磁共振成像PROPELLER采样数据重建中的运动估计新算法

PROPELLER（推进器）采样技术能够利用K空间中心重叠采样区域的数据来估计采集过程中受检查者的运动进而加以补偿,对运动伪影的消除效果非常显著。然而,由于其重建时的运动估计是基于最大化频域空间上相关系数的配准算法,该算法为了实现旋转估计与平移估计的分离,在进行旋转估计时,仅仅采用K空间数据的模,在数据量有限的情况下造成估计精度较低,在重建图像上表现为模糊及星条状伪影。本研究基于最大化图像空间上的互信息提出一种PROPELLER采样数据的运动估计新算法,首先由每个K空间带进行傅立叶逆变换后取模重建出系列临时图像,对这些图像进行模糊增强后以互信息作为相似性测度迭代搜索最优的运动参数。实验证明,该方法能显著提高PROPELLER采样数据重建中运动估计与补偿的精度,从而更好地消除伪影,特别是用于有运动时T1加权头部成像时。     

Recent advances in breast MRI and MRS

Breast MRI is an area of intense research and is fast becoming an important tool for the diagnosis of breast cancer. This review covers recent advances in breast MRI, MRS, and image post-processing and analysis. Several studies have explored a multi-parametric approach to breast imaging that combines analysis of traditional contrast enhancement patterns and lesion architecture with novel methods such as diffusion, perfusion, and spectroscopy to increase the specificity of breast MRI studies. Diffusion-weighted MRI shows some potential for increasing the specificity of breast lesion diagnosis and is even more promise for monitoring early response to therapy. MRS also has great potential for increasing specificity and for therapeutic monitoring. A limited number of studies have evaluated perfusion imaging based on first-pass contrast bolus tracking, and these clearly identify that vascular indices have great potential to increase specificity. The review also covers the relatively new acquisition technique of MR elastography for breast lesion characterization. A brief survey of image processing algorithms tailored for breast MR, including registration of serial dynamic images, segmentation and extraction of morphological features of breast lesions, and contrast uptake modeling, is also included. Recent advances in MRI, MRS, and automated image analysis have increased the utility of breast MR in diagnosis, screening, management, and therapy monitoring of breast cancer.     

Bilateral symmetry analysis of breast MRI

Mammographic interpretation often uses symmetry between left and right breasts to indicate the site of potential tumour masses. This approach has not been applied to breast images obtained from MRI. We present an automatic technique for breast symmetry detection based on feature extraction techniques which does not require any efforts to co-register breast MRI data. The approach applies computer-vision techniques to detect natural biological symmetries in breast MR scans based on three objective measures of similarity: multiresolution non-orthogonal wavelet representation, three-dimensional intensity distributions and co-occurrence matrices. Statistical distributions that are invariant to feature localization are computed for each of the extracted image features. These distributions are later compared against each other to account for perceptual similarity. Studies based on 51 normal MRI scans of randomly selected patients showed that the sensitivity of symmetry detection rate approached 94%. The symmetry analysis procedure presented in this paper can be applied as an aid in detecting breast tissue changes arising from disease.     

Computerized method for nonrigid MR-to-PET breast-image registration

We have developed and tested a new simple computerized finite element method (FEM) approach to MR-to-PET nonrigid breast-image registration. The method requires five-nine fiducial skin markers (FSMs) visible in MRI and PET that need to be located in the same spots on the breast and two on the flanks during both scans. Patients need to be similarly positioned prone during MRI and PET scans. This is accomplished by means of a low gamma-ray attenuation breast coil replica used as the breast support during the PET scan. We demonstrate that, under such conditions, the observed FSM displacement vectors between MR and PET images, distributed piecewise linearly over the breast volume, produce a deformed FEM mesh that reasonably approximates nonrigid deformation of the breast tissue between the MRI and PET scans. This method, which does not require a biomechanical breast tissue model, is robust and fast. Contrary to other approaches utilizing voxel intensity-based similarity measures or surface matching, our method works for matching MR with pure molecular images (i.e. PET or SPECT only). Our method does not require a good initialization and would not be trapped by local minima during registration process. All processing including FSMs detection and matching, and mesh generation can be fully automated. We tested our method on MR and PET breast images acquired for 15 subjects. The procedure yielded good quality images with an average target registration error below 4 mm (i.e. well below PET spatial resolution of 6-7 mm). Based on the results obtained for 15 subjects studied to date, we conclude that this is a very fast and a well-performing method for MR-to-PET breast-image nonrigid registration. Therefore, it is a promising approach in clinical practice. This method can be easily applied to nonrigid registration of MRI or CT of any type of soft-tissue images to their molecular counterparts such as obtained using PET and SPECT.     

抑制MR成像中刚体运动伪影的新方法

目的从成像空间出发,提出了一种新的纠正MRI刚性平移运动伪影的方法.方法利用梯度读出方向和相位编码方向的运动特性不同,分别采用不同的方法来消除刚性运动伪影.首先读出方向的运动,通过追踪频谱边缘非零区域和零区域的偏差进行估计,然后在频谱反方向移动相同的量来消除;利用改进的Snake算法,即梯度向量场方法提取目标区域的边界,然后利用相位迭代恢复算法消除残留的读出方向亚像素级伪影和相位编码方向的伪影.结果按国际通用方法生成SL头颅模板,通过对模板的仿真试验,证明修正后图像的信噪比大大提高,验证了方法的有效性和可靠性.结论本文提出的方法能够有效消除MR图像平移运动伪影,与传统的相位迭代恢复算法相比,对于较大运动伪影修正效果更好.     

Extracting MRS discriminant functional features of brain tumors

The current challenge in automatic brain tumor classification based on MRS is the improvement of the robustness of the classification models that explicitly account for the probable breach of the independent and identically distributed conditions in the MRS data points. To contribute to this purpose, a new algorithm for the extraction of discriminant MRS features of brain tumors based on a functional approach is presented. Functional data analysis based on region segmentation (RSFDA) is based on the functional data analysis formalism using nonuniformly distributed B splines according to spectral regions that are highly correlated. An exhaustive characterization of the method is presented in this work using controlled and real scenarios. The performance of RSFDA was compared with other widely used feature extraction methods. In all simulated conditions, RSFDA was proven to be stable with respect to the number of variables selected and with respect to the classification performance against noise and baseline artifacts. Furthermore, with real multicenter datasets classification, RSFDA and peak integration (PI) obtained better performance than the other feature extraction methods used for comparison. Other advantages of the method proposed are its usefulness in selecting the optimal number of features for classification and its simplified functional representation of the spectra, which contributes to highlight the discriminative regions of the MR spectrum for each classification task. Copyright © 2012 John Wiley & Sons, Ltd.     

Motion correction for three-dimensional chemical exchange saturation transfer imaging without direct water saturation artifacts

In chemical exchange saturation transfer (CEST) MRI, motion correction is compromised by the drastically changing image contrast at different frequency offsets, particularly at the direct water saturation. In this study, a simple extension for conventional image registration algorithms is proposed, enabling robust and accurate motion correction of CEST-MRI data. The proposed method uses weighted averaging of motion parameters from a conventional rigid image registration to identify and mitigate erroneously misaligned images. Functionality of the proposed method was verified by ground truth datasets generated from 10 three-dimensional in vivo measurements at 3 T with simulated realistic random rigid motion patterns and noise. Performance was assessed using two different criteria: the maximum image misalignment as a measure for the robustness against direct water saturation artifacts, and the spectral error as a measure of the overall accuracy. For both criteria, the proposed method achieved the best scores compared with two motion-correction algorithms specifically developed to handle the varying contrasts in CEST-MRI. Compared with a straightforward linear interpolation of the motion parameters at frequency offsets close to the direct water saturation, the proposed method offers better performance in the absence of artifacts. The proposed method for motion correction in CEST-MRI allows identification and mitigation of direct water saturation artifacts that occur with conventional image registration algorithms. The resulting improved robustness and accuracy enable reliable motion correction, which is particularly crucial for an automated and carefree evaluation of spectral CEST-MRI data, e.g., for large patient cohorts or in clinical routines.     

基于频域约束和交叉融合特征网络的磁共振图像超分辨率重建算法

磁共振（MR）图像常用于临床医学诊断，获得高分辨率MR图像有利于进行医学分析。目前主流的基于参考的图像超分辨率重建算法重建的图像，其视觉效果取得了明显的提升，但仍存在明显的伪影问题。针对该问题，提出频域约束和交叉融合特征网络（FCCF）模型，即引入频域损失函数作为约束条件，并构建一种多分辨率特征融合机制，通过交叉融合不同分辨率的图像特征来提高生成图像的质量，使重建结果具有更清晰的细节，没有明显的伪影。在合成和真实的MR图像数据集上分别用PSNR和SSIM指标进行评估，实验结果明显优于现有的超分辨率重建方法。     

Quantitative three‐dimensional ultrashort echo time cones imaging of the knee joint with motion correction

Knee degeneration involves all the major tissues in the joint. However, conventional MRI sequences can only detect signals from long T2 tissues such as the superficial cartilage, with little signal from the deep cartilage, menisci, ligaments, tendons and bone. It is highly desirable to develop new sequences that can detect signal from all major tissues in the knee. We aimed to develop a comprehensive quantitative three‐dimensional ultrashort echo time (3D UTE) cones imaging protocol for a truly “whole joint” evaluation of knee degeneration. The protocol included 3D UTE cones actual flip angle imaging (3D UTE‐Cones‐AFI) for T₁ mapping, multiecho UTE‐Cones with fat suppression for T₂* mapping, UTE‐Cones with adiabatic T_1ρ (AdiabT_1ρ) preparation for AdiabT_1ρ mapping, and UTE‐Cones magnetization transfer (UTE‐Cones‐MT) for MT ratio (MTR) and modeling of macromolecular proton fraction (f). An elastix registration technique was used to compensate for motion during scans. Quantitative data analyses were performed on the registered data. Three knee specimens and 15 volunteers were evaluated at 3 T. The elastix motion correction algorithm worked well in correcting motion artifacts associated with relatively long scan times. Much improved curve fitting was achieved for all UTE‐Cones biomarkers with greatly reduced root mean square errors. The averaged T₁, T₂*, AdiabT_1ρ, MTR and f for knee joint tissues of 15 healthy volunteers were reported. The 3D UTE‐Cones quantitative imaging techniques (ie, T₁, T₂*, AdiabT_1ρ, MTR and MT modeling) together with elastix motion correction provide robust volumetric measurement of relaxation times, MTR and f of both short and long T₂ tissues in the knee joint.     

Pathohistology of costal cartilage and immunomorphologic characteristics of collagen in funnel chest

Biopsy samples from the costal cartilage tissue were studied for 68 children with funnel deformity and from 20 children with normally formed chest. The authors present general morphologic features characteristic of the costal cartilage structure in norm and in case of funnel chest. These features include vast acellular sites, map-like areas, unmasked chondrin fibers and "marrow" cavities. In funnel chest they develop, however, 6-7 years earlier, than normally, and are consequent stages of the accelerated costal cartilage involution. Costal cartilage matrix was found to have an increased content of fibronectin and V collagen type in case of funnel chest. Besides III and IV procollagen types were noted in the chondrocytes. The accelerated growth of costal cartilages is involved in the formation of funnel chest.     

Binary Segmentation Masks Can Improve Intrasubject Registration Accuracy of Bone Structures in CT Images

Registration of bone structures is a common problem in medical research as well as in clinical applications. Intrasubject rigid 3D monomodality registration of segmented bone structures of CT images and multimodality registration of μMR and segmented μCT bone images were performed with the multiresolution intensity-based technique implemented in ITK. The registration results for binary volumes of interest (VOI) masks and for segmented gray value VOIs were compared. To determine the registration quality, in the monomodality case the sum of squared difference, the sum of absolute differences, and the normalized symmetric difference of binary masks and in the multimodality case Mattes mutual information were applied. The use of binary VOI masks was significantly superior to the use of gray value VOIs.     

Tensile Force Transmission in Human Patellar Tendon Fascicles Is Not Mediated by Glycosaminoglycans

Costal cartilage is much understudied compared with the load-bearing cartilages. Abnormally grown costal cartilages are associated with the inherited chest wall deformities pectus excavatum and pectus carinatum resulting in sunken and pigeon chests, respectively. A lack of understanding of the ultrastructural and molecular biology of costal cartilage is a major confounder in predicting causes and outcomes of these disorders. This study analyzed the structure of marginal human costal cartilage (ribs 6–10) through scanning electron and atomic force microscopes and identified the presence of straw-like structures running longitudinally. We also demonstrated that chondrocytes tend to occur singly or as doublets and that centrally located cells produce high levels of aggrecan compared with more peripherally located cells measured using immunohistochemistry. Gene expression from mRNA extracted from cartilage showed high levels of decorin expression, likely associated with the large, complex tubular structures running through this cartilage type. COL2A1, ACAN, and TIMP1 also showed higher levels of expression compared with ACTB. Analysis of gene expression ratios demonstrate that costal cartilage is under differentiated compared with published ratios for articular cartilage, likely due to the vastly different biomechanical environments of each cartilage type. Further studies need to establish whether findings described here from the costal margins are significantly different than the cartilage of the “true ribs” and how these values change with age.     

New Cardiac MRI Gating Method Using Event-Synchronous Adaptive Digital Filter

When imaging the heart using MRI, an artefact-free electrocardiograph (ECG) signal is not only important for monitoring the patient’s heart activity but also essential for cardiac gating to reduce noise in MR images induced by moving organs. The fundamental problem in conventional ECG is the distortion induced by electromagnetic interference. Here, we propose an adaptive algorithm for the suppression of MR gradient artefacts (MRGAs) in ECG leads of a cardiac MRI gating system. We have modeled MRGAs by assuming a source of strong pulses used for dephasing the MR signal. The modeled MRGAs are rectangular pulse-like signals. We used an event-synchronous adaptive digital filter whose reference signal is synchronous to the gradient peaks of MRI. The event detection processor for the event-synchronous adaptive digital filter was implemented using the phase space method—a sort of topology mapping method—and least-squares acceleration filter. For evaluating the efficiency of the proposed method, the filter was tested using simulation and actual data. The proposed method requires a simple experimental setup that does not require extra hardware connections to obtain the reference signals of adaptive digital filter. The proposed algorithm was more effective than the multichannel approach.     

一种基于互信息量的并行磁共振图像重建新算法

平方和算法是多线圈采集技术与并行成像中常用的一种图像重建方法。但是,在数据采集过程中,某些类型的运动常常会使个别位置上的线圈数据发生异常,采用平方和算法会对最终的重建图像质量产生很大影响。本研究提出一种新的并行磁共振图像重建算法——加权平方和方法。算法以线圈图像间最大互信息量作为判据来统计破坏数据,在之后的图像结合过程中赋予不同的权值,最大限度地降低破坏数据对最终结合图像造成的影响,有效解决运动对原有算法造成的破坏。本算法分别对多线圈并行采集的体模数据与真实脑部数据进行了实验,结果显示:相比于现有方法,新算法可以有效地抑制破坏数据在重建图像中产生的伪影,重建图像在细节分辨率上也有更好的表现。     

结合卷积神经网络与图卷积网络的乳腺癌病理图像分类研究

目的乳腺癌的精确诊断对于后续治疗具有重要临床意义,组织病理学分析是肿瘤诊断的金标准。卷积神经网络(convolution neural network,CNN)具有良好的局部特征提取能力,但无法有效捕捉细胞组织间的空间关系。为了有效利用这种空间关系,本文提出一种新的结合CNN与图卷积网络(graph convolution network,GCN)的病理图像分类框架,应用于乳腺癌病理图像的辅助诊断。方法首先对病理图像进行卷积及下采样得到一组特征图,然后将特征图上每个像素位置的特征向量表示为1个节点,构建具有空间结构的图,并通过GCN学习图中蕴含的空间结构特征。最后,将基于GCN的空间结构特征与基于CNN的全局特征融合,并同时对整个网络进行优化,实现基于融合特征的病理图像分类。结果本研究在提出框架下进行了3种GCN的比较,其中CNN-sGCN-fusion算法在2015生物成像挑战赛乳腺组织学数据集上获得93.53%±1.80%的准确率,在Databiox乳腺数据集上获得78.47%±5.33%的准确率。结论与传统基于CNN的病理图像分类算法相比,本文提出的结合CNN与GCN的算法有效融合了病理图像的全局特征与空间结构特征,从而提升了分类性能,具有潜在的应用可行性。     

Preclinical MR fingerprinting (MRF) at 7 T: effective quantitative imaging for rodent disease models

High‐field preclinical MRI scanners are now commonly used to quantitatively assess disease status and the efficacy of novel therapies in a wide variety of rodent models. Unfortunately, conventional MRI methods are highly susceptible to respiratory and cardiac motion artifacts resulting in potentially inaccurate and misleading data. We have developed an initial preclinical 7.0‐T MRI implementation of the highly novel MR fingerprinting (MRF) methodology which has been described previously for clinical imaging applications. The MRF technology combines a priori variation in the MRI acquisition parameters with dictionary‐based matching of acquired signal evolution profiles to simultaneously generate quantitative maps of T₁ and T₂ relaxation times and proton density. This preclinical MRF acquisition was constructed from a fast imaging with steady‐state free precession (FISP) MRI pulse sequence to acquire 600 MRF images with both evolving T₁ and T₂ weighting in approximately 30 min. This initial high‐field preclinical MRF investigation demonstrated reproducible and differentiated estimates of in vitro phantoms with different relaxation times. In vivo preclinical MRF results in mouse kidneys and brain tumor models demonstrated an inherent resistance to respiratory motion artifacts as well as sensitivity to known pathology. These results suggest that MRF methodology may offer the opportunity for the quantification of numerous MRI parameters for a wide variety of preclinical imaging applications. Copyright © 2015 John Wiley & Sons, Ltd.     

Evaluation of deformable image registration and a motion model in CT images with limited features

Deformable image registration (DIR) is increasingly used in radiotherapy applications and provides the basis for a previously described model of patient-specific respiratory motion. We examine the accuracy of a DIR algorithm and a motion model with respiration-correlated CT (RCCT) images of software phantom with known displacement fields, physical deformable abdominal phantom with implanted fiducials in the liver and small liver structures in patient images. The motion model is derived from a principal component analysis that relates volumetric deformations with the motion of the diaphragm or fiducials in the RCCT. Patient data analysis compares DIR with rigid registration as ground truth: the mean ± standard deviation 3D discrepancy of liver structure centroid positions is 2.0 ± 2.2 mm. DIR discrepancy in the software phantom is 3.8 ± 2.0 mm in lung and 3.7 ± 1.8 mm in abdomen; discrepancies near the chest wall are larger than indicated by image feature matching. Marker's 3D discrepancy in the physical phantom is 3.6 ± 2.8 mm. The results indicate that visible features in the images are important for guiding the DIR algorithm. Motion model accuracy is comparable to DIR, indicating that two principal components are sufficient to describe DIR-derived deformation in these datasets.     

一种基于稀疏分解去除EEG信号中MRI伪迹的新方法

在脑电图(Electroencephalography,EEG)和功能磁共振成像(Functional magnetic resonance imaging, FMRI)同时记录时,如何有效的去除混入EEG信号中的强磁共振(Magnetic resonance imaging,MRI)伪迹干扰信号是当前在EEG和FMRI的联合研究中面临的一个信号前期处理难点。主要从MRI干扰信号和EEG信号在时空上的差别出发,提出了一种基于混合过完备库的稀疏成分分析的分解方法,实现了强MRI干扰下的EEG信号的估计。在方法实现中,首先利用小波和离散余弦构造能体现MRI干扰和EEG时空特性差别的混合过完备库,然后通过匹配追踪(Matching pursuit,MP)方法在混合过完备库中的学习,实现MRI伪迹的消除。对模拟数据以及真实记录的混入了MRI干扰的EEG信号的估计实验结果,证实了该方法的有效性。