All patients should hold empty stomach for 3～4 hours before scanning, drink 250ml water 15～30min before scanning, and given oral chloral hydrate(0.5ml/kg) for sedation.By using the technique of 3-D automatic tube current modulation (ATCM), the first group were scanned by "High Quality"exposure condition in Sure Exp 3D,and the second group were scanned by "Standard Quality" exposure condition in Sure Exp 3D.The sanning parameters were unified in these two groups, with the scanning scope from the upper limit of the liver to the spina iliaca anterior superior. All the patients had the abdomen CT plain scan, arterial phase, portal vein phase and equilibrium phase after injection of contrast agent.4.Post-reconstruction and Analysis All volume data of CT plain scan and contrast scan of all different phases were transferred to vitrea workstation to perform multi-planar reconstruction (MPR).There sophisticated radiologists double-blindly evaluated the MPR images quality from plain scan and contrast scan of all different phases. Pictures showing smooth biliary wall,with little false defect or stenosis, and with clear first scale branch were classified as good images, while showing coarse biliary wall,with much false defect or stenosis, and with unclear first scale branch were classified as bad images.5.Statistic analysis:Statistic analyses were performed with SPSS 13.0 software.All measurement data were recorded as standard±deviation. And statistic significance is defined as P<0.05.Sex and distribution of diseases between the high quality group and the standard group were compared by Chi-square test. Image quality of plain scan and contrast scan from every phase between the two groups were compared by Mann-Whitney test and Chi-square test. And radiation doses were compared by Chi-square test and by least significance difference(LSD)-t test(Not Paired t test)Result1.No sex deviation was found between the high quality group and the standard group(P>0.05).No difference of distribution of diseases was found between the two groups(Z=-0.848,P>0.05).2.No diifference of image quality was found between the high quality group and the standard group(P=0.146>0.05).Diifference of image quality was found in different phases of scanning in the two respective groups(P=0.000<0.05),which can be concluded that image quality from the portal phase was much better than other phases.3.With homogeneity of variance, difference of radiation doses between the high quality group and the standard group were found by least significance difference(LSD)-t test(P=0.000<0.05)Conclusion1.No difference was found in the image quality between of the high quality group and the standard group classified based on the Sure Exp 3D exposure technique.2.In all phases from the high quality group and the standard group,image quality from the portal phase was much better than other phases.3.Radiation doses of the high quality group were much higher than that of the standard group.Difference was significant by statistical analysis.4.Satisfactory images could be obtained in portal phase with standard scanning condition by using the technique of 3-D automatic tube current modulation (ATCM) in infants.PART TWO A Study on 64-slice Negative CT Cholangiopancreatography in Normal InfantObjectivesBy using reasonable conditions of N-CTCP, evaluate the presentation of N-CTCP and its biliary system showing abilities in the normal infant group.Material and Methods1.Subjects:All subjects were from out-patients or in-patients of Guangzhou Women and Children’s Hospital during April 2009 to December 2009, who had abdominal contrast scan.We chose from them excluding obstructive jaundice, biliary obstruction or other biliary diseases.30 subjects were recruided in this study.2.Equitment:All studies were done using Toshiba Aquilion 64-slice spiral CT, and post-reconstruction work were done on the vitrea2.0 workstation provided by Toshiba.3.Scan methods and Conditions:Preparation, scanning method and scanning condition are the same as that in Study Part One.4.Post-reconstruction and AnalysisAll volume data of CT plain scan and contrast scan of all different phases were transferred to vitrea workstation to perform multi-planar reconstruction (MPR) and the features of N-CTCP in normal infants.5.Statistic analysis:Statistic analyses were performed with SPSS 13.0 software. All measurement data were recorded as standard±deviation. And statistic significance is defined as P<0.05.Result30 cases of normal infants all showed normal hepatic size, with filling gallbladder and clear configuration of porta hepatic, exceptionally with abnormal densities and widened interspace of portal vein. Part of them displayed extrahepatic bile ducts and left or right hepatic ducts, with smooth main stems. No one could show the second scale branches of hepatic ducts.Conclusion1.Normal infants can either display extrahepatic bile ducts or not, when they do, they would display continuous bile stem and smooth margin.2.The second scale branches wouldn’t be displayed in normal infants using different reconstruction methods.3.There is no difference between in the display rate of common bile duct and that of bilateral hepatic duct in normal infants. 4. There is no statistical significance in the inner diameter of common bile duct, right and left hepatic duct.PART THREE A Study on 64-slice Negative CT Cholangiopancreatography in Biliary atresia InfantObjectivesEvaluate the imaging presentation of N-CTCP in biliary atresia infant group, and investigate its diagnositic value.Material and Methods1.Subjects:All subjects were from out-patients or in-patients of Guangzhou Women and Children’s Hospital during April 2009 to December 2009, who had abdominal contrast scan and suspected biliary atresia.52 subjects were recruided in this study. Among them,32 cases were proven as biliary atresia by post-surgery pathology, and 20 cases were proven as infantile hepatitis.2.Equitment:The same as Study Part two.3.Scan methods and Conditions:Preparation, scanning method and scanning condition are the same as that in Study Part Two.4.Post-reconstruction and AnalysisAll volume data of CT plain scan and contrast scan of all different phases were transferred to vitrea workstation to perform MinlP.There sophisticated radiologists double-blindly evaluated the MinIP images quality from plain scan and contrast scan of all different phases. Evaluating standard were the same as that of Study Part One. 5.Statistic analysis:Statistic analyses were performed with SPSS 13.0 software. All measurement data were recorded as standard±deviation.Evaluate gallbladder sizes and displaying ratio of and biliary structures (including intrahepatic bile ducts and left or right hepatic ducts) in N-CTCP from all cases.And compare the sensitivity, specificity, and accuracy of different biliary atresia signs.Compare image quality from the three different reconstruction methods including MPR, MinIP and VR by Mann-Whitney test.Result1.No difference between dignostic evaluation and post-surgery pathology were found in 32 biliary atresia patients, by McNemar test P=1.000>0.05.2.Imaging feature of biliary atresia include:1)No gallbladder or small gallbladder;2)Appearing triangular low density around the porta hepatic area;3)Widened interspace of portal vein, appearing as"double-track sign"or"target sign";4)No common bile duct were recognizable in all biliary atresia patients;5) Hepatosplenomegaly. Difference of small gallbladder, triangular sign and widened interspace of portal vein were found between BA and HIS patients by statistical analysis, while no difference of any recognizable extrahepatic bile ducts or hepatosplenomegaly were found between them.3.Specificity of triangular sign in biliary atresia patients was as high as 100%, but with low sensitivity. Sensitivity, specificity, and accuracy of small gallbladder sign (including no gallbladder sign) were comparatively high. Specificity of double-track sign was also high, but sensitivity and accuracy were relatively low. Therefore, signs of hepatosplenomegaly and no recognizable extrahepatic bile ducts are not so valuable for diagnosis of biliary atresia. Signs of triangular low density, small gallbladder, double-track and no recognizable extrahepatic bile ducts are helpful for conforming diagnosis, particularly triangular sign.Conclusion1.By comparing N-CTCP diagnostic and post-surgery pathology results, goodness of fit of the two diagnostic methods is significant and is rather strong.2.Triangular low density around the porta hepatic area, small gallbladder, widened interspace of portal vein, appearing as"double-track sign"or"target sign" are the key signs of diagnosis of BA, while which could be excluded if the common bile duct is recognizable.3.Among the MPR, MinIP and VR reconstruction methods, MinIP showed best in infants of BA.更多还原