基于深度递归级联卷积神经网络的并行磁共振成像方法

快速磁共振成像是磁共振研究领域重要的课题之一.随着大数据和深度学习的兴起,神经网络成为快速磁共振技术的重要方法.然而网络性能表现和网络参数量之间较难取得平衡,且对于多通道数据重建的并行成像问题,相关研究较少.本文构建了一种深度递归级联卷积神经网络结构,用于处理并行成像问题.这种网络结构在减少网络参数量的同时,能够尽可能地提高网络的表达能力,提高网络重建的精确度.实验结果表明,相较于传统并行成像方法,通过训练好的神经网络对欠采样磁共振数据进行重建,可以得到更准确的重建结果,且重建时间大大缩短.     

结合残差U-Net神经网络和DIP的PET图像降噪

由于正电子发射型计算机断层显像(PET)噪声较大,现有图像降噪效果不理想,提出了一种结合残差U-Net神经网络和深度图像先验(DIP)的PET图像降噪.在U-Net网络中引入残差学习,提高网络表达能力和收敛速度;提出一种无训练数据的DIP算法,将神经网络解释为图像的参数化,利用图像噪声参数化后呈现高阻抗的特性将其去除,...     

基于深度学习的混沌加密灰度图像重建方法

混沌加密由于其初始值敏感性、伪随机性和运动轨迹的不可预测性而被广泛应用于图像加密领域。提出了一种通过深度学习来攻击Lorenz混沌加密系统的灰度图像重建方法,通过残差网络实现了对一系列明文-密文对数据集进行训练,从而拟合出密文到明文的过程,然后将训练好的网络应用在独立于训练集的密文上,恢复出与明文非常接近的图像。数值仿真结果验证了这种灰度图像重建方法的有效性。     

基于Dense-UNet++的关节滑膜磁共振图像分割

为解决以往基于深度学习的滑膜磁共振图像分割模型存在的分割精度较低、鲁棒性较差、训练耗时等问题,本文提出了一种基于Dense-UNet++网络的新模型,将DenseNet模块插入UNet++网络中,并使用Swish激活函数进行训练．利用1 036张滑膜磁共振图像数据增广后的14 512张滑膜图像对模型进行训练,并利用68张图像进行测试．结果显示,模型的平均DSC系数为0.819 9,交叉联合度量（IOU）为0.927 9．相较于UNet、ResUNet和VGG-UNet++网络结构,DSC系数和IOU均有提升,DSC振荡系数降低．另外在应用于相同滑膜图像数据集和使用相同的网络结构时,Swish函数相比ReLu函数有助于提升分割精度．实验结果表明,本文提出的算法对于滑膜磁共振图像的病灶区域的分割有较好的效果,能够辅助医生对病情做出判断．     

基于改进深度残差网络的计算断层扫描图像分类算法

基于改进的深度残差网络(ResNet),提出更加适合肺部组织的计算断层扫描(CT)图像模式分类模型。为克服医学图像分析中可用数据集稀少的困难,采用迁移学习方法来减小神经网络模型对数据量大的需求,以减小过拟合。迁移学习的策略是将肺内大量可用的无标签区域作为预训练的数据,使用深度互信息最大化和先验分布匹配的方法进行无监督表征学习。通过对比实验发现,改进的深度ResNet可以得到更高的分类精度,迁移学习算法可以有效地利用肺内无标签区域的数据,从而提升网络模型的分类表现。     

基于虚拟线圈和卷积神经网络的多层同时激发图像重建

本文提出一种基于虚拟共轭线圈（Virtual Coil Concept,VCC）技术和k空间插值鲁棒人工神经网络（Robust Artificial-neural-networks for k-space Interpolation,RAKI）的图像重建方法,用于磁共振多层同时激发成像（Simultaneous Multi-Slice imaging,SMS）,该方法能够有效提升重建图像的质量,被命名为VIRGINIA（VIRtual conjuGate coIls Neural-networks InterpolAtion）.为了得到更高质量的SMS图像,本文提出的VIRGINIA方法利用磁共振线圈数据的复数共轭对称性质扩展了SMS所获取的多通道数据,并将扩展后的数据用于RAKI网络的训练,利用训练后的网络实现高质量的SMS图像重建.本文将VIRGINIA方法和其他SMS图像重建方法（RAKI和Slice-GRAPPA方法）进行了对比,并采用结构相似指数（Structural Similarity Index,SSIM）、峰值信噪比（Peak Signal-to-Noise Ratio,PSNR）和均方根误差（Root Mean Square Error,RMSE）对不同方法的重建图像进行了量化对比分析.结果显示,在相同的SMS加速倍数下,使用VIRGINIA方法进行重建的图像质量均好于RAKI方法,且远好于传统Slice-GRAPPA方法.     

基于先验图像约束的多光谱压缩感知

在低采样率、低信噪比(SNR)的探测条件下,多光谱重建图像噪声增多,重建质量大幅度降低。为了提高多光谱图像的重建质量,提出了一种基于先验图像约束的多光谱压缩感知(PICHCS)重建方法。PICHCS利用多光谱图像的空间相关性和谱间相关性重建出初始图像,并将相邻谱段的初始图像取平均获得高信噪比的先验图像。先验图像与目标图像相减可以使优化目标稀疏化,并使得重建结果具有与先验图像类似的高信噪比特性。通过数值模拟和实验验证了该重建算法的可行性,并在不同的采样率、信噪比条件下和全变差低秩联合重建算法进行了对比研究。结果表明,PICHCS可以在低采样率低信噪比情况下提高多光谱图像的重建质量,从而降低对数据采样率和系统信噪比的要求。     

利用CAD模型的不完全扫描CT图像重建

工业计算机断层成像(CT)扫描大尺寸和高密度工件时,会出现穿不透、扫描角度有限导致的投影数据不完备、重建伪影严重等问题。基于此,提出一种将工件的CAD设计模型作为先验知识,来实现该类对象的有限角CT重建优质图像的方法。由工件CT扫描的断层位置计算CAD模型对应的分层位置,并得到模型的像素截面;根据CT系统X射线能量和模型中各部分材质,确定和生成一幅衰减系数图像,并将其配准到一幅零灰度图像中,得到先验图像;最后对扫描投影数据进行SART+TVM+PRIOR算法重建,得到重建图像。仿真实验和实际工件扫描实验的结果显示,加入先验图像后重建的图像质量要远远优于未加入先验图像的重建结果,边缘结构更加清晰,并极大地减少了伪影。     

双树小波变换与小波树稀疏联合的低场CS-MRI算法

压缩感知理论常用在磁共振快速成像上,仅采样少量的K空间数据即可重建出高质量的磁共振图像.压缩感知磁共振成像技术的原理是将磁共振图像重建问题建模成一个包含数据保真项、稀疏先验项和全变分项的线性组合最小化问题,显著减少磁共振扫描时间.稀疏表示是压缩感知理论的一个关键假设,重建结果很大程度上依赖于稀疏变换.本文将双树复小波变换和小波树稀疏联合作为压缩感知磁共振成像中的稀疏变换,提出了基于双树小波变换和小波树稀疏的压缩感知低场磁共振图像重建算法.实验表明,本文所提算法可以在某些磁共振图像客观评价指标中表现出一定的优势.     

基于正则化迭代的并行磁共振图像重建算法

为了解决并行磁共振成像过程的病态性和图像信噪比下降问题,降低重建过程中噪声放大和异常值的干扰造成的图像信噪比的损失,提出了一种基于正则化共轭梯度迭代的并行磁共振成像重建算法;该算法基于最小二乘理论,引入正则化,优化方程,进而进行迭代重建;采用了不同加速因子的人脑磁共振K空间欠采样数据以验证该算法的重建性能,仿真结果表明了该算法相较于最小二乘法,能较大限度地降低噪声对重建结果的干扰,具有信噪比更高、误差更小、成像效果更好等特征;重建图像质量得到了较好的改善,对临床诊断更具有适用性。     

High-performance rapid MR parameter mapping using model-based deep adversarial learning

PurposeTo develop and evaluate a deep adversarial learning-based image reconstruction approach for rapid and efficient MR parameter mapping.MethodsThe proposed method provides an image reconstruction framework by combining the end-to-end convolutional neural network (CNN) mapping, adversarial learning, and MR physical models. The CNN performs direct image-to-parameter mapping by transforming a series of undersampled images directly into MR parameter maps. Adversarial learning is used to improve image sharpness and enable better texture restoration during the image-to-parameter conversion. An additional pathway concerning the MR signal model is added between the estimated parameter maps and undersampled k-space data to ensure the data consistency during network training. The proposed framework was evaluated on T₂ mapping of the brain and the knee at an acceleration rate R = 8 and was compared with other state-of-the-art reconstruction methods. Global and regional quantitative assessments were performed to demonstrate the reconstruction performance of the proposed method.ResultsThe proposed adversarial learning approach achieved accurate T₂ mapping up to R = 8 in brain and knee joint image datasets. Compared to conventional reconstruction approaches that exploit image sparsity and low-rankness, the proposed method yielded lower errors and higher similarity to the reference and better image sharpness in the T₂ estimation. The quantitative metrics were normalized root mean square error of 3.6% for brain and 7.3% for knee, structural similarity index of 85.1% for brain and 83.2% for knee, and tenengrad measures of 9.2% for brain and 10.1% for the knee. The adversarial approach also achieved better performance for maintaining greater image texture and sharpness in comparison to the CNN approach without adversarial learning.ConclusionThe proposed framework by incorporating the efficient end-to-end CNN mapping, adversarial learning, and physical model enforced data consistency is a promising approach for rapid and efficient reconstruction of quantitative MR parameters.     

Non-blind image deconvolution using natural image gradient prior

Since non-blind image deconvolution is inherently ill-posed, the results of unregularized methods are often contaminated by noise and ringing artifacts. To reach a stable solution, we adopt the natural image gradient prior to regularize the latent image and obtain an improved version of the Richardson–Lucy (RL) algorithm. We use both synthetic and real world blurred images to test the proposed method. Experimental results show that the negative artifacts are significantly suppressed and the restored images are of high quality.     

Transfer learning in deep neural network based under-sampled MR image reconstruction

In Magnetic Resonance Imaging (MRI), the success of deep learning-based under-sampled MR image reconstruction depends on: (i) size of the training dataset, (ii) generalization capabilities of the trained neural network. Whenever there is a mismatch between the training and testing data, there is a need to retrain the neural network from scratch with thousands of MR images obtained using the same protocol. This may not be possible in MRI as it is costly and time consuming to acquire data. In this research, a transfer learning approach i.e. end-to-end fine tuning is proposed for U-Net to address the data scarcity and generalization problems of deep learning-based MR image reconstruction. First the generalization capabilities of a pre-trained U-Net (initially trained on the human brain images of 1.5 T scanner) are assessed for: (a) MR images acquired from MRI scanners of different magnetic field strengths, (b) MR images of different anatomies and (c) MR images under-sampled by different acceleration factors. Later, end-to-end fine tuning of the pre-trained U-Net is proposed for the reconstruction of the above-mentioned MR images (i.e. (a), (b) and (c)). The results show successful reconstructions obtained from the proposed method as reflected by the Structural SIMilarity index, Root Mean Square Error, Peak Signal-to-Noise Ratio and central line profile of the reconstructed images.     

Noise suppression and details enhancement for infrared image via novel prior

Infrared images always suffer from blurring edges, fewer details and low signal-to-noise ratio. So, sharpening edges and suppressing noise become the urgent techniques in infrared image technology field. However, they are contradictories in most cases. Hence, to depict correctly infrared image features under low signal-to-noise ratio circumstance, a novel prior, which is immune to noise, is presented in this paper. The proposed method scopes noise suppression and details enhancement. In noise suppression, the prior is introduced into Bayesian model to obtain optimal estimation through iteration. In details enhancement, based on the proposed prior, the final image is obtained by the improved unsharp mask algorithm which enhances adaptively details and edges of optimal estimation. The effectiveness and robustness of the proposed method is analyzed by testing the infrared images obtained from different signal-to-noise ratio conditions. Compared with other well-established methods, the proposed method shows a significant performance in terms of noise suppression, actual scene reappearance, enhancing the details and sharpening edges.     

Incorporating prior knowledge via volumetric deep residual network to optimize the reconstruction of sparsely sampled MRI

For sparse sampling that accelerates magnetic resonance (MR) image acquisition, non-linear reconstruction algorithms have been developed, which incorporated patient specific a prior information. More generic a prior information could be acquired via deep learning and utilized for image reconstruction. In this study, we developed a volumetric hierarchical deep residual convolutional neural network, referred to as T-Net, to provide a data-driven end-to-end mapping from sparsely sampled MR images to fully sampled MR images, where cartilage MR images were acquired using an Ultra-short TE sequence and retrospectively undersampled using pseudo-random Cartesian and radial acquisition schemes. The network had a hierarchical architecture that promoted the sparsity of feature maps and increased the receptive field, which were valuable for signal synthesis and artifact suppression. Relatively dense local connections and global shortcuts were established to facilitate residual learning and compensate for details lost in hierarchical processing. Additionally, volumetric processing was adopted to fully exploit spatial continuity in three-dimensional space. Data consistency was further enforced. The network was trained with 336 three-dimensional images (each consisting of 32 slices) and tested by 24 images. The incorporation of a priori information acquired via deep learning facilitated high acceleration factors (as high as 8) while maintaining high image fidelity (quantitatively evaluated using the structural similarity index measurement). The proposed T-Net had an improved performance as compared to several state-of-the-art networks.     

A multi-scale residual network for accelerated radial MR parameter mapping

A deep learning MR parameter mapping framework which combines accelerated radial data acquisition with a multi-scale residual network (MS-ResNet) for image reconstruction is proposed. The proposed supervised learning strategy uses input image patches from multi-contrast images with radial undersampling artifacts and target image patches from artifact-free multi-contrast images. Subspace filtering is used during pre-processing to denoise input patches. For each anatomy and relaxation parameter, an individual network is trained. in vivo T₁ mapping results are obtained on brain and abdomen datasets and in vivo T₂ mapping results are obtained on brain and knee datasets. Quantitative results for the T₂ mapping of the knee show that MS-ResNet trained using either fully sampled or undersampled data outperforms conventional model-based compressed sensing methods. This is significant because obtaining fully sampled training data is not possible in many applications. in vivo brain and abdomen results for T₁ mapping and in vivo brain results for T₂ mapping demonstrate that MS-ResNet yields contrast-weighted images and parameter maps that are comparable to those achieved by model-based iterative methods while offering two orders of magnitude reduction in reconstruction times. The proposed approach enables recovery of high-quality contrast-weighted images and parameter maps from highly accelerated radial data acquisitions. The rapid image reconstructions enabled by the proposed approach makes it a good candidate for routine clinical use.     

Single image deblurring using novel image prior constraints

Single image deblurring is a highly ill-posed problem and requires to be regularized. Many common forms of image prior have a major drawback that is unable to make full use of local image information. In this paper, we propose a single image deblurring method using novel image prior constraints. We establish a probabilistic model by enforcing inspired image prior constraints and adopt an advanced iterative scheme that alternates between blur kernel estimation and non-blind image restoration. To suppress ringing artifacts caused by inevitable blur kernel estimated errors, our method employs total variation image restoration and presents an alternation half-quadratic algorithm to solve the non-convex cost function. Finally, experiments show that our method has good performance in suppressing ringing artifacts, and makes a good balance between alleviating staircase effects and preserving image details.     

Blind image deconvolution using the Fields of Experts prior

In this paper, we present a method for single image blind deconvolution. To improve its ill-posedness, we formulate the problem under Bayesian probabilistic framework and use a prior named Fields of Experts (FoE) which is learnt from natural images to regularize the latent image. Furthermore, due to the sparse distribution of the point spread function (PSF), we adopt a Student-t prior to regularize it. An improved alternating minimization (AM) approach is proposed to solve the resulted optimization problem. Experiments on both synthetic and real world blurred images show that the proposed method can achieve results of high quality.     

IKWI-net: A cross-domain convolutional neural network for undersampled magnetic resonance image reconstruction

Magnetic resonance imaging (MRI) is widely used to get the information of anatomical structure and physiological function with the advantages of high resolution and non-invasive scanning. But the long acquisition time limits its application. To reduce the time consumption of MRI, compressed sensing (CS) theory has been proposed to reconstruct MRI images from undersampled k-space data. But conventional CS methods mostly use iterative methods that take lots of time. Recently, deep learning methods are proposed to achieve faster reconstruction, but most of them only pay attention to a single domain, such as the image domain or k-space. To take advantage of the feature representation in different domains, we propose a cross-domain method based on deep learning, which first uses convolutional neural networks (CNNs) in the image domain, k-space and wavelet domain simultaneously. The combined order of the three domains is also first studied in this work, which has a significant effect on reconstruction. The proposed IKWI-net achieves the best performance in various combinations, which utilizes CNNs in the image domain, k-space, wavelet domain and image domain sequentially. Compared with several deep learning methods, experiments show it also achieves mean improvements of 0.91 dB in peak signal-to-noise ratio (PSNR) and 0.005 in structural similarity (SSIM).     

Compressed sensing MR image reconstruction using a motion-compensated reference

Compressed sensing (CS)-based methods have been proposed for image reconstruction from undersampled magnetic resonance data. Recently, CS-based schemes using reference images have also been proposed to further reduce the sampling requirement. In this study, we propose a new reference-constrained CS reconstruction method that accounts for the misalignment between the reference and the target image to be reconstructed. The proposed method uses a new image model that represents the target image as a linear combination of a motion-dependent reference image and a sparse difference image. We then use an efficient iterative algorithm to jointly estimate the motion parameters and the difference image from sparsely sampled data. Simulation results from a numerical phantom data set and an in vivo data set show that the proposed method can accurately compensate the motion effects between the reference and the target images and improve reconstruction quality. The proposed method should prove useful for several applications such as interventional imaging, longitudinal imaging studies and dynamic contrast-enhanced imaging.