冠状动脉血管的二维提取和运动跟踪

提出了一种基于图像脊提取和snake模型的复合式方法来实现X射线造影图像序列中冠状动脉血管的二维提取和运动跟踪,并分别对临床采集图像序列和模拟图像进行了实验.结果说明,与经典模型相比本算法自动化程度和精度都提高许多.     

基于活动轮廓模型的序列超声图像的心内壁运动跟踪方法

提出一种对心脏序列超声图像中的心内壁进行运动跟踪的方法,采用活动轮廓模型将前一帧图像中snake的停留位置作为当前帧snake的初始位置,选择适当的能量函数,使能量函数最小snake变形得到当前时刻的心内壁轮廓,实验结果论证了该算法的可行性.     

变形模型在冠状动脉二维运动跟踪中的应用

本文提出一种采用变形模型对X射线造影图像序列中的冠状动脉血管进行运动跟踪的方法.首先采用多尺度高斯滤波对造影图像序列进行预处理,将血管区域从背景中提取出来,灰度图像被转化成二值图像.然后,采用改进的离散活动轮廓模型对图像序列进行感兴趣血管段的运动跟踪,其中模型的初始位置是自动获取的.通过在外部能量中引入对端点的灰度相似度和对中间点的运动相似度的测量,有效地避免了模型沿血管收缩或漂移的现象.同时为了保证模型的连续性和光滑性,降低变形结果对权重参数的敏感性,在每次变形结束后对模型进行重新采样.对临床得到的造影序列进行了实验,结果证明了方法的可行性.     

血管造影图像序列中冠状动脉的三维运动估计

提出了由两个角度的单面血管造影图像序列估计冠状动脉骨架树三维运动的算法。首先对冠状动脉造影图像序列进行二维预处理和二维运动估计。然后根据冠脉造影系统的透视投影模型得到两幅不同角度的造影图像之间的几何变换关系，以及空间点三维坐标的计算方法。最后，在对整个图像序列进行分析的过程中，将三维运动估计与重建结合起来，得到各骨架点的三维运动向量。采用临床得到的冠状动脉造影图像序列对算法进行了验证，并分析了误差源。     

冠状动脉的三维运动特征提取

提出一种根据X射线造影图像序列提取冠状动脉血管三维运动特征的方法.首先对由两个角度的造影图像重建得到的三维血管骨架进行运动估计,计算出两个时刻间骨架上各点的三维运动向量.然后结合心脏解剖和运动的先验知识,对血管运动进行定性分析,如位移方向、振幅及运动模式.提取、解释运动信息,并采用符号表达,方便医生进行观察和分析.最后给出了对临床得到的造影图像序列进行实验的结果.     

基于投影图像和CT容积图像的三维冠状动脉运动跟踪新方法

目的利用双平面X射线投影图像序列和参考CT容积图像进行冠状动脉的三维运动跟踪建模。方法①提取投影图像序列中的冠状动脉树;②利用多尺度滤波和血管函数提取CT容积图像中的动脉血管;③采用基于B样条配准的方法进行三维运动建模。结果将双投影图像中血管的运动估计结果的三维重建形态与三维运动模型预测出的结果进行了量化比较,调整配准的B样条参数后,两者误差处于有效范围之内。结论由此表明采用投影图像和对应容积图像的联合先验知识进行冠状动脉的三维运动建模是有效的。     

冠状动脉的二维运动分析

本文提出了两种由造影图像序列估计冠状动脉运动参数的算法:光流法和动态规划法.采用临床得到的单面冠状动脉造影图像序列对两种算法进行了验证,通过比较其估计结果,分析了它们的优缺点及适用范围.     

基于X射线造影图像的冠状动脉运动分析

我们提出了一种采用X射线造影图像对冠状动脉的非刚性运动进行分析的方法,通过寻找两帧图像之间血管上点的对应关系,得到各点的运动向量。匹配中同时考虑了运动场的光滑性和血管的形状相似度。同时由于变形前后血管的长度可能不同,因而加入了对无匹配部分的处理,保证了匹配的均匀性。采用动态规划完成对最优匹配的搜索,减少了计算量。采用仿真图像和临床得到的造影图像分别对该算法进行了验证,对仿真图像运动向量估计的均方根误差在1个像素左右。对临床造影图像的实验也得到了较为满意的结果。     

ICUS图像序列中冠状动脉的三维重建和形态测量

针对冠状动脉内超声(ICUS)图像序列中血管的三维重建问题,提出一种从连续回撒超声导管获取的、覆盖多个心动周期的ICUS序列中准确重建冠脉血管、并定量测量其形态结构参数的方法.首先利用在导管回撤路径起点采集的两幅近似垂直的X射线冠脉造影(CAG)图像之间的交叉信息,重建出导管的回撤路径,然后从各帧ICUS图像中提取出血管壁的内外轮廓;在对ICUS序列中由心脏运动所致的运动伪影进行补偿之后,选择出在各心动周期的同一时相处采集的ICUS帧,并将其沿三维导管路径顺序排列;最后采用NURBS曲面拟合技术,完成血管的三维重建.根据该三维血管模型,对临床重要的血管形态参数进行测量,采用临床图像数据进行验证.实验结果证明,较之仅采用CAG图像或ICUS序列重建出的血管模型的测量结果,该方法更为精确.     

一种基于超声造影剂的血管灌注区域的提取方法

子宫肌瘤超声造影图像中的血管灌注分布提供了有用的病理和生理信息,提取出血管灌注区域有助于量化评价子宫肌瘤的血供情况。血管灌注区域像素灰度变化有别于其它区域,本文因此提出一种提取肌瘤中血管灌注区域的方法。首先对图像序列进行去噪处理,接着采用Brox光流法对相邻两帧图像进行运动估计,根据得出的位移场作warp运动校正,最后根据超声造影图像序列中富血供区域与乏血供区域图像信号的差异,从周围背景中提取出血管灌注区域。实验结果表明,本文算法能精确地提取出血管灌注区域,可达到临床对灌注区域识别的精度,且计算量小,实现简单。     

Comparison of active contour models for image segmentation in X-ray coronary angiogram images

Snakes, or active contours, are used extensively in computer vision and image processing applications, particularly to locate object boundaries. However, most present snake models cannot provide better capture range and evolution stop mechanism. This project presents a new external force for active contours, largely solving both problems. An extension of the gradient vector flow snake (GVF snake) method is presented. First, the adaptive balloon force has been developed to increase the GVF snake's capture range and convergence speed. Then, a dynamic GVF force is introduced to provide an efficient evolution-stop mechanism. In this way, we prevent the snake from breaking through the correct surface and locking to other salient feature points. The active contour models have been applied on X-ray coronary angiogram images. The segmentation results demonstrate the potential of improved GVF method is comparison with all previous active contour methods. Texture parameters have been calculated and results are compared with all active contour models.     

Comparison of active contour models for image segmentation in X-ray coronary angiogram images

Snakes, or active contours, are used extensively in computer vision and image processing applications, particularly to locate object boundaries. However, most present snake models cannot provide better capture range and evolution stop mechanism. This project presents a new external force for active contours, largely solving both problems. An extension of the gradient vector flow snake (GVF snake) method is presented. First, the adaptive balloon force has been developed to increase the GVF snake's capture range and convergence speed. Then, a dynamic GVF force is introduced to provide an efficient evolution-stop mechanism. In this way, we prevent the snake from breaking through the correct surface and locking to other salient feature points. The active contour models have been applied on X-ray coronary angiogram images. The segmentation results demonstrate the potential of improved GVF method is comparison with all previous active contour methods. Texture parameters have been calculated and results are compared with all active contour models.     

一种分割冠状动脉X射线造影图像的有效方法

目的 提出一种简单有效的方法进行冠状动脉X射线造影图像的分割。方法 基于Hessian矩阵的多尺度滤波和区域增长等算法,其中多尺度滤波用来增强造影图像中的血管,然后利用多种子点区域增长算法从增强后的图像中提取冠状动脉树。结果 该方法对于造影图像中血管状结构非常敏感,能够清晰提取出冠状动脉树中较细小的末梢,并能有效抑制噪声。结论 该方法适合于分割冠状动脉造影,适用于冠状动脉造影的精确量化分析。     

Endothelial dysfunction and increased carotid intima-media thickness in the patients with slow coronary flow

Flow mediated brachial dilatation (FMD) and carotid intima-media thickness (IMT) have been a surrogate for early atherosclerosis. Slow coronary flow in a normal coronary angiogram is not a rare condition, but its pathogenesis remains unclear. A total of 85 patients with angina were evaluated of their brachial artery FMD, carotid IMT and conventional coronary angiography. Coronary flow was quantified using the corrected thrombosis in myocardial infarction (TIMI) frame count method. Group I was a control with normal coronary angiography (n = 41, 56.1 ± 8.0 yr) and group II was no significant coronary stenosis with slow flow (n = 44, 56.3 ± 10.0 yr). Diabetes was rare but dyslipidemia and family history were frequent in group II. Heart rate was higher in group II than in group I. White blood cells, especially monocytes and homocysteine were higher in group II. The FMD was significantly lower in group II than in group I. Elevated heart rate, dyslipidemia and low FMD were independently related with slow coronary flow in regression analysis. Therefore, endothelial dysfunction may be an earlier vascular phenomenon than increased carotid IMT in the patients with slow coronary flow.     

A case of acute myocardial infarction after blunt chest trauma in a young man

Coronary artery injury rarely occurs after blunt chest trauma, but it can lead to extensive myocardial infarction and be frequently overlooked. A 16-yr-old man was presented with comatose mental state and rapid respiration rate. He ran into guard rail while riding a motorcycle. In routine examination, his electrocardiogram showed Q wave and 2 mm ST segment elevation in all precordial leads, I and aVL. The cardiac enzymes were also elevated: creatine kinase (CK)-MB was 300 U/L, and cardiac specific troponin I was 5.7 ng/mL. Two-dimensional echocardiography showed anteroseptal akinesia with severely depressed left ventricular function, ejection fraction of 28%. He could not receive any anticoagulation or thrombolytic therapy because of his brain lesion. Three weeks later, his mental state improved. A diagnostic coronary angiogram revealed total occlusion in the proximal left anterior descending artery (LAD) with collaterals from the right coronary artery and left circumflex artery. We successfully performed a percutaneous coronary intervention for the LAD lesion, and the final angiogram showed a good coronary flow without residual stenosis.     

心血管造影图像中的心血管提取

利用心血管造影图像质量描述心血管病变和重建三维心血管，首要任务是将造影图像中的心血管表述成单个象素宽的心血管骨架，而心血管的提取又是心血管骨架提取和决定其提取准确性的关键。首先采用了旋转高斯函数对图像进行增强。然后提出了自适应跟踪圆模板对增强后的心血管图像进行心血管提取。该提取方法与以往的提取方法相比，鲁棒性增强，为心血管骨架提取、心血管病变的定量描述和心血管三维重建奠定了基础。