Bayesian image processing in magnetic resonance imaging.

In the past several years, image processing techniques based on Bayesian models have received considerable attention. In our earlier work, we developed a novel Bayesian approach which was primarily aimed at the processing and reconstruction of images in positron emission tomography. In this paper, we describe how the technique has been adopted to process magnetic resonance images in order to reduce noise and artifacts, thereby improving image quality. In this framework, the image is assumed to be a statistical variable whose posterior probability density conditional on the observed image is modeled by the product of the likelihood function of the observed data with a prior density based our prior knowledge. A Gibbs random field incorporating local continuity information and with edge-detection capability is used as the prior model. Based on the formalism of the posterior density, we can compute an estimate of the image using an iterative technique. We have implemented this technique and applied it to phantom and clinical images. Our results indicate that the approach works reasonably well for reducing noise, enhancing edges, and removing ringing artifact.     

A piecewise local regularized Richardson-Lucy algorithm for remote sensing image deconvolution

As it is not possible to obtain an accurate point spread function (PSF) in remote sensing imaging, classic deconvolution methods such as Wiener filtering often introduce strong noise and ringing artifacts, which contaminate the restored images. In this paper, we modify the standard Richarson-Lucy (RL) algorithm with a piecewise local regularization term and combine it with residual deconvolution method. Experimental results show that it is effective in suppressing negative effects, and images with rich details and sharp edges are obtained.     

Improved elimination of phase effects from background field inhomogeneities for susceptibility weighted imaging at high magnetic field strengths

To enhance susceptibility-related contrast of magnetic resonance images, the phase of susceptibility weighted data needs to be corrected for background inhomogeneities and phase wraps caused by them. Current methods either use homodyne filtering or a combination of phase unwrapping and high pass filtering. The drawback of homodyne filtering is incomplete elimination of phase wraps in areas with steep phase topography produced by background inhomogeneities of the static magnetic field. The disadvantage of phase unwrapping is that it requires subsequent high pass filtering, which introduces artifacts in areas with very steep transitions, such as areas near interfaces between parenchyma and bone or air. A method is proposed that reduces the artifacts associated with high pass filtering without sacrificing the advantages of phase unwrapping. This technique is demonstrated with phantom data at 1.5 T and with human data at 1.5, 3 and 7 T.     

An improved Richardson-Lucy algorithm based on local prior

Ringing is one of the most common disturbing artifacts in image deconvolution. With a totally known kernel, the standard Richardson-Lucy (RL) algorithm succeeds in many motion deblurring processes, but the resulting images still contain visible ringing. When the estimated kernel is different from the real one, the result of the standard RL iterative algorithm will be worse. To suppress the ringing artifacts caused by failures in the blur kernel estimation, this paper improves the RL algorithm based on the local prior. Firstly, the standard deviation of pixels in the local window is computed to find the smooth region and the image gradient in the region is constrained to make its distribution consistent with the deblurring image gradient. Secondly, in order to suppress the ringing near the edge of a rigid body in the image, a new mask was obtained by computing the sharp edge of the image produced using the first step. If the kernel is large-scale, where the foreground is rigid and the background is smoothing, this step could produce a significant inhibitory effect on ringing artifacts. Thirdly, the boundary constraint is strengthened if the boundary is relatively smooth. As a result of the steps above, high-quality deblurred images can be obtained even when the estimated kernels are not perfectly accurate. On the basis of blurred images and the related kernel information taken by the additional hardware, our approach proved to be effective.     

An improved image restoration approach using adaptive local constraint

How to restore images well with reducing ringing and noise remains difficult. An improved image restoration algorithm based on adaptive local constraint is proposed. We consider the statistical property that first order of derivatives of image obeying Gaussian distributions to control the noise. In order to suppress the artifacts around regions where gray values vary severely, spatial weighted matrix is adopted which can impose a local constraint on restoration and smoothing. According to experiments of simulation and real images, this approach can well restore edge regions well as well as reduce the noise in smooth area, which means that the approach does validly in suppressing ringing ripples and preserve more details.     

结合基于梯度的振铃评价算法的总变分最小化图像分块复原法

为了消除退化函数随空间变化发生变化模糊图像分块复原法子块之间的不平滑拼接缝,提出了一种结合了基于梯度的振铃评价算法梯度振铃评价(GRM)的总变分(TV)最小化分块复原法.根据图像分布及退化类型将模糊图像划分为矩形、环形或其他形状的子块,图像子块之间要留有一定的重叠区;然后对每一个图像子块进行复原,GRM方法是基于图像梯度结构相似度的图像质量评价算法,以GRM作为TV复原算法迭代过程中的收敛条件,可以更好地控制复原图像的振铃;最后去除复原图像子块含振铃波纹的重叠区,拼接得到完整图像.并以矩形分块及环形分块为例,证明该方法可以很好地抑制图像边界振铃效应,克服分块复原法本身的缺陷,得到拼接平滑的完整图像.     

可编程成像系统中莫尔条纹处理方法研究

使用对傅里叶变换频谱进行空间滤波的方法,根据“切口”滤波器在频谱突变位置出现振铃现象的缺点,设计了基于“切口”滤波器的多点高斯滤波器.实验表明:使用该方法可以有效降低莫尔条纹对系统采集到图像质量的影响,得到的重构图像清晰,基本无振铃现象,可以满足系统需要.使用该方法运算量小、处理简单,能够由嵌入式系统完成.     

An adaptive super-resolution method based on regional pixel information and ringing artifacts suppression

Multi-frame image super-resolution (SR) aims to utilize information from a set of low-resolution (LR) images to compose a high-resolution (HR) one. In this paper, a novel multi-frame image super-resolution algorithm is proposed based on regional pixel information and ringing artifacts suppression. Firstly, a new regularization term which adopts Regional Adaptive Weight Coefficients (RAWC) is produced to keep edges and flat regions. After detailed analysis, an iterative process is given for image reconstruction. Then an adaptive term according to the local variance of iterative correction image is designed to evaluate the ringing artifacts. Finally, the original iteration is updated by adding the restraint term for better visual effects and lower noise of reconstructive HR image. Thorough experimental results show the proposed algorithm is effective for SR reconstruction and ringing artifacts suppression.     

一种抑制合成孔径雷达图像相干斑的各向异性扩散滤波方法

本文提出了一种抑制合成孔径雷达图像乘性相干斑噪声的各向异性扩散滤波新方法.该方法将受自适应耦合函数控制的平均曲率运动嵌入到传统相干斑抑制的各向异性扩散方程中,形成了一种可有效抑制边缘区域相干斑与同质区块效应现象的各向异性扩散新方程,同时在新建的扩散方程中,引入了由改进Frost滤波与局部方向比率联合构建的一种带方向约束的新扩散函数,进一步削弱了块效应现象且明显改善了抑斑图像的边缘抖动扭曲问题.实验结果表明该方法在有效保护图像边缘的同时,能充分平滑同质区与边缘区域的相干斑,明显削弱块效应现象,有效改善抑斑图像边缘抖动扭曲问题,而抑斑图像无论视觉效果还是参数指标均比多种传统抑斑方法更具优势.     

An improved image deconvolution approach using local constraint

Conventional deblurring approaches such as the Richardson–Lucy (RL) algorithm will introduce strong noise and ringing artifacts, though the point spread function (PSF) is known. Since it is difficult to estimate an accurate PSF in real imaging system, the results of those algorithms will be worse. A spatial weight matrix (SWM) is adopted as local constraint, which is incorporated into image statistical prior to improve the RL approach. Experiments show that our approach can make a good balance between preserving image details and suppressing ringing artifacts and noise.     

Reduced aliasing artifacts using shaking projection k-space sampling trajectory

Radial imaging techniques, such as projection-reconstruction （PR）, are used in magnetic resonance imaging （MRI） for dynamic imaging, angiography, and short-T2 imaging. They are less sensitive to flow and motion artifacts, and support fast imaging with short echo times. However, aliasing and streaking artifacts are two main sources which degrade radial imaging quality. For a given fixed number of k-space projections, data distributions along radial and angular directions will influence the level of aliasing and streaking artifacts. Conventional radial k-space sampling trajectory introduces an aliasing artifact at the first principal ring of point spread function （PSF）. In this paper, a shaking projection （SP） k-space sampling trajectory was proposed to reduce aliasing artifacts in MR images. SP sampling trajectory shifts the projection alternately along the k-space center, which separates k-space data in the azimuthal direction. Simulations based on conventional and SP sampling trajectories were compared with the same number projections. A significant reduction of aliasing artifacts was observed using the SP sampling trajectory. These two trajectories were also compared with different sampling frequencies. ASP trajectory has the same aliasing character when using half sampling frequency （or half data） for reconstruction. SNR comparisons with different white noise levels show that these two trajectories have the same SNR character. In conclusion, the SP trajectory can reduce the aliasing artifact without decreasing SNR and also provide a way for undersampling recon- struction. Furthermore, this method can be applied to three-dimensional （3D） hybrid or spherical radial k-space sampling for a more efficient reduction of aliasing artifacts.     

Autocalibrated parallel imaging reconstruction with sampling pattern optimization for GRASE: APIR4GRASE

PurposeTo reduce artifacts and scan time of GRASE imaging by selecting an optimal sampling pattern and jointly reconstructing gradient echo and spin echo images.MethodsWe jointly reconstruct images for the different echo types by considering these as additional virtual coil channels in the novel Autocalibrated Parallel Imaging Reconstruction with Sampling Pattern Optimization for GRASE (APIR4GRASE) method. Besides image reconstruction, we identify optimal sampling patterns for the acquisition. The selected optimal patterns were validated on phantom and in-vivo acquisitions. Comparison to the conventional GRASE without acceleration, and to the GRAPPA reconstruction with a single echo type was also performed.ResultsUsing identified optimal sampling patterns, APIR4GRASE eliminated modulation artifacts in both phantom and in-vivo experiments; mean square error (MSE) was reduced by 78% and 94%, respectively, compared to the conventional GRASE with similar scan time. Both artifacts and g-factor were reduced compared to the GRAPPA reconstruction with a single echo type.ConclusionAPIR4GRASE substantially improves the speed and quality of GRASE imaging over the state-of-the-art, and is able to reconstruct both spin echo and gradient echo images.     

Coherent cross-polarization theory for a spin-12 coupled to a general object

This work describes a segmented radial turbo-spin-echo technique (DW-rTSE) for high-resolution multislice diffusion-weighted imaging and quantitative ADC mapping. Diffusion-weighted images with an in-plane resolution of 700 microm and almost free of bulk motion can be obtained in vivo without cardiac gating. However, eddy currents and pulsatile brain motion cause severe artifacts when strong diffusion weighting is applied. This work explains in detail the artifacts in projection reconstruction (PR) imaging arising from eddy currents and describes an effective eddy current compensation based on the adjustment of gradient timing. Application of the diffusion gradients in all three orthogonal directions is possible without degradation of the images due to eddy current artifacts, allowing studies of the diffusional anisotropy. Finally, a self-navigation approach is proposed to reduce residual nonrigid body motion artifacts. Five healthy volunteers were examined to show the feasibility of this method.     

A hybrid technique for spectroscopic imaging with reduced truncation artifact

Traditionally, Fourier spectroscopic imaging is associated with a small k-space coverage which leads to truncation artifacts such as "bleeding" and ringing in the resultant image. Because substantial truncation artifacts mainly arise from regions having intense signals, such as the subcutaneous lipid in the head, effective reduction of truncation artifacts can be achieved by obtaining an extended k-space coverage for these regions. In this paper, a hybrid technique which employs phase-encoded spectroscopic imaging (SI) to cover the central portion of the k-space and echo-planar spectroscopic imaging (EPSI) to measure the peripheral portion of the k-space is developed. EPSI, despite its inherently low SNR characteristics, provides a sufficient SNR for outer high-spatial frequency components of the aforementioned high signal regions and supplies an extended k-space coverage of these regions for the reduction of truncation artifacts. The data processing includes steps designed to remove inconsistency between the two types of data and a previously described technique for selectively retaining only outer k-space information for the high signal regions during the reconstruction. Experimental studies, in both phantoms and normal volunteers, demonstrate that the hybrid technique provides significant reduction in truncation artifacts.     

Correction of concomitant gradient artifacts in experimental microtesla MRI

Magnetic resonance imaging (MRI) suffers from artifacts caused by concomitant gradients when the product of the magnetic field gradient and the dimension of the sample becomes comparable to the static magnetic field. To investigate and correct for these artifacts at very low magnetic fields, we have acquired MR images of a 165-mm phantom in a 66-microT field using gradients up to 350 microT/m. We prepolarize the protons in a field of about 100 mT, apply a spin-echo pulse sequence, and detect the precessing spins using a superconducting gradiometer coupled to a superconducting quantum interference device (SQUID). Distortion and blurring are readily apparent at the edges of the images; by comparing the experimental images to computer simulations, we show that concomitant gradients cause these artifacts. We develop a non-perturbative, post-acquisition phase correction algorithm that eliminates the effects of concomitant gradients in both the simulated and the experimental images. This algorithm assumes that the switching time of the phase-encoding gradient is long compared to the spin precession period. In a second technique, we demonstrate that raising the precession field during phase encoding can also eliminate blurring caused by concomitant phase-encoding gradients; this technique enables one to correct concomitant gradient artifacts even when the detector has a restricted bandwidth that sets an upper limit on the precession frequency. In particular, the combination of phase correction and precession field cycling should allow one to add MRI capabilities to existing 300-channel SQUID systems used to detect neuronal currents in the brain because frequency encoding could be performed within the 1-2 kHz bandwidth of the readout system.     

Imaging of subcutaneous microcirculation vascular network by double correlation Optical Coherence Tomography

Spatial distribution of superficial blood vessels in human skin in vivo has been observed by using the double correlation Optical Coherence Tomography (OCT). To remove background noise, reduce the artifacts associated with patient motions and to increase the overall quality of the experimental OCT images an adaptive Wiener filtering technique has been employed. Fourier domain correlation has been subsequently applied to enhance spatial resolution of images of vascular network in human skin in vivo. Image processing has been performed on Graphics Processing Units (GPUs) utilizing Compute Unified Device Architecture (CUDA) framework in the frequency‐domain. This approach allows carrying out image processing in parallel significantly speeding up the computations. The presented results show that the double correlation method permits obtaining 2D/3D OCT images of subcutaneous microcirculation vascular network and its spatial distribution within the human skin with higher spatial resolution compare to the other OCT correlation‐based techniques developed earlier.     

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.     

基于指数律分布的图像快速盲复原算法

针对现有图像盲复原迭代算法多存在耗时较长和难以保证收敛性等问题,提出一种改进的快速算法.首先根据指数律重建原始图像的频谱,然后利用原始图像和降质图像的频谱关系,采用多方向综合估计方法得到点扩散函数.多方向综合估计方法可降低估计误差,增加算法的稳定性.最后利用得到的点扩散函数和维纳滤波法进行图像重建.与现有算法的对比实验结果表明,针对适合大量成像系统的G类点扩散函数,本算法可以得到更准确的点扩散函数估计,且降低了振铃效应的影响,取得更好的图像复原效果.     

减少光电成像系统周期目标莫尔赝像的双倍率法

光电成像系统对周期目标欠抽样成偈时,由于频谱混淆将产生莫尔条纹式的赝像.介绍了利用两帧不同样频率下的混淆图像以减少莫尔赝像恢复原周期目标的双倍率方法,并给出了计算模拟结果.