基于血液双对比度的空间信息完全匹配的亮血黑血磁共振成像技术在颈动脉粥样硬化成像中的研究

Atherosclerotic disease is the leading cause of death and disability worldwide. Outside of the coronary circulation, the carotid arteries are likely the most clinically significant site of atherosclerosis. To diagnose carotid stenosis using magnetic resonance imaging (MRI), bright-blood and black-blood (also called vessel wall imaging) magnetic resonance angiography (MRA) are usually jointly performed to depict vessel morphology, detect carotid artery thrombus and evaluate plaque size and its constituents such as fibrous cap, lipid core so on. . However, as independent MRI techniques, bright-blood MRI and black-blood imaging are always implemented separately. Thus, spatial misregistration between bright- and black-blood scans has become a limitation of interpreting images in multiple blood contrast imaging protocol for clinical MRI examination. The reason to cause the misregistration is the deteriorated image quality of bright-blood or black-blood images resulting from imaging subject’s motion such as swallowing, respiration and neck movements during the long scan time of the two scans. . Based on our previous study on compressed sensing (CS) based simultaneous black- and gray-blood carotid vessel wall MR imaging, we create a hypothesis that the perfect spatial registered bright- and black-blood images can be achieved in one single scan by means of post-processing the black- and gray-blood images obtained by the simultaneous black- and gray-blood MRI sequence. The technique to generate such bright- and black-blood images is termed compressed sensing based simultaneous bright- and black-blood carotid artery MRI (CS-siBrBlack). In this study, there are three aims: (1) to demonstrate the feasibility of CS-siBrBlack by combining CS based three-dimensional (3D) MRA sequence with the image segmentation and fusion of carotid lumen and wall for a large number of clinical cases; (2) to evaluate the quality of CS-siBrBlack bright-blood images obtained by post-processing black- and gray-blood images; (3) to explore the characteristics and advantages of the perfect spatial registered bright- and black-blood images for the clinical diagnosis and the clinical application prospect. The importance of the technique for the clinical diagnosis of carotid atherosclerosis will then be elaborated according to the statistical results on the large number of clinical cases. CS-siBrBlack will be a promising tool for the clinical diagnosis of vascular disease and will provide a new perspective for the diagnosis.

磁共振颈动脉粥样硬化的临床诊断需结合亮血和黑血图像（血管壁图像）来进行。然而亮血和黑血是两个独立的成像技术，需分两次扫描，因此如何得到空间位置完全匹配的亮血和黑血图像是磁共振颈动脉成像上一个急需解决的难题。我们基于前期在黑血灰血双对比度成像序列方面的研究，提出假设：通过将黑血灰血双对比度成像序列与图像处理相结合的方法，就能够在一次扫描中得到空间位置完全匹配的亮血和黑血成像。本课题拟从基于压缩感知的磁共振三维血管成像序列、颈动脉管腔和管壁的图像分割和融合以及大量临床病例三个部分的研究来验证亮血黑血同时成像技术在临床应用的可行性；确定由黑血灰血成像序列与图像处理相结合所得的亮血成像质量；探讨空间位置完全匹配的亮血黑血图像对于临床诊断的特点和优势以及临床应用前景。本研究拟通过大量病例研究来阐明空间位置完全匹配的亮血黑血图像在颈动脉粥样硬化诊断中的重要意义，从而为血管疾病诊断提供新思路。

磁共振颈动脉粥样硬化的临床诊断需结合亮血和黑血图像(血管壁图像)来进行。然而亮血和黑血是两个独立的成像技术，需分两次扫描，因此如何得到空间位置完全匹配的亮血和黑血图像是磁共振颈动脉成像上一个急需解决的难题。基于前期在黑血灰血双对比度成像序列方面的研究，我们提出了CS-siBrBlack序列方法在一次扫描中得到空间位置完全匹配的亮血和黑血成像。CS-siBrBlack技术利用压缩感知磁共振快速成像序列来得到黑血和灰血图像，然后利用血管分割方法得到血管壁和管腔，再通过图像融合而得到最终与黑血图像空间位置完全一致的亮血图像。为了能够较好地分割血管壁，在实验中我们首先优化黑血序列RECS-3D MERGE使血管壁图像信号得到显著提高。在图像对比以及统计学分析中我们发现，在血液抑制效果方面，基于压缩感知的快速血管壁成像技术（RECS-3D MERGE）要显著优于传统血管壁成像技术（3D MERGE），在血管壁与血管腔对比度方面，RECS-3D MERGE也显著优于3D MERGE。然后在此基础上，我们加入灰血成像部分到序列中，再通过图像分割算法，从而得到黑血亮血图像。并且，分割融合图像通过设置血管腔系数（LF）和血管壁系数（WF）还可以来调整血管壁和血管腔的对比度以及与背景组织的对比度。实验发现通过设置相应的LF和WF，融合图像的血管壁显示明显比黑血、灰血以及3D TOF亮血图像要清晰可辩；在LF=5，WF=1的融合图像中，血管腔信号区域边界清晰，对比度显著高于亮血图像，且血管壁显示比亮血图像完整。对一位颈动脉管壁增厚患者的成像对比显示，LF=6，WF=3分割融合图像的血管壁显示清晰，血管增厚部分显示清晰，同时血管腔也清晰显示，且信号均匀。设置LF=1，WF=6的分割融合图像血管壁显示清晰，血管壁血管腔对比度清晰。血管腔与管壁分界显示清晰，血管壁与周边组织边界清晰。总结，CS-siBrBlack技术能够得到与黑血图像空间位置完全一致地亮血图像，并且通过设置LF和WF，能够得到比现有3D TOF亮血成像质量更好的亮血成像，是一个能够为临床提供高质量血管图像的血管成像技术，有较大的临床应用潜力。