建立基于MRI图像的胎盘血管网数字化三维模型技术的研究及其在胎儿生长受限胎盘血管网结构研究中的应用

Vascular abnormality of placenta acts as the pathogenesises of many pregnancy complications , however ,there doesn't exist a suitable method used to study placenta vascular network.In this research, elites from obstetrics, imaging department and anatomy department jointly explore a new method- - digital three-dimensional reconstruction technique based on MRI images- - to study placenta vascular network.We reconstruct the digital three-dimensional models of placentas based on MRI images and use these models to explore the distinctions of placenta vascular network between the placentas before delivery and those after delivery in the following aspects：structure, branch, vessel grading, vessel capacity, vessel diameter and so on. We also compare the fetal growth restriction（FGR） placentas and normal placentas to find out their differences in placenta vascular network. Methods: Abdominal MRI scanning for pregnant women before delivery-digital reconstruction of placenta before delivery with the Mimics 14.0 software- - MRI scanning for placenta after delivery-digital reconstruction of placenta-placenta vascular network perfusion-enhanced MRI scanning- - digital reconstruction of placenta-establish the placenta vascular network casting model. Using the placenta vascular network casting model as the standard, we determine the accuracy of digital three-dimensional models of placentas based on MRI scanning and enhanced scanning, and continuously modify the parameter of scanning and reconstruction, finally to obtain the best solution of digital reconstruction of placenta vascular network. The results of our research can not only establish a new method to study placentas vascular network, but also play important roles in substantiating the placenta histology theory and the pathogenesis of FGR.

很多妊娠合并症或并发症的发病与胎盘血管异常有关，但胎盘血管研究却缺乏良好手段。我们集合产科、影像学科和解剖学科精英从数字医学角度入手共同研究一种用于胎盘血管网研究的新方法- - 基于MRI数据的胎盘血管网数字化三维模型，并利用该模型研究产前与产后胎盘血管网在结构、分支、血管分级、血管容量、血管管径等方面是否存在差异，研究胎儿生长受限胎盘与正常胎盘血管网的不同。方法为：分娩前孕妇MRI盆腔扫描- - 利用Mimics软件数字化重建分娩前胎盘- - 分娩后胎盘MRI扫描- - 胎盘数字化重建- - 胎盘血管灌注- - 增强MRI扫描- - 胎盘数字化重建- - 建立胎盘血管铸型模型。以胎盘血管铸型为实体对照，确定基于MRI平扫和增强扫描的胎盘数字化模型的准确性，不断修正扫描条件和重建条件，最终得到胎盘血管网数字化重建的最佳方案。本研究结果不仅可以建立一种胎盘血管研究的新方法，还对充实胎盘组织学理论和FGR发病机理有重要意义。

为了解决胎盘血管研究缺乏良好手段问题，本研究着手建立基于MRI扫描数据的胎盘血管网数字化三维模型（3D-PVNM），利用该模型研究正常胎盘娩出前后血管网变化和FGR胎盘与正常胎盘血管网区别。在3D-PVNM研究中，我们先利用血管显示较为清晰的增强CT扫描，建立出效果满意的离体胎盘3D-PVNM。以此模型为参照，进一步探索利用MRI数据建立离体和在体胎盘3D-PVNM条件，并完成MRI数据下离体胎盘3D-PVNM、在体胎盘-子宫-骨盆拟合模型、正常孕鼠和FGR孕鼠在体和离体胎盘3D-PVNM。比较孕鼠分娩前后胎盘血管变化、FGR孕鼠与正常孕鼠胎盘血管网差异。结果显示，以自凝牙托材料为血管填充剂、先脐动脉后脐静脉为灌注顺序、0.75mm为扫描层厚和层间距获得的增强CT扫描数据，可在Mimics软件上构建出立体感强、血管光滑饱满、分支清晰、可任意缩放和旋转的离体胎盘3D-PVNM，与同胎盘血管铸型相似度达100%。以无造影剂材料填充血管、钆喷酸葡胺混合液浸泡胎盘、e-THRIVE为扫描序列、脐动静脉分次灌注扫描获得MRI数据，用Mimics软件以“点配准”法融合，可获得满意的MRI数据3D-PVNM。以BaSO4为造影剂、Micro-CT常规参数扫描孕鼠胎盘，利用胎盘血管重建技术可获得正常和FGR孕鼠在体和离体胎盘3D-PVNM。比较孕鼠分娩前后胎盘血管分级、密度、容量、管径无明显变化；FGR孕鼠孕20天胎盘血管指标较孕14天略有减少，但无统计学意义（P＞0.05）；FGR孕鼠离体胎盘血管各指标比正常孕鼠明显减少（P＜0.01）。以T2WISSFSE序列对120例孕妇行MRI扫描，重建胎盘-子宫-骨盆数字模型，在模型上行子宫虚化、前置胎盘定位分型、胎盘边缘线和子宫切口模拟线勾画，形成指导手术用“子宫数字化导航模型”。以该模型指导42例“凶险型前置胎盘”手术，出血量、输血量、手术时间、ICU转入率、出院时间、住院费均明显少于常规手术，子宫保留率100%。本研究各项结果均为国内外首次完成。3D-PVNM血管测量的准确性弥补了既往血管研究方法的不足，并可作为良好教学工具。胎盘娩出前后血管变化情况的明确和FGR胎盘血管减少的验证，为胎盘组织学研究及FGR发病机理研究提供理论依据。“数字化导航”下的前置胎盘手术，效果明显优于传统手术，是解决“凶险型前置胎盘”手术难题的有效方法。