节点文献
64层螺旋CT三维重建盆腔动脉的形态学研究及临床应用
Morphological Study and Application of3D Reconstruction of Pelvic Arteries by64-slice Spiral CT
【作者】 夏红;
【导师】 何援利;
【作者基本信息】 南方医科大学 , 妇产科学(专业学位), 2013, 博士
【摘要】 研究背景女性盆腔动脉是一个庞大复杂的动脉系统,尤其是髂内动脉分支密集、起源复杂、迂曲多变、动脉细小,一直是解剖学和妇产科研究的难点及重点。不同个体之间盆腔动脉存在差异,个别甚至出现明显的变异,很难形成一个统一、固定的标准,因而对盆腔动脉系统的认识需要提高到个体化和立体化的概念上。对女性盆腔动脉的研究,主要有彩色多普勒超声、数字减影血管造影(Digital subtraction angiography, DSA)、磁共振血管造影(magnetic resonance angiography, MRA)和计算机X射线断层血管造影(computed tomography angiography, CTA)。彩色多普勒超声检查无创经济,但不能连续立体的显示盆腔血管图像,成像质量受检查者技术的影响较大。DSA一直被作为动脉研究的“金标准”。但DSA是有创性检查,操作复杂,临床上仅用于计划行选择性动脉栓塞的病人中。磁共振血管造影(magnetic resonance angiography, MRA)和计算机X射线断层血管造影(computed tomography angiography, CTA)为在体血管的研究提供了很好的研究手段和研究基础。近年来,有学者利用MRA技术对在体盆腔动脉血管网的构建进行了有益的探索。2008年Naguib等利用MRA技术成功构建了拟行子宫动脉栓塞术(uterine artery embolization, UAE)患者的盆腔动脉血管网,将MRA图像与UAE插管密切相关的髂内动脉分支相比,发现具有很高的符合率。2010年Mori等对比非增强MRA与DSA中子宫动脉起源的显示率,结果MRA中有97%子宫动脉起源显示清晰,认为利用MRA构建的盆腔动脉血管网可以较好的显示髂内动脉、子宫动脉等与UAE密切相关的粗大血管。但MRA成像区域相对较小,难以同时显示盆腔外的供血情况,如一些子宫肌瘤存在来自肾动脉来源的卵巢动脉供血,盆腔MRA图像难以将肾脏和盆腔图像同时包括。而且MRA所示的血管图像,仅显示血管,周围骨性参照以及病种图像难以同时显示,使得血管的辨认过程较为困难。CTA是在螺旋CT的基础上发展起来的一种崭新的血管成像技术,由于拥有相对无创、简便快捷、信息量大等优点,目前正广泛应用于临床中。特别是64层螺旋CT由于达到了真正的各向同性,使得CTA有了长足的发展,对疾病的诊断有了更加广泛的空间。到目前为止,已有不少学者利用该技术对人体多个部位如头颅、甲状腺等的动脉血管网进行了详细的解剖学研究。更有研究表明其在心脏及脑动脉血管网的显示方面基本可以替代传统的DSA技术。借鉴CTA技术在心脏及脑动脉等血管网中的成功应用,为人体盆腔血管形态学研究提供了新的研究手段和发展空间。到目前为止,将该技术应用于盆腔血管形态学的研究并不多。2006年国内学者使用16层螺旋CT构建了16名正常受试者(不分男女)盆腔动脉血管网模型的3-4级分支,但由于CT设备的限制,未能显示出较为细小的子宫动脉血管网。2011年另一学者基于64层CTA数据集,利用Mimics软件,对170例妇科住院病人的CTA进行女性髂内动脉解剖学分型,但未提供盆腔内其它细小侧枝血管的解剖、变异和影像学资料。2011年国外学者通过采用不同的成像方式CTA及DSA)对21例男性髂内动脉主要分支显示情况进行对比研究,认为CTA较MRA可以更好地显示精细的盆腔动脉血管网,较DSA更加快捷。该研究提供了一定的影像解剖资料,特别是对盆腔动脉的细小侧枝血管,如闭孔动脉、髂腰动脉等进行了详实的描述。遗憾的是,该研究受试对象仅为男性。因此,对于女性盆腔动脉血管网详实的CTA影像解剖资料研究,至今未见报道。为此,我们在原有女性在体盆腔血管CTA研究的基础上,利用CT自带三维重建技术,通过观察分析33例成年女性正常盆腔动脉血管的CTA图像,对成年女性正常盆腔动脉的各级分支血管,尤其是细小侧枝动脉的起源、走行及其毗邻结构等进行解剖学特点的研究,以识别盆腔内动脉血管的CTA表现,为三维重建技术应用于女性盆腔疾病的诊断和治疗提供平台,为妇科介入放射治疗及腹腔镜微创手术提供一定的解剖学和影像学依据。第一章64层螺旋CT三维重建盆腔动脉的形态学研究目的:基于CTA数据集,利用CT自带三维重建技术,构建33例成年女性正常盆腔动脉血管网,对女性盆腔动脉的各级分支血管,尤其是细小侧枝动脉的起源、走行及其毗邻结构等进行解剖形态学特点的研究,以识别盆腔动脉血管的CTA表现,为三维重建技术应用于女性盆腔疾病的诊断和治疗提供平台,为妇科介入放射治疗及腹腔镜微创手术提供一定的解剖学和影像学依据。方法:1.研究对象:收集南方医科大学珠江医院青年志愿者、番禺中心医院体检科、泌尿外科及胃肠外科2010年10月至2012年12月期间,因各种原因行中下腹部CTA检查的33例成年女性正常盆腔(无盆腔疾病和手术史,CTA结果提示无盆腔病变)的CTA数据。年龄20岁~61岁,平均(37.36±11.97)岁。所有受试对象均行64排螺旋CT扫描并进行盆腔血管三维重建,数据采集前均未接受放化疗或手术。该研究遵循的程序符合我院负责人体试验的委员会所指定的伦理学标准并得到该委员会的批准,并签署知情同意书。2.设备:(1)CT扫描仪:日本东芝Aquilion64层螺旋CT扫描仪(探测器组合为0.5mm×64)。(2)双筒高压注射器:美国MEDRAD公司。(3)图像后处理工作站:Aquilion64层螺旋CT自带StartVitrea2后处理工作站。(4)优维显(370mg/ml):德国先灵公司生产。(5)CT阅片软件:中国华海医信Med Viewer影像工作站。3.CT扫描参数设定及扫描方法:检查前禁食4h-6h,适度充盈膀胱(未予其他特殊处理)。扫描时患者取仰卧位,身体正中矢状面与床面垂直,且位于床面中央,双手抱头,双腿伸直并拢,扫描范围从第4腰椎中段至股骨大转子下3cm。扫描条件:管电压120KV、管电流250mAs,采用0.5×64排探测器组合,层厚0.5mm、层间距0.3mmm,准直螺距0.984,球管旋转一周时间0.4s。先行常规平扫,再用高压注射器经肘正中静脉进行团注非离子碘造影剂优维显370(Ultravist含碘370mg/ml),剂量1.5ml/kg,注射速率4.5ml/s,对比剂注射完毕后再以相同速率加注生理盐水20ml。应用对比剂示踪法在腹主动脉分叉上缘层面选择感兴趣区(Region of Interest, ROI),动态监测CT值,设定当ROI内CT值达到200Hu后3s为采集点,动脉期扫描延迟时间为25s。4.后期处理:将扫描后的断层图像发送至StartVitrea2后处理工作站,重建方法为:①最大密度投影法(maximum intensity projection, MIP);②多平面重建(Multi-Planar Reconstruction, MPR);③容积再现技术(volume rendering, VR)。根据以上三种重建方法,调整不同域值和模式,重建出满意的CTA图像,依据系统解剖学知识,对影像在三维空间做任意角度的观察,分析盆腔血管的全貌,观察盆腔动脉的显示情况和形态学差异。5.影像判断标准及分析项目:全部动脉血管鉴定均参考妇产科临床解剖学图谱、人体断层解剖学图谱及妇产科放射介入治疗学图谱等。全部CTA图像均经两位影像科资深医生评估并达成一致。分析盆腔动脉血管网的各级分支显示情况,包括双侧髂总动脉的长度,双侧髂内动脉的开口位置和分支类型,双侧子宫动脉的起源、开口显像类型及走形,两侧臀上动脉、臀下动脉、阴部内动脉、髂腰动脉、闭孔动脉、脐动脉、骶外侧动脉、骶正中动脉、卵巢动脉等的起源、走形、及毗邻情况。6.统计学分析;应用SPSS13.0软件对数据进行统计分析。计数资料用百分率表示;两样本率的比较采用R×C表配对资料的χ2检验及Fisher’s确切概率检验。计量资料以均值±标准差(x±s)表示,样本均数比较采用独立样本的t检验,以P≤0.05为有统计学意义。结果:1.成年女性盆腔动脉血管网的CTA构建情况:33例全部成功构建出完整的盆腔动脉血管网图像。该血管网图像清晰逼真、形态规则、走行自然、立体感强。可以清楚地显示生理状态下的腹主动脉、直肠上动脉、髂总动脉、骶正中动脉、髂外动脉、髂内动脉、髂内动脉前干的子宫动脉、闭孔动脉、脐动脉、膀胱上动脉及膀胱下动脉、髂内动脉后干的髂腰动脉、臀上动脉、臀下动脉、骶外侧动脉、阴部内动脉、股动脉及其分支的腹壁下动脉、旋髂浅动脉、旋髂深动脉、旋股内动脉、阴部外动脉、股深动脉、股外侧动脉及其分支等。通过对CTA图像进行任意的缩放及3600旋转,能够清晰地辨认出盆腔动脉及其4~5级血管分支的开口、起源、走形及毗邻关系。2.髂总动脉长度变化分型根据左右髂总动脉的长度及两者间的关系,分为5种类型:(1)两髂总动脉颇长型:3例,占9.1%。(2)两髂总动脉甚短型:1例,占3.0%。(3)髂总动脉左长于右型:13例,占39.4%。(4)髂总动脉右长于左型:1例,占3.0%。(5)左右髂总动脉等长型:15例,占45.5%。3.髂内动脉的解剖(1)髂内动脉的开口位置及高度测量:髂内动脉开口位于第五腰椎体上1/2至第一骶椎下1/2之间。比较两侧髂内动脉开口的高度,右侧高于左侧者占42.4%(14/33),左侧高于右侧者占6.1%(2/33),双侧同一高度者占51.5%(17/33)。(2)髂内动脉的分支类型:参考Adachi分型方法,依据髂内动脉的三大分支(臀上动脉、臀下动脉和阴部内动脉)的组合情况,将髂内动脉分为五大类型:Ⅰ型:髂内动脉先发出臀上动脉,后发出臀下阴部干;Ⅱ型:髂内动脉依次发出臀上动脉、臀下动脉和阴部内动脉;Ⅲ型:髂内动脉先发出臀干,后发出阴部内动脉;Ⅳ型:髂内动脉发出一个总干,总干再分出臀上动脉、臀下动脉和阴部内动脉;Ⅴ型:臀下动脉有2支,两支起源不一致,一支起自臀上动脉,另一支与阴部内动脉共干。本研究33例成年女性正常盆腔动脉血管网,仅见Ⅰ型、Ⅱ型及Ⅲ型,未见Ⅳ型及V型。在前3种分型中,左侧髂内动脉分支类型所占比例分别为:63.6%(21/33)、30.3%(10/33)及6.1%(2/33),右侧分别为:69.7%(23/33)、21.2%(7/33)及9.1%(3/33)。其中63.6%(21/33)的患者双侧髂内动脉分型相同,对比双侧髂内动脉各种分型所占的比率,差异无统计学意义(χ2=6.398,P=0.171)(3)髂内动脉后干分支的CTA表现:22例CTA图像(66.7%)可显示出髂腰动脉,它是髂内动脉最高位的分支,最常见起源于髂内动脉主干及后干。24例CTA图像(72.7%)可辨认出骶外侧动脉,该动脉最常见起源于髂内动脉主干末端,其次为臀上动脉和臀下动脉。33例受试者共66支闭孔动脉,55支(83.3%)均可清晰显示。闭孔动脉起源变异较大,绝大部分起自髂内动脉或其分支,其余部位是臀上动脉、臀下阴部动脉干、髂内动脉主干、臀下动脉等。66支(100%)臀上动脉全部显示,臀上动脉为髂内动脉后干的终末分支,也是髂内动脉最大的分支。从CTA图像可见,该动脉绝大部分起自髂内动脉主干,仅7.5%的受试者与臀下动脉组成臀干。(4)髂内动脉前干分支的CTA表现:脐动脉是髂内动脉主干的延续,近侧端发出膀胱上动脉,然后分成2-3支垂向内下,达耻骨上支上方,弯向内下方。66支阴部内动脉均可清楚显示。它是髂内动脉前干两条末梢分支中的较小分支,常与臀下动脉共干,其次是单一干起始,极少数与臀上动脉共干。臀下动脉为髂内动脉前干最大的末梢分支,多数与阴部内动脉共干起自髂内动脉前干,部分以1支单干起自髂内动脉。4.子宫动脉的解剖:(1)子宫动脉的起源:66支(100%)子宫动脉均可显示开口、走行、粗细及主要分支。子宫动脉起源变异比较大,起源于髂内动脉主干31支(左侧15支,右侧16支),占47.0%;起源于臀下动脉阴部干20支(左侧11支,右侧9支),占30.3%;起源于脐动脉11支(左侧6支,右侧5支),占16.7%;起源于阴部内动脉1支(右侧1支),占1.5%;其余起源为臀上动脉或臀下动脉等,共3支(左侧1支,右侧2支),占4.5%。本研究中51.5%(17/33)的受试者左右子宫动脉起源不对称。(2)子宫动脉的开口显像类型:Ⅰ呈锐角51支(左侧25支,右侧26支),占77.4%;Ⅱ呈直角3支(左侧3支,右侧0支),占4.5%;Ⅲ呈旋转状9支(左侧4支,右侧5支),占13.6%;Ⅳ呈螺旋状3支(左侧1支,右侧2支),占4.5%5.卵巢动脉:33例受试者中,显示卵巢动脉者4例,占12.1%。33例受试者共66支卵巢动脉,共有4支卵巢动脉显影(左侧3支,右侧1支)。卵巢动脉的显示率为6.1%(4/66)。结论:1.基于64层螺旋CT数据,应用CT自带软件,可以构建出理想的女性盆腔动脉血管网模型,是研究在体女性盆腔动脉的好方法,值得临床推广。2.基于64层CTA构建的盆腔动脉血管网模型,为盆腔动脉形态学研究提供了一定的影像解剖学依据,为盆腔微创手术和盆腔血管内介入治疗提供个体化的准确的形态学依据,为临床医生对盆腔血管性疾病的诊疗提供了新技术,为个体化治疗的开展奠定了基础。第二章64层螺旋CT三维重建盆腔动脉血管网的临床应用第一节64层螺旋CT三维重建盆腔动脉血管网在产后出血中的应用及意义目的:利用21例保守治疗无效的产后出血患者的64层CTA数据集,采用CT自带软件进行盆腔动脉血管网重建,研究产后出血盆腔动脉血管网构建的方法及意义,观察分析卵巢动脉在产后出血中参与病变供血的情况。方法:1.一般资料:选取2012年1月至2012年12月在番禺中心医院产科住院分娩,因各种原因产后出血保守治疗无效(经药物、按摩子宫、清宫等)而行盆腔64层CTA检查的患者21例(合并胎盘植入3例),全部患者于剖宫产术后或分娩后24h内出血大于500mL,经常规保守治疗后阴道出血未减少或减少后再次出血增多。年龄21岁~34岁,平均(27.24±3.45)岁。其中初产妇12例,经产妇9例;阴道分娩7例,剖宫产14例;宫缩乏力13例,胎盘因素6例,软产道裂伤1例,子宫切口血肿1例。出血量565ml-330ml,平均(813.10±182.08)ml。2.仪器与方法:患者扫描前无需特殊处理。其余均同第一章。3.图像后处理:同第一章。4.统计学方法同第一章。结果:21例患者全部成功构建出完整的盆腔动脉血管网,血管网血管边缘清晰、形态规则、走行自然、立体感强,通过任意缩放及360°旋转观察,可以清晰地显示髂内动脉及其3-4级血管分支,可对包括卵巢动脉在内的所有盆腔血管进行三维显像。21例患者中,有12例患者的CTA图像中可显示出卵巢动脉,占57.1%,21例患者共42支卵巢动脉,其中两侧显示卵巢动脉共12支,左右单侧显示共6支,卵巢动脉的显示率为42.9%(18/42)。在显示卵巢动脉的12例患者中,CTA图像见卵巢动脉管径增粗,直径在2.0mm~11.6mm之间,其中2例患者OA异常怒张,见卵巢静脉提前显影,提示存在卵巢动静脉畸形。结论:1.基于64层CTA数据集可以构建出理想的产后出血患者的盆腔动脉血管网,为产后出血患者的诊疗提供了新技术,为介入手术前评估提供了解剖学依据。2.卵巢动脉参与产后病变的供血,可能是导致常规栓塞两侧髂内动脉-子宫动脉不能控制出血的原因之一。第二节64层螺旋CT重建供血动脉在巨大盆腔肿瘤诊断中的应用目的:利用64层螺旋CTA对经手术病理证实的巨大盆腔肿瘤进行盆腔动脉三维重建,通过观察肿瘤的血管及供血动脉显示情况,评价该技术在巨大盆腔肿瘤诊断中的价值,为临床治疗方案的制定提供帮助。方法:1.一般资料搜集2010年10月至2012年10月于我院行64层螺旋CTA检查,并经手术和病理证实的11例女性盆腔巨大肿瘤患者临床资料,年龄27岁-54岁,平均(41.73±8.58)岁。肿块大小8.5cm~27.0cm,平均(13.64±6.07)cm。子宫肌瘤4例,卵巢肿瘤7例(包括浆液性囊腺瘤1例,交界性粘液性囊腺瘤1例,浆液性囊腺癌2例,粘液性囊腺癌2例,转移性卵巢癌1例)。临床症状:腹部包块、腹胀腹痛、下坠感、月经异常、不规则阴道流血等。所有患者均签署知情同意书。2.仪器与方法:同第一章。3.图像后处理:同第一章。4.供血动脉判断标准:动脉增粗,分支伸入瘤体内呈网格状或放射状分布者,称为肿瘤供血动脉。肿瘤主要由子宫动脉子宫支(uterine branch, UB)供血被认为肿瘤来源于子宫。肿瘤主要由卵巢动脉(ovarian artery, OA)和/或子宫动脉卵巢支(ovarian branch, OB)供血,被认为肿瘤起源于卵巢。5.统计学方法:同第一章。结果:1.11例患者均成功构建出完整的盆腔动脉血管图像,血管边缘清晰、形态规则、空间立体感强、走行自然,可任意缩放,360°旋转观察盆腔血管及供血动脉的情况。11例患者血管图像均能清楚的显示腹主动脉、髂总动脉、髂外动脉、髂内动脉、子宫动脉、子宫动脉卵巢支或卵巢动脉。11例患者均显示出供血动脉。2.巨大子宫肌瘤:4例。肿瘤最大直径8.5cm~27.0cm,平均(14.63±8.38)cm。血管后处理重建均显示肿瘤存在单侧或双侧UB供血,发生率为100%(4/4),其中1例由双侧UB及右侧OA共同参与供血,发生率25.0%(1/4),表现为双侧UB及OA末梢分支于瘤体。肿瘤定位诊断准确率为100%(4/4)。3.巨大卵巢肿瘤:7例,肿瘤最大直径9.0cm~22.0cm,平均(13.07±5.04)cm。7例患者均显示供血动脉,其中两侧显示3例,单侧显示4例,共10侧。血管后处理重建显示UB、OB及OA共同参与供血2侧,OA、OB共同供血3侧,OA供血1侧,OB供血4侧。OA供血发生率为50.0%(5/10),OB供血发生率为90.0%(9/10)。肿瘤定位诊断准确率为100.0%(10/10)。结论:基于64层螺旋CT数据,利用MPR及VR等三维重建技术,能清楚地显示巨大盆腔肿瘤的供血动脉,对巨大盆腔肿瘤的定位诊断提供更准确的诊断依据。第三节剖宫产术后子宫疤痕妊娠盆腔动脉血管网的构建及意义目的:利用CTA三维重建技术,构建剖宫产术后子宫瘢痕妊娠(cesarean scar pregnancy, CSP)患者的盆腔动脉血管网模型,探讨该血管网在CSP中的方法及意义。方法:1.一般资料选取2010年10月至2012年10月在番禺中心医院妇科住院并行盆腔CTA检查的5例CSP患者的CTA数据集。5例患者年龄28岁~37岁,平均(33.40±3.58)岁。孕次3次~6次,平均(4.20±±1.30)次。产次1次~3次,平均(1.80±0.84)次。5例患者均有停经史,最短34天,最长112天,平均(68.6±31.9)天,此次妊娠距前次剖宫产间隔时间最短6个月,最长11年,平均(4.94±4.32)年。2.诊断标准:①剖宫产史和停经史;②阴道流血:自然情况下淋漓出血或大出血,人流术中大量出血不止药流无胚胎组织排出,清宫时大出血不止;③腹痛:无或轻微;④血清β-hCG升高;⑤超声或彩色多普勒检查:宫腔及宫颈管内均未见妊娠囊、妊娠囊位于子宫峡部且峡部肌层变薄;或妊娠囊和膀胱之间缺乏子宫肌层或肌层有缺陷;彩色多普勒提示孕囊或包块周围血流丰富、高速。3.治疗经过:1例患者合并不完全子宫破裂急诊行子宫下段疤痕部位切开取胚术,1例行腹腔镜下子宫下段疤痕部位切开取胚术,2例接受MTX治疗。1例因经济困难要求返回当地医院治疗。4.仪器与方法:同第一章。5.图像后处理:同第一章。6.统计学方法同第一章。结果5例CSP患者均成功构建出完整的盆腔动脉血管网图像,图像血管边缘清晰、形态规则、空间立体感强、走行自然,不仅可以清楚地显示骨盆的骨性结构、骶髂关节、骶尾关节、耻骨联合等,也能够真实地再现生理状态下的髂总动脉、髂外动脉、髂内动脉、子宫动脉等血管分支。通过对血管网立体结构任意缩放,360°旋转,可以清楚的显示疤痕妊娠病灶的血管分布、血供丰富程度。结论:基于64层CTA数据集可以成功构建出CSP盆腔动脉血管网模型,为CSP患者选择个体化的治疗方式提供参考。第四节64层螺旋CT血管重建在子宫动静脉畸形中的应用目的:通过64排螺旋CT与三维重建技术诊断2例子宫动静脉畸形(uterine arteriovenous malformation,UAVM)患者的资料,旨在探讨64排螺旋CT及图像后处理技术对UAVM的临床价值。方法:一般资料:选取2例UAVM患者的64排螺旋CT数据集.例1:38岁,因“不规则阴道流血2个月,加重1天”入院。患者G2P1A1,2年前因左宫角妊娠行人流术,因组织物残留又行两次清宫,清宫后彩超CDFI呈五彩镶嵌信号。入院后彩超示左宫角内液性暗区,CDFI呈五彩镶嵌信号。盆腔CT左附件-子宫壁间见不规则团状异常粗大瘤样血管影。动脉期血管即显影且浓密,供血源于左侧髂内动脉分支,经明显增粗的左侧卵巢静脉回流入左侧肾静脉,薄层重建后确诊为UAVM。初步诊断:不规则阴道流血查因:子宫动静脉畸形。患者坚决要求切除子宫。术中见左侧宫角处见一圆形腔隙,25mm×25mm,与宫腔有管道相通,左侧宫角处肌壁间血管丰富,左侧阔韧带内血管、左侧子宫动静脉明显增粗,怒张。病理检查:左侧宫角处肌壁间血管畸形,丰富,符合子宫动静脉畸形诊断。例2:28岁,因“咳嗽伴下腹胀1月,阴道流血25天”入院。患者G2P1A1,剖宫产及人流各1次。盆腔CT示子宫190mm×160mm×80mm,子宫壁均匀变薄,子宫内部蜂窝状密度影,增强后内见不均匀蚯蚓状强化,肿瘤右供血动脉明显增粗迂曲。双肺有散在多个类圆形结节样密度增高影。血HCG:541.62IU/L,孕酮163.27μg/L。初步诊断:绒毛膜癌,多发肺转移瘤。予EMA-CO方案化疗,2天后自行排出2500ml葡萄籽样物。复查B超示宫腔内大量异常回声,予清宫术,术中出血约300ml。复查血常规,HB:63g/L, PLT6×10^9/L。血HCG:884679IU/L。立即转ICU抢救,严密监测病情后好转。病理提示完全性水泡状胎块,滋养叶细胞轻度增生。考虑病情复杂,追加1mm薄层重建,确诊为侵蚀性葡萄胎合并UAVM。1.仪器与方法:同第一章2.图像后处理:同第一章。结果:2例患者经CTA均明确诊断为UAVM。2例患者盆腔动脉三维重建图像均能清晰显示骨盆、髂总动脉、髂内动脉、髂外动脉及其分支,重建图像显示血管连续,血管壁光滑,走行自然,分支清晰。通过任意缩放及360°旋转图像,可以确定畸形血管团的位置、大小、供血动脉和引流静脉以及毗邻的血管、骨结构的立体空间关系。结论:64排螺旋CT结合三维重建技术能提供清晰立体的影像资料,对UAVM的诊断和治疗具有明显的优势。全文小结:1.基于64层CTA数据集,应用CT自带软件,可以构建出一定的女性盆腔动脉血管网模型,为盆腔动脉的形态学研究提供个体化、准确的解剖学依据。2.基于64层CTA构建的盆腔动脉血管网模型,是研究活体女性盆腔动脉的好方法,值得临床推广。3.基于64层CTA数据集,可以构建出理想的产后出血的盆腔动脉血管网,为产后出血患者的诊疗提供了新技术。卵巢动脉参与产后病变的供血,可能是导致常规栓塞两侧髂内动脉-子宫动脉不能控制出血的原因之一。4.64层螺旋CT血管重建可以清楚显示巨大盆腔肿瘤的供血动脉,能提供较准确的诊断依据。5.基于64层CTA数据集,可以构建出个体化的剖宫产术后子宫疤痕妊娠的盆腔动脉血管网,为剖宫产术后子宫疤痕妊娠患者选择合适的治疗方式提供形态学参考。6.64排螺旋CT结合三维重建技术能提供清晰立体的影像资料,对子宫动静脉畸形的诊断和治疗具有明显的优势。
【Abstract】 Research backgroundFemale pelvic arterial system is a large and complex system, and has been the difficulty and keystone in anatomical and gynecologic&obstetric researches due to dense internal iliac arteries, complex origins, tortuosity, polytropy and small arteries. Pelvic arteries are different for different individuals, some individuals even have obvious differences, thus it is difficult to form a unified and fixed standard, as a result, knowledge about the pelvic arterial system needs to be improved to individualized and3D concept.At present, main pelvic artery research methods include color doppler ultrasonography, digital subtraction angiography (DSA), magnetic resonance angiography (MRA) and computed tomography angiography (CTA). Color doppler ultrasonography is non-invasive, economic and convenient, but can not continuously display3D pelvic vessel images, and the imaging quality is greatly influenced by the skills of examiners. DSA has been deemed a "gold standard" for arterial research. However, DSA is an invasive examination and complex in operation, and2D images obtained by DSA can’t be evaluated for overlapping and other reasons, thus significantly restricting its application to morphological study of pelvic arteries, and DSA is only used for subjects to be subject to clinical selective arterial embolization. MRA is radiation-free, and has higher resolution for soft tissues. In recent years, some scholars have performed beneficial research on the construction of pelvic arterial vascular networks based on MRA. In2008, Naguib et al. successfully constructed pelvic arterial vascular networks of subjects to be subject to uterine artery embolization (UAE), and found higher coincidence rate after comparing MRA images with internal iliac arteries closely related to UAE catheterization. In2010, Mori et al. compared the display rate of uterine artery origins between the unenhanced MRA and DSA, found that97%of uterine artery origins were clearly displayed in MRA, and found that ovarian arteries of5subjects were involved in blood supply of hysteromyoma, thus they considered that unenhanced MRA can provide UAE with reliable information on uterine artery and ovarian artery. However, MRA has disadvantages of complex imaging sequence, long examination time, unclear uterine artery and hysteromyoma angiography, and can not analyze blood supply distribution for hysteromyoma in details. Therefore, the pelvic arterial vascular network constructed by MRA can display internal iliac arteries, uterine arteries and other enlarged blood vessels closely related to UAE well, but there is no report on successful construction of smaller artery vascular network. In addition, poor skeleton display of MRA makes vessel recognition difficult.CTA is a brand new angiography developed based on spiral CT in the1990s, and now widely applied to clinical practice due to its advantages of relative non-invasiveness, convenience and quickness, large amounts of information, strong3D effect, high performance cost ratio, etc. Especially after64-slice spiral CT was applied to clinical practice in2004, temporal resolution and special resolution have been obviously improved, achieving real isotropy, enabling CTA to progress greatly, and allowing broader space for diagnosis of diseases. So far, may scholars have performed detailed anatomical researches on arterial vascular networks of multiple body parts such as skull and thyroid gland using this technology. Some studies show that CTA can basically replace the traditional DSA technology in terms of display of cardiac and cerebral arterial vascular networks.The successful application of CTA technology to cardiac and cerebral arterial vascular networks, etc. provides new research methods and development space for morphological study of human pelvic vessels. So far, application of this technology to morphological study of pelvic vessels is rare. In2006, Gao Chengjie et al constructed Level3-4branches of pelvic arterial vascular network models of16normal subjects (male and female) by64-slice spiral CT. However, smaller uterine arterial vascular networks were not displayed due to restrictions of the CT equipment. In2011, Chen Chunlin et al constructed a pelvic arterial vascular network model of one patient with adenomyosis using Mimics software based on64-slice CTA data set, but did not provide detailed anatomical and image data, and did not perform detailed anatomical and morphological studies on the origins, ostia and routes of all pelvic arterial vessels. In the same year, Chen Chunlin et al performed anatomical classification of female internal iliac arteries on CTA of170gynecological insubjects, described main branch types of internal iliac arteries, but did not provide anatomical, variation and image data of other small collateral vessels of pelvic organs. In2011, Bilhim et al performed comparative studies on major branches of internal iliac arteries of21male subjects by different imaging methods (i.e. MRA, CTA and DSA), and considered that CTA can display fine pelvic arterial vascular networks better compared with MRA, and was faster than DSA. The study provided detailed image and anatomical data, especially described small collateral vessels of pelvic arteries such as obturator arteries and iliolumbar arteries in details. Unfortunately, all objects of the study were male. Therefore, detailed CTA image and anatomical data studies on female pelvic arterial vascular networks have not yet been reported, especially studies on large samples.For this reason, we studied anatomical characteristics of the origins, routes, adjacent structures, etc. of adult female pelvic arterial vascular networks, especially small collateral arteries through observation and analysis of CTA images of normal pelvic arterial systems of33adult female using embedded software of CT machines based on the original CTA studies on normal female pelvic vessels for the purpose of recognizing CTA manifestation of pelvic arterial vessels, providing a platform for application of3D reconstruction technology to diagnosis and treatment of female pelvic diseases, and providing detailed anatomical and imaging basis for gynecological interventional radiotherapy and minimally invasive laparoscopic surgery.Chapter1Morphological Study and Clinical Verification of3D Reconstruction of Pelvic Arteries by64-slice Spiral CTPurpose:To construct normal pelvic arterial vascular networks of33adult female using embedded software of CT machines based on CTA data set, and study anatomical and morphological characteristics of the origins, routes, adjacent structures, etc. of all branch vessels of female pelvic arteries, especially small collateral arteries so as to recognize CTA manifestation of pelvic arterial vessels, provide a platform for application of3D reconstruction technology to diagnosis and treatment of female pelvic diseases, and provide detailed anatomical and imaging basis for gynecological interventional radiotherapy and minimally invasive laparoscopic surgery.Method:1. Research object:Collect64-slice spiral CTA data of33adult female with normal pelvic cavity (no history of pelvic disease and operation, and no pelvic lesion based on CTA results) who were subject to middle and lower abdominal CTA examination for various reasons from Department of Gynecology and Obstetrics, Urinary Surgery and Gastrointestinal Surgery of Zhu Jiang Hospital of Southern Medical University and Pan-Yu Central Hospital from October2010to December2012. The subjects were aged from20to61years old, with an average age of(37.36±11.97)years. All subjects were accepted64-slice spiral CT scanning and3D reconstruction of pelvic vessels, and did not undergo any chemoradiotherapy or operation before data acquisition. Governing procedures of the study conform to ethical standards specified by the committee in charge of human trials in our hospital, in addition, the committee approved the procedures and signed information consent form.2. Apparatus: (1) CT scanner:Toshiba Aquilion64-slice spiral CT scanner (the detector assembly is0.5mm×64).(2) Double-syringe high pressure injector:MEDRAD(3) Image post-processing workstation:Start Vitrea2post-processing workstation of Aquilion64-slice spiral CT scanner(4) Ultravist (370mg/ml):manufactured by Schering AG.(5) CT viewing software:Med "Viewer image workstation of Huahai.3. CT scanning parameter setting and scanning method:Subjects were subject to fasting for4-6h before examination to properly fill their bladders (no other special treatment). The subjects were in supine positions, with median sagittal plane perpendicular to bed surface, and lay in the center of the bed surface, with heads in hands and legs stretched out together, and the scanning range was from midpiece of the fourth lumbar vertebra to3cm below greater trochanter of femur. Scanning conditions are as follows:tube voltage:120KV, tube current:250mAs, detector assembly:0.5×64row, slice thickness:0.5mm, inter-slice spacing:0.3mm, alignment pitch:0.984, and tube rotating period per cycle:0.4s. The subjects were subject to conventional scanning, then bolus injection of nonionic iodine contrast medium ultravist370(ultravist contains370mg/ml iodine) through median cubital veins by high pressure injectors at a dose of1.5ml/kg and injection rate of4.5ml/s, and injection of20ml normal saline at the same rate after injection of the contrast medium. Region of interest (ROI) was selected on upper margin slice of abdominal aortic bifurcation by contrast medium tracing method to dynamically monitor CT values and set3s after CT values within the ROI reach200Hu as an acquisition point, with scan delay time during the arterial phase of25s.4. Post-processing:Fault images from scanning was sent to the StartVitrea2post-processing workstation, and the reconstruction methods were as follows:(1) maximum intensity projection (MIP),(2) multi-planar reconstruction (MPR), and (3) volume rendering (VR). Different thresholds and models were adjusted according to the above three reconstruction methods to reconstruct satisfactory CTA images, and images were observed at any angle in a3D space based on systematic anatomical knowledge to analyze the panorama of pelvic vessels, and observe the display situation and morphological differences of pelvic arteries.5. Image analysis items:All arterial identification refer to clinical obstetrics and Gynecology map, Atlas of human anatomy atlas and interventional radiology obstetrics map, etc. The display situation of all branches of pelvic arterial vascular networks was analyzed, including length of bilateral common iliac arteries, ostium position and branch type of bilateral internal iliac arteries, origins, quantity, ostium imaging types and routes of bilateral uterine arteries as well as origins, routes and adjacent situations of bilateral superior gluteal arteries, inferior gluteal arteries, internal pudendal arteries, iliolumbar arteries, obturator arteries, umbilical arteries, lateral sacral arteries, median sacral arteries, ovarian arteries, etc.6. Statistical analysis:SPSS13.0software was used for statistical analysis of data. Enumeration data was expressed in%, and χ2tests were performed on data in2×4table for comparison of two sample rates. The measurement data were expressed by mean±standard deviation (x±s), t tests were performed on independent samples for comparison of sample means, and P<0.05was considered to be statistically significant.Results:1. CTA construction of adult female pelvic arterial vascular networks:Complete pelvic arterial vascular network images of the33subjects were successfully constructed. The arterial vascular networks were characterized by clearness and reality, regular morphology, natural routes and strong3D effect, and can clearly display abdominal aorta, superior rectal arteries, common iliac arteries, median sacral arteries, external iliac arteries, internal iliac arteries, uterine arteries on anterior branches of internal iliac arteries, obturator arteries, umbilical arteries, superior vesical arteries and inferior vesical arteries, iliolumbar arteries on posterior branches of internal iliac arteries, superior gluteal arteries, inferior gluteal arteries, lateral sacral arteries, internal pudendal arteries, femoral arteries and inferior epigastric arteries of their branches, superficial iliac circumflex arteries, deep iliac circumflex arteries, medial femoral circumflex arteries, external pudendal arteries, deep femoral arteries, lateral femoral circumflex arteries and their branches, etc. under physiological conditions. The ostia, origins, routes and adjacent relationship of pelvic arteries and their Level3-4vascular branches can be clearly recognized by arbitrary scaling and360°rotation of CTA images.2. Length changes and types of common iliac arteriesCommon iliac arteries were divided into5types according to their length and relationship:(1) Long common iliac arteries:3subjects, accounting for9.1%.(2) Short common iliac arteries:1subjects, accounting for3.0%.(3) Longer left common iliac artery than right common iliac artery:13subjects, accounting for39.4%.(4) Longer right common iliac artery than left common iliac artery:1subjects, accounting for3.0%.(5) Equal common iliac arteries:15subjects, accounting for45.5%.3. Anatomy of internal iliac arteries(1) Ostium position and height measurement of internal iliac arteries:The ostium of internal iliac artery is located between1/2above the fifth lumbar vertebra and1/2below the first sacral vertebra. Comparison of the ostium heights of bilateral internal iliac arteries showed that the subjects with ostia of the right internal iliac arteries higher than those of the left internal iliac arteries accounted for42.4%(14/33), the subjects with ostia of the left internal iliac arteries higher than those of the right internal iliac arteries accounted for6.1%(2/33), and the subjects with equal height of ostia of bilateral internal iliac arteries accounted for51.5%(17/33).(2) Branch type of internal iliac artery:By reference to Adachi typing method and combination of three branches of the internal iliac arteries (superior gluteal artery, inferior gluteal artery and internal pudendal artery), internal iliac arteries were divided into five types. For Type I, the internal iliac arteries originated the superior gluteal artery and the inferior pudendal trunk successively; for Type Ⅱ, the internal iliac arteries originated the superior gluteal artery, the inferior gluteal artery and the internal pudendal artery successively; for Type Ⅲ, the internal iliac arteries originated the gluteal trunk and the internal pudendal artery successively; for Type Ⅳ, the internal iliac arteries originated a main trunk which further originated the superior gluteal artery, the inferior gluteal artery and the internal pudendal artery; and for Type V, two inferior gluteal arteries had different origins, with one inferior gluteal artery originating from the superior gluteal artery and the other one having sharing trunks with the internal pudendal artery. In this study, only Type Ⅰ, Type Ⅱ and Type Ⅲ were seen, and Type IV and Type V were not seen. The percentage of the first three types in left internal iliac arteries accounted for63.6%(21/33)、30.3%(10/33)及6.1%(2/33) respectively, while the percentage of the three subtypes in right internal iliac arteries accounted for69.7%(23/33)、21.2%(7/33)及9.1%(3/33) respectively. Although63.6%(21/33) of subjects had the same types in bilateral internal iliac arteries, this difference was not statistically significant when it was compared with the percentage of various types to bilateral internal iliac arteries (χ2=6.398, P=0.171) (3) CTA manifestation of posterior trunk of internal iliac artery:CTA images of22subjects (66.7%) can display iliolumbar artery which is the branch at the highest position in internal iliac arteries and most commonly seen in the main trunk and posterior trunk of internal iliac arteries. CTA images of24subjects (72.7%) can display lateral sacral artery which most commonly originated from trunk terminal of internal iliac artery, followed by superior gluteal artery and inferior gluteal artery. Among66branches of obturator arteries of33subjects,55branches (83.3%) can be clearly displayed. The origins of obturator arteries varied greatly, and most of the obturator arteries originated from internal iliac arteries or their branches, superior gluteal artery, trunk of inferior gluteal pudendal artery, main trunk of internal iliac artery, inferior gluteal artery, etc. As the terminal branch of the posterior trunk of the internal iliac artery and the biggest branch of the internal iliac artery,66branches of superior gluteal arteries (100%) were totally displayed. It can be seen from CTA images that most of superior gluteal arteries originated from the internal iliac artery, and formed gluteal trunk with inferior gluteal arteries for only7.5%of subjects.(4) CTA manifestation of anterior trunk of internal iliac artery:The umbilical artery is extended from main trunk of the internal iliac artery, originates superior vesical artery at proximal end, and bends inward into2-3branches to reach over superior ramus of pubis and bend inward. The66internal pudendal arteries can be clearly displayed. As a smaller branch of two terminal branches of anterior trunk of the internal iliac artery, the internal pudendal artery had co-trunk with inferior gluteal artery or originated from single trunk, but very few internal pudendal arteries have co-trunks with superior gluteal arteries. As the biggest terminal branch of anterior trunk of the internal iliac artery, inferior gluteal arteries had co-trunk with the internal pudendal artery originated from anterior trunk of the internal iliac artery in most subjects, or single branch originated from the internal iliac artery in a few subjects.4. Anatomy of uterine artery (1)Origin of uterine artery:The images can show ostia, routes, size and main branches of66uterine arteries (100%). The uterine arteries had varied origins, in which31branches originated from main trunk of internal iliac artery (15branches at left and16branches at right), accounting for47.0%,20branches originated from pudendal trunk of inferior gluteal artery (11branches at left and9branches at right), accounting for30.3%,11branches originated from umbilical artery (6branches at left and5branches at right), accounting for16.7%,1branches originated from internal pudendal artery (1branches at right), accounting for1.5%, and other3branches originated from superior gluteal artery or inferior gluteal artery (1branches at left and2branches at right), accounting for4.5%. In this study,51.5%of subjects (17/33) had asymmetric origins of left and right uterine arteries.(2) Ostium types of uterine arteries:Among66branches:①51branches appeared to be acute angle (25branches at left and26branches at right), accounting for77.4%,②3branches appeared to be right angle (3branches at left and0branches at right), accounting for4.5%,③9branches appeared to be rotary (4branches at left and5branches at right), accounting for13.6%, and④3branches appeared to be spiral (1branches at left and2branches at right), accounting for4.5%.5.ovarian artery:The33subjects had66ovarian arteries, and4subjects showed ovarian arteries, accounting for12.1%, with3subjects having left developing and1subjects having right developing. The developing rate of ovarian arteries was6.1%(4/66).Conclusions:1.Data of64-slice spiral CT and the embedded software of CT machine can help construct ideal female pelvic arterial vascular network models, so64-slice spiral CT is a good method for researching living female pelvic artery and worthwhile for promotion in clinical practice.2.The pelvic arterial vascular network models constructed based on64-slice CTA can provide detailed imaging anatomy basis for morphological study of pelvic arteries, provide individualized and accurate morphological basis for minimally invasive pelvic surgery and vascular interventional treatment of pelvic cavity, provide new diagnosis and treatment technique of pelvic vascular diseases for clinicians and lay a foundation for practice of individualized treatment.Chapter2Clinical Application of3D Reconstruction of Pelvic arterial vascular Network by64-slice Spiral CT Section1Application and Significance of3D Reconstruction of Pelvic arterial vascular Network by64-slice Spiral CT to Postpartum HemorrhagePurpose:Based on64-slice data set of21subjects with postpartum hemorrhage suffering from invalid conservative treatment, embedded software of CT machine was used to reconstruct pelvic arterial vascular networks to study construction method and significance of pelvic arterial vascular networks for subjects with postpartum hemorrhage, and observe angiographic manifestations of ovarian arteries as sources of postpartum hemorrhage.Method:1. General data:from21subjects (including3subjects with concomitant placenta implantation) who were hospitalized for delivery in Department of Obstetrics of Pan-Yu Central Hospital from January2012to December2012, and underwent pelvic64-slice CTA for invalid conservative treatment upon postpartum hemorrhage for various reasons (drug, uterus massage, curettage of uterine cavity, etc.). All subjects had over500mL bleeding within24h upon cesarean section or delivery. After conventional conservative treatment, colporrhagia did not decrease or slightly decreased and increased again. They aged from21to34years, with average age of27.24±3.45years. Among the subjects,12subjects were primipara,9subjects were multipara;7subjects underwent vaginal delivery and14subjects underwent cesarean section;13subjects suffered from uterine inertia,6subjects suffered from placental factors,1patient suffered from laceration of soft birth canal and1patient suffered from cesarean scar hematoma. Among21subjects,3subjects suffered from concomitant placenta implantation, with amount of bleeding up to565ml-330ml and average amount of bleeding up to813.10±182.08ml.2. Apparatus and method:special treatment was not required before scanning. Others are the same as those described in Chapter1.3. Image post-processing:the same as those described in Chapter1.4. Statistical method:the same as those described in Chapter1.Results:For21subjects, complete pelvic arterial vascular network models were constructed successfully. The vascular network had clear edge, regular morphology, natural route and strong3D effect, can be arbitrarily scaled and rotated by360°for observation, thus clearly displaying internal iliac arteries and Level3to4branch vessels, and displaying3D images of all pelvic vessels including ovarian arteries. Among21subjects, CTA images of12subjects can show ovarian arteries, accounting for57.1%, with6subjects having bilateral display and6subjects having unilateral display at left or right. The display rate of ovarian arteries was42.9%(18/42). Among12subjects with ovarian arteries showed, ovarian arteries were enlarged, with diameter up to2.0-11.6mm,2subjects having abnormal ovarian arteries distension, with diameter up to11.6mm, early developing of ovarian vein was seen, indicating possibility of ovarian arteriovenous malformation.Conclusions:1. CTA data set and the embedded software of CT machine can help construct ideal pelvic arterial vascular networks of subjects with postpartum hemorrhage, thus providing new diagnosis and treatment technique for clinicians and anatomical basis for pre-interventional operation evaluation of subjects with postpartum hemorrhage.2. ovarian artery can participate in blood supply of postpartum pelvic lesion, which is probably one of the reasons for uncontrollable hemorrhage of bilateral internal iliac arteries-uterine arteries in conventional embolism.Section2Application of Vascular Reconstruction by3D Spiral CT to Feeding Arteries of Giant Pelvic TumorsPurpose:At present, there are many reports on CT diagnosis of pelvic tumors at home and abroad, but research on application of3D vascular reconstruction by64-slice spiral CT to giant pelvic tumors is rarely reported at home and abroad. For this article, vascular network was reconstructed for patients with giant pelvic tumors confirmed by surgical pathology using64-slice spiral CT3D reconstruction technique to observe display conditions of tumor vessels and feeding arteries and evaluate the value of the technique in diagnosis of giant pelvic tumors, thus helping formulate clinical therapy.Method:1. Collection of general data:clinical data of11female patients with giant pelvic tumors who were aged from27to54years with an average age of41.73±8.58years, subject to64-slice spiral CTA examination from October2010to October2012and confirmed by surgical pathology. The size of tumors was8.5cm-27.0cm, with an average size of (13.64±6.07) cm. Four patients had hysteromyoma. Seven patients had ovarian tumors, including one patient with serous cystadenoma, one patient with borderline mucinous cystadenoma, two subjects with serous cystadenocarcinoma and two subjects with mucinous cystadenocarcinoma. One patient had metastatic oophoroma (primary gastric cancer). Clinical symptoms: abdominal mass, abdominal distention, abdominal pain, sense of bearing down, menoxenia, irregular colporrhagia, etc. All subjects had signed CT examination consent forms.2. Apparatus and method:the same as those described in Chapter1.3. Image post-processing:the same as those described in Chapter1.4. Judgment criterion of feeding artery:the enlarged artery with branches extending into tumors and distributed as net shape or radial pattern is called as feeding artery. Tumors with blood supplied by uterine branch (UB) were supposed to originate from uterus. Tumors with blood mainly supplied by ovarian artery (OA) and/or ovarian branch (OB) were supposed to originate from ovary.5. Statistical method:the same as those described in Chapter1.Results:1.Complete pelvic arterial vessel images of11patients were successfully constructed, with clear vessel edges, regular morphology, strong3D effect and natural route. The images can be arbitrarily scaled and rotated by360°to observe the pelvic vessels and feeding arteries. The vessel images of11subjects can show abdominal aorta, common iliac artery, external iliac artery, internal iliac artery, uterine artery, ovarian branch of uterine artery or ovarian artery clearly, and also displayed feeding arteries of11subjects.2.Giant hysteromyoma:4patients. The maximum diameter of the tumors was8.5-27.0cm, and average diameter was14.63±8.38cm. Vascular post-processing reconstruction showed that unilateral and bilateral UB blood supply was present in the tumors, with the incidence of100%(4/4). Blood supply of one patient was jointly provided by bilateral UB and right OA, with the incidence of25.0%(1/4), representing that bilateral UB and OA peripheral branch in the tumor body. The accuracy rate of anatomical diagnosis of tumors was100%(4/4).3. Giant ovarian tumor:7patients including6patients with primary ovarian tumor and1patient with metastatic ovarian tumor. The maximum diameter of the tumors was9.0-22.0cm, and the average diameter was13.07±5.04cm. Feeding arteries were displayed in the vessel images of7subjects,3patients had bilateral feeding arterials and4patients had unilateral feeding arterials,10sides in total. Vascular post-processing reconstruction showed bilateral blood supply from UB, OB and OA, trilaterial blood supply from OA and OB, unilateral blood supply from OA and quadrilateral blood supply from OB. The blood supply incidence of OA was50.0%(5/10), and that of OB was90.0%(9/10). The accuracy rate of anatomical diagnosis of tumors was100%(10/10).Conclusions:As a noninvasive imaging examination method,3D-CTA can clearly display the feeding arteries of giant pelvic tumors and can provide accurate diagnostic basis owing to obvious advantages in diagnosis of giant pelvic tumor.Section3Construction and Significance of Pelvic Arterial Vascular Network with Cesarean Scar PregnancyPurpose:Uterine artery embolization (UAE) has become a safe and effectively method for treating cesarean scar pregnancy (CSP). It can be used as the first treatment choice for young CPS subjects requiring preserving fertility and having extraordinarily abundant local blood flow at scars prompted by ultrasound. Many researches have demonstrated that3D reconstruction of pelvic arterial vascular network using CTA before UAE and the implementation of UAE directly instructed by the network accordingly can improve success ratio of surgery, and also can obviously reduce exposure time of X-ray dose and the dosage of contrast medium, thus playing a very important instructive role in treatment. At present, the research on application of the technology to CSP has not been reported. For this reason, we successfully built pelvic arterial vascular network models of5patients with CSP using CTA3D reconstruction technology, and discussed the value of applying the vascular network to CSP. Method:1. Selection of general data CTA data set of5patients with CSP who were hospitalized in Obstetric and Gynecologic Department of Pan-Yu Central Hospital from October2010to October2012and subject to CTA examination of pelvic cavity. Five patients were aged from28to37years, with an average age of33.40±3.58years. Gravidity was3-6, with an average of4.20±1.30. Parity was1-3, with an average of1.80±0.84. Five subjects had history of menostasia, with the shortest menolipsis being34days, longest menolipsis being112days and an average of68.6±31.9days. The interval time of the pregnancy and the previous caesarean section was6months at least and11years at most, with an average of4.94±4.32years.2. Diagnostic criteria:①history of cesarean section and menostasia;②colporrhagia:natural profuse bleeding or massive bleeding, massive and endless bleeding in abortion operation without embryonic tissue discharged in drug abortion, and massive and endless bleeding in clearing uterus;③stomachache:no stomachache or slight stomachache;④enhanced serum β-hCG;④color doppler ultrasonography:gestational sac was not seen in uterine cavity and cervical canals, gestational sac was located at isthmus of uterus, and muscular layer of isthmus became thin; or myometrium lacked between gestational sac and bladder or muscular layer had defect; color Doppler prompted abundant and high-rate blood flow around foetal sac or mass.3. Course of treatment:one patient with concomitant incomplete rupture of uterus was subject to embryo remove surgery at cesarean scar in emergency treatment, one patient was subject to embryo remove surgery at cesarean scar under a laparoscope, and two patients were subject to MTX treatment. One patient required to return back to local hospital for treatment due to financial difficulty.4. Apparatus and method:the same as those described in Chapter1.5. Image post-processing:the same as those described in Chapter1.6. Statistical method:the same as those described in Chapter1. Results:Complete pelvic arterial vascular network images of5patients with CSP were successfully constructed, with clear vessel edge, regular morphology, strong3D effect and natural routes. Bone structure of pelvis, cacroiliac joint, sacrococcygeal joint, pubic symphysis, etc. can be clearly showed; and common iliac artery, external iliac artery, internal iliac artery, uterine artery and other branch vessels in physiological condition can be reproduced truly. The vessel distribution and richness of blood supply of scar pregnancy lesions can be clearly displayed by arbitrarily scaling and rotating the3D structure of vessel networks by360°.Conclusions:Ideal pelvic arterial vascular network models of patients with CSP can be constructed with the embedded software of CT machine based on CTA data set, thus providing accurate morphological basis for selecting proper therapy and pre-interventional operation evaluation for subjects with CSP.Section4Application of Vascular Reconstruction by64-slice Spiral CT to Uterine Arteriovenous MalformationPurpose:At present,3D reconstruction based on
【Key words】 pelvic cavity; artery; angiography; 3D reconstruction; computerized tomography; X-ray;