利用相关比相似性测度多分辨率配准MR和PET医学图像的方法

采用了全新的相关比相似性测度作为配准的测度准则，提出了有效的磁共振(MR)和正电子发射端层扫描(PET)临床医学图像配准方法。具体设计时，采用了加速的多分辨率的配准方案，对方案中涉及的几何变换选取、重采样、多分辨率体数据表达及最优化方法进行了详细的设计分析。最后，利用多分辨率配准方法，对MR和PET临床医学图像进行配准，给出了令人满意的效果，同时和基于体素灰度的直接配准法相比，配准速度也有了很大提高。     

基于互信息的2D-3D医学图像配准

介绍一种基于互信息的2D-3D医学图像配准方法，将此方法用于X线透射图(由DRR模拟)与CT体数据的配准，对具体实现方法进行了探讨，用数据采样和灰度级压缩的方法对体数据进行预处理，通过改变平行光入射角度，计算不同方向的DRR图像，用互信息方法配准DRR图像，得到了较好的2D-3D图像配准效果。     

基于最大互信息的人脑MR-PET图像配准方法

利用最大互信息法进行多模医学图像配准近来成为医学图像处理领域的热点。MR和PET图像配准对研究神经组织的结构关系和引导神经外科手术有着重要的指导意义。本文描述了一种基于互信息的人脑MR－PET图像配准方法。我们将这种方法应用于图像的几何对准并给出了初步的评估结果。由于不需要对不同成像模式下的图像灰度间的关系作任何假设,最大互信息法是一种稳健性强,可广泛应用于基于体素的多模图像的配准方法。     

利用边缘和表面特征实现脑图像的自动配准

本文中我们使用基于CT、MR和PET图像等值特征表面的配准算法对多模医学图像进行了配准研究，在CT、MR和PET的原始图像中提取等值特征表面，进行图像的几何对准，并对结果进行初步评估，同时对该算法的稳健性，搜索最近点策略和采样策略进行了研究，结果表明；这种方法能够达到亚像素级的配准精度，是一种稳健、高精度、全自动的配准方法。     

多模态医学图像配准技术研究

目的:通过对CT和PET图像进行配准实验,得出CT和PET配准的最优插值搜索方法以及粗配准对二者配准结果的影响。方法:首先利用改进的力矩与主轴法对CT和PET图像进行粗配准,得到CT图像的平移量与旋转量。然后针对不同的插值和搜索方法实现互信息配准,并将所得平移量与旋转量作为Powell优化初值进行实验,与使用默认值的Powell优化方法进行对比。结果:实验结果表明:使用双线性插值算法时,两实验的结果基本一致;采用部分体积插值法时,基于粗配准的Powell优化配准效果优于基于默认值的方法;而使用立方卷积插值法时二者配准效果差别不大,但基于粗配准的方法时间更短。结论:基于粗配准的Powell优化方法(部分体积插值与Brent搜索)有更好的配准速度及效果。     

基于多层次变换和优化方法的PET-CT图像弹性配准

目的为校正PET/CT一体化仪器的误配准和形变问题。方法本文提出一种基于多层次变换和优化的弹性配准方法,以互信息为相似性测度,先进行全局刚性变换使用单纯形优化策略,在刚性变换的基础上使用B样条形变变换并使用有限内存的BFGSB(limited memory Broyden Fietcher Goldfarb Shanno bound,LBFGSB)优化算法。选取3个患者PET/CT联合扫描下的6组数据对本文算法进行验证,并与采用传统单一变换方法和单一优化策略的方法进行了对比。结果采用本文提出的配准算法得到的平均配准时间为19.78 s,平均互信息值为0.3115,较传统仅使用单一变换和单一优化的配准方法在配准速度和精度上有一定的提高。结论基于多层次变换和优化方法的图像配准方法能够较好地解决PET产生的形变问题,配准速度快精度高,可以用来校正PET/CT联合扫描由于呼吸运动和脏器运动产生的误配准和形变问题。     

灰度级别对基于互信息医学图像配准方法的影响

医学图像配准在医学图像处理领域中已经被广泛使用。基于互信息配准的方法具有自动化程度高、配准精度高等优点。基于互信息的配准方法实质上是一种进行灰度统计和计算的方法 ,因此同一图像采用不同的灰度表示必然会影响配准结果。在分析灰度级别的压缩对于图像质量的影响和基于互信息配准方法的影响的基础上 ,进行了一系列的多模态医学图像配准试验 ,从配准精度和计算时间两个方面比较了不同的灰度级别对图像配准的影响。在详细分析和比较不同级别图像配准结果的基础上 ,给出了基于互信息配准时所采用的合理灰度级别的建议。     

基于多目标交叉变异粒子群算法的多模医学图像非刚性配准

以互信息为相似性测度,采用B样条变换对多模态医学图像进行非刚性配准时,由于噪声及图像插值等原因造成的互信息局部极值使得传统优化方法不能搜索到最佳配准参数。为此,使用粒子群智能优化方法作为搜索策略,以降低对图像预处理的要求,进一步提高基于互信息的非刚性配准的鲁棒性。为了克服粒子群算法受初始值选取等因素的影响易陷于局部最优的缺点,使用LBFGS优化得到的结果构造初始粒子群,采用多目标优化方法结合交叉变异策略加以改进,使得算法在解空间搜索的遍历性得到改善,优化结果更接近全局最优。MR-T2与MR-PD图像的配准实验证明,该方法提高了基于互信息的B样条非刚性配准的鲁棒性,配准率达到94%;CT与PET图像的配准实验表明该方法相比惯性权重粒子群算法提高了配准精度,互信息增加了0.026;另外,CT与CBCT图像的配准实验也验证了本方法的有效性。     

基于互信息的人脑图像配准研究

近来利用互信息进行多模医学图像配准已成为医学图像处理领域的热点,人脑多模医学图像配准对研究神经组织的结构功能关系和引导神经外科手术有着重要的指导意义,本文描述了一种基于互信息的人脑图像配准方法,我们将这种方法应用于图像的几何对准并给出了初步的评估结果,同时,我们还就归一化互信息、多分辨率策略,多种插值和优化算法对配准速度和精度的影响作了讨论,由于不需要对不同成像模式下的图像灰度间的关系作任何假设,互信息法是一种稳健性强、可广泛应用于基于体素的多模医学图像的配准方法。     

鲍威尔和模拟退火优化算法结合的多分辨率三维图像配准

基于互信息的全局配准算法是近年图像配准研究的热点之一，此法具有精度高、鲁棒性强的特点。现将互信息作为相似性测度，采用模拟退火与鲍威尔相结合的多分辨率搜索优化策略，对临床使用的三维CT、MR图像进行了刚体配准。实验结果表明此方法能有效地防止优化陷入局部极值，具有良好的实用性。     

基于三次样条层间插值的多模态医学图像配准

我们从PET-CT多模态图像序列的特点出发,提出了一种全新的图像配准及融合方法,它采用三次样条插值法对PET-CT图像进行层间插值,然后再利用最大互信息法进行配准,最后应用改进的主成分分析(PCA)法融合PET-CT图像用以增强PET显像效果,从而得到满意的配准以及融合结果。用三次样条插值法进行层间插值并恢复层间缺失图像的信息,弥补了现有配准方法的不足,提高了配准精度,使融合后的图像更加接近实际的物理断层。该方法已经成功应用于三维适形放疗(3D-CRT)系统的开发中。     

基于自由变形法的多模态医学图像的配准与融合

本研究提出了一种自动识别颈部PET-CT图像特征点的算法，它应用自由变形（FFD）方法以CT图像的特征点为参考使PET图像产生变形，再结合最大互信息法对颈部PET与CT图像进行非刚体配准，最后用改进的小波图像融合法把两者进行融合得出视觉效果比较理想的融合图像。经实际计算得出的变形PET图像与对应CT图像的互信息量大于原始PET图像，并且最后用改进的小波图像融合法得出的融合图像的信息量比一般小波融合大，由此证明本研究所用方法是有效的。     

Non-rigid image registration using a median-filtered coarse-to-fine displacement field and a symmetric correlation ratio

Conventional approaches to image registration are generally limited to image-wide rigid transformations. However, the body and its internal organs are non-rigid structures that change shape due to changes in the body's posture during image acquisition, and due to normal, pathological and treatment-related variations. Inter-subject matching also constitutes a non-rigid registration problem. In this paper, we present a fully automated non-rigid image registration method that maximizes a local voxel-based similarity metric. Overlapping image blocks are defined on a 3D grid. The transformation vector field representing image deformation is found by translating each block so as to maximize the local similarity measure. The resulting sparsely sampled vector field is median filtered and interpolated by a Gaussian function to ensure a locally smooth transformation. A hierarchical strategy is adopted to progressively establish local registration associated with image structures at diminishing scale. Simulation studies were carried out to evaluate the proposed algorithm and to determine the robustness of various voxel-based cost functions. Mutual information, normalized mutual information, correlation ratio (CR) and a new symmetric version of CR were evaluated and compared. A T1-weighted magnetic resonance (MR) image was used to test intra-modality registration. Proton density and T2-weighted MR images of the same subject were used to evaluate inter-modality registration. The proposed algorithm was tested on the 2D MR images distorted by known deformations and 3D images simulating inter-subject distortions. We studied the robustness of cost functions with respect to image sampling. Results indicate that the symmetric CR gives comparable registration to mutual information in intra- and inter-modality tasks at full sampling and is superior to mutual information in registering sparsely sampled images.     

基于深度学习的2D/3D医学图像配准研究

2D/3D配准在临床诊断和手术导航规划中有着广泛的应用,可解决医学图像领域中不同维度图像存在信息缺失的问题,能辅助医生在术中精准定位患者的病灶。常规的2D/3D配准方法主要依赖于图像的灰度进行配准,但非常耗时,不利于临床实时性的需求,并且配准过程中容易陷入局部最优值。提出用深度学习的方法来解决2D/3D医学图像配准问题。采用一个基于深度学习的卷积神经网络,通过网络对数字影像重建技术(DRR)进行训练并自动学习图像特征,预测X光图像所对应的参数,从而实现配准。以人体骨盆的模型骨为实验对象,根据骨盆的CT数据生成36000张DRR图像作为训练集,同时通过C臂采集模型骨的50张X光图像作为验证。结果显示,深度学习算法在相关系数、归一化互信息、欧式距离3个精度评价指标上的测试值分别为0.82±0.07、0.32±0.03、61.56±10.91,而常规2D/3D算法对应的测试值分别为0.79±0.07、0.29±0.03、37.92±7.24,说明深度学习算法的配准精度优于常规2D/3D算法的配准精度,且不存在陷入局部最优值的问题。同时,深度学习的配准时间约为0.03s,远低于常规2D/3D配准的时间,可满足临床对于实时配准的需求,未来将进一步开展临床数据的2D/3D配准研究。     

Robust and fast shell registration in PET and MR/CT brain images

A robust and fast hybrid method using a shell volume that consists of high contrast voxels with their neighbors is proposed for registering PET and MR/CT brain images. Whereas conventional hybrid methods find the best matched pairs from several manually selected or automatically extracted local regions, our method automatically selects a shell volume in the PET image, and finds the best matched corresponding volume using normalized mutual information (NMI) in overlapping volumes while transforming the shell volume into an MR or CT image. A shell volume not only can reduce irrelevant corresponding voxels between two images during optimization of transformation parameters, but also brings a more robust registration with less computational cost. Experimental results on clinical data sets showed that our method successfully aligned all PET and MR/CT image pairs without losing any diagnostic information, while the conventional registration methods failed in some cases.     

基于改进最大互信息法的MR切片图像配准

医学图像配准是医学图像处理分析的关键步骤,是医学图像融合首先要解决的问题。本研究的主要目的是实现帕金森患者深脑部刺激手术前后MR图像的配准。将互距离引入互信息测度,实现手术前后两组MR切片图像的对应匹配,然后将对应的两组MR切片系列重建三维图像,最后用Powell优化算法对重建的三维图像进行配准。通过术前术后MR三维图像的配准,可以定量的分析手术后植入电极和手术前丘脑底核的相对位置关系,从而实现对深脑部刺激手术质量的科学评估。     

Automatic deformable PET/MRI registration for preclinical studies based on B-splines and non-linear intensity transformation

PET images deliver functional data, whereas MRI images provide anatomical information. Merging the complementary information from these two modalities is helpful in oncology. Alignment of PET/MRI images requires the use of multi-modal registration methods. Most of existing PET/MRI registration methods have been developed for humans and few works have been performed for small animal images. We proposed an automatic tool allowing PET/MRI registration for pre-clinical study based on a two-level hierarchical approach. First, we applied a non-linear intensity transformation to the PET volume to enhance. The global deformation is modeled by an affine transformation initialized by a principal component analysis. A free-form deformation based on B-splines is then used to describe local deformations. Normalized mutual information is used as voxel-based similarity measure. To validate our method, CT images acquired simultaneously with the PET on tumor-bearing mice were used. Results showed that the proposed algorithm outperformed affine and deformable registration techniques without PET intensity transformation with an average error of 0.72?±?0.44 mm. The optimization time was reduced by 23% due to the introduction of robust initialization. In this paper, an automatic deformable PET-MRI registration algorithm for small animals is detailed and validated.

Open image in new window

Graphical abstract ?

     

利用数学形态学工具配准人脑MR-PET图像

数学形态学是以形态为基础对图像进行分析的数学工具。我们利用简单的形态学操作从PET和MR图像中提取对应的解剖结构--皮质层，以归一化互信息为相似性测度，进行医学PET/MR图像的几何对准。评估结果证明，该方法的配准精度达到亚像素精度，并能极大地节省运行时间，是一种稳健性强、精度高、全自动的多模医学图像配准新方法。     

Matching PET and CT scans of the head and neck area: development of method and validation

Positron emission tomography (PET) provides important information on tumor biology, but lacks detailed anatomical information. Our aim in the present study was to develop and validate an automatic registration method for matching PET and CT scans of the head and neck. Three difficulties in achieving this goal are (1) nonrigid motions of the neck can hamper the use of automatic ridged body transformations; (2) emission scans contain too little anatomical information to apply standard image fusion methods; and (3) no objective way exists to quantify the quality of the match results. These problems are solved as follows: accurate and reproducible positioning of the patient was achieved by using a radiotherapy treatment mask. The proposed method makes use of the transmission rather than the emission scan. To obtain sufficient (anatomical) information for matching, two bed positions for the transmission scan were included in the protocol. A mutual information-based algorithm was used as a registration technique. PET and CT data were obtained in seven patients. Each patient had two CT scans and one PET scan. The datasets were used to estimate the consistency by matching PET to CT1, CT1 to CT2, and CT2 to PET using the full circle consistency test. It was found that using our method, consistency could be obtained of 4 mm and 1.3 degrees on average. The PET voxels used for registration were 5.15 mm, so the errors compared quite favorably with the voxel size. Cropping the images (removing the scanner bed from images) did not improve the consistency of the algorithm. The transmission scan, however, could potentially be reduced to a single position using this approach. In conclusion, the represented algorithm and validation technique has several features that are attractive from both theoretical and practical point of view, it is a user-independent, automatic validation technique for matching CT and PET scans of the head and neck, which gives the opportunity to compare different image enhancements.