An anisotropic diffusion method for denoising dynamic susceptibility contrast-enhanced magnetic resonance images

A Survey on State-of-the-art Denoising Techniques for Brain Magnetic Resonance Images

The accuracy of the magnetic resonance (MR) image diagnosis depends on the quality of the image, which degrades mainly due to noise and artifacts. The noise is introduced because of erroneous imaging environment or distortion in the transmission system. Therefore, denoising methods play an important role in enhancing the image quality. However, a tradeoff between denoising and preserving the structural details is required. Most of the existing surveys are conducted on a specific MR image modality or on limited denoising schemes. In this context, a comprehensive review on different MR image denoising techniques is inevitable. This survey suggests a new direction in categorizing the MR image denoising techniques. The categorization of the different image models used in medical image processing serves as the basis of our classification. This study includes recent improvements on deep learning-based denoising methods alongwith important traditional MR image denoising methods. The major challenges and their scope of improvement are also discussed. Further, many more evaluation indices are considered for a fair comparison. An elaborate discussion on selecting appropriate method and evaluation metric as per the kind of data is presented. This study may encourage researchers for further work in this domain.

New image-restoration method using a simultaneous algebraic reconstruction technique: comparison with the Richardson-Lucy algorithm

We developed a new image-restoration method that incorporates the point spread function (PSF) into the simultaneous algebraic reconstruction technique (SART-PSF). Additionally, through simulation studies, we investigated the usefulness of the method in comparison with the Richardson-Lucy (RL) algorithm. In the simulation studies, degraded images were generated by convolving magnetic resonance imaging-based brain images with PSF and adding Gaussian or Poisson noise to them to simulate various noise levels. The effects of the number of iterations N, noise, and PSF error on the processed images were quantitatively evaluated using the percent root mean square error (PRMSE) and mean structural similarity index (mSSIM). After applying the SART-PSF to images degraded using Gaussian noise, the PRMSE value and increase thereof, when N was increased, were smaller than those when using the RL algorithm. The mSSIM value was higher and its decrease upon increasing N was smaller than that of the RL algorithm. When Poisson noise was assumed, the differences in PRMSE and mSSIM between both methods were smaller than those when Gaussian noise was assumed. When the PSF error was negative, its effect on PRMSE and mSSIM was similar for both methods. However, when it was positive, the deterioration of these parameters for the SART-PSF was less than that for the RL algorithm in both the Gaussian and Poisson noise cases. The results suggest that the SART-PSF is more robust against noise and a PSF error than the RL algorithm and, thus, can be used as an alternative to the RL algorithm.

Laplacian Eigenmaps Network-Based Nonlocal Means Method for MR Image Denoising

Magnetic resonance (MR) images are often corrupted by Rician noise which degrades the accuracy of image-based diagnosis tasks. The nonlocal means (NLM) method is a representative filter in denoising MR images due to its competitive denoising performance. However, the existing NLM methods usually exploit the gray-level information or hand-crafted features to evaluate the similarity between image patches, which is disadvantageous for preserving the image details while smoothing out noise. In this paper, an improved nonlocal means method is proposed for removing Rician noise in MR images by using the refined similarity measures. The proposed method firstly extracts the intrinsic features from the pre-denoised image using a shallow convolutional neural network named Laplacian eigenmaps network (LEPNet). Then, the extracted features are used for computing the similarity in the NLM method to produce the denoised image. Finally, the method noise of the denoised image is utilized to further improve the denoising performance. Specifically, the LEPNet model is composed of two cascaded convolutional layers and a nonlinear output layer, in which the Laplacian eigenmaps are employed to learn the filter bank in the convolutional layers and the Leaky Rectified Linear Unit activation function is used in the final output layer to output the nonlinear features. Due to the advantage of LEPNet in recovering the geometric structure of the manifold in the low-dimension space, the features extracted by this network can facilitate characterizing the self-similarity better than the existing NLM methods. Experiments have been performed on the BrainWeb phantom and the real images. Experimental results demonstrate that among several compared denoising methods, the proposed method can provide more effective noise removal and better details preservation in terms of human vision and such objective indexes as peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM).

High-SNR multiple T2 (*)-contrast magnetic resonance imaging using a robust denoising method based on tissue characteristics

PURPOSE

To develop an effective method that can suppress noise in successive multiecho T₂ (*)-weighted magnetic resonance (MR) brain images while preventing filtering artifacts.

MATERIALS AND METHODS

For the simulation experiments, we used multiple T₂ -weighted images of an anatomical brain phantom. For in vivo experiments, successive multiecho MR brain images were acquired from five healthy subjects using a multiecho gradient-recalled-echo (MGRE) sequence with a 3T MRI system. Our denoising method is a nonlinear filter whose filtering weights are determined by tissue characteristics among pixels. The similarity of the tissue characteristics is measured based on the l₂ -difference between two temporal decay signals. Both numerical and subjective evaluations were performed in order to compare the effectiveness of our denoising method with those of conventional filters, including Gaussian low-pass filter (LPF), anisotropic diffusion filter (ADF), and bilateral filter. Root-mean-square error (RMSE), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were used in the numerical evaluation. Five observers, including one radiologist, assessed the image quality and rated subjective scores in the subjective evaluation.

RESULTS

Our denoising method significantly improves RMSE, SNR, and CNR of numerical phantom images, and CNR of in vivo brain images in comparison with conventional filters (P < 0.005). It also receives the highest scores for structure conspicuity (8.2 to 9.4 out of 10) and naturalness (9.2 to 9.8 out of 10) among the conventional filters in the subjective evaluation.

CONCLUSION

This study demonstrates that high-SNR multiple T₂ (*)-contrast MR images can be obtained using our denoising method based on tissue characteristics without noticeable artifacts. Evidence level: 2 J. MAGN. RESON. IMAGING 2017;45:1835-1845.

A wavelet multiscale denoising algorithm for magnetic resonance (MR) images

Based on the Radon transform, a wavelet multiscale denoising method is proposed for MR images. The approach explicitly accounts for the Rician nature of MR data. Based on noise statistics we apply the Radon transform to the original MR images and use the Gaussian noise model to process the MR sinogram image. A translation invariant wavelet transform is employed to decompose the MR 'sinogram' into multiscales in order to effectively denoise the images. Based on the nature of Rician noise we estimate noise variance in different scales. For the final denoised sinogram we apply the inverse Radon transform in order to reconstruct the original MR images. Phantom, simulation brain MR images, and human brain MR images were used to validate our method. The experiment results show the superiority of the proposed scheme over the traditional methods. Our method can reduce Rician noise while preserving the key image details and features. The wavelet denoising method can have wide applications in MRI as well as other imaging modalities.

Denoising of dynamic contrast-enhanced MR images using dynamic nonlocal means

This paper presents a new algorithm for denoising dynamic contrast-enhanced (DCE) MR images. It is a novel variation on the nonlocal means (NLM) algorithm. The algorithm, called dynamic nonlocal means (DNLM), exploits the redundancy of information in the temporal sequence of images. Empirical evaluations of the performance of the DNLM algorithm relative to seven other denoising methods-simple Gaussian filtering, the original NLM algorithm, a trivial extension of NLM to include the temporal dimension, bilateral filtering, anisotropic diffusion filtering, wavelet adaptive multiscale products threshold, and traditional wavelet thresholding-are presented. The evaluations include quantitative evaluations using simulated data and real data (20 DCE-MRI data sets from routine clinical breast MRI examinations) as well as qualitative evaluations using the same real data (24 observers: 14 image/signal-processing specialists, 10 clinical breast MRI radiographers). The results of the quantitative evaluation using the simulated data show that the DNLM algorithm consistently yields the smallest MSE between the denoised image and its corresponding original noiseless version. The results of the quantitative evaluation using the real data provide evidence, at the alpha = 0.05 level of significance, that the DNLM algorithm yields the smallest MSE between the denoised image and its corresponding original noiseless version. The results of the qualitative evaluation provide evidence, at the alpha = 0.05 level of significance, that the DNLM algorithm performs visually better than all of the other algorithms. Collectively the qualitative and quantitative results suggest that the DNLM algorithm more effectively attenuates noise in DCE MR images than any of the other algorithms.

A new denoising method for dynamic contrast-enhanced MRI

This paper presents a new algorithm for denoising dynamic contrast-enhanced (DCE) MR images. The algorithm is called Dynamic Non-Local Means and is a novel variation on the Non-Local Means (NL-Means) algorithm. It exploits the redundancy of information in the DCE-MRI sequence of images. An evaluation of the performance of the algorithm relative to six other denoising algorithms-Gaussian filtering, the original NL-Means algorithm, bilateral filtering, anisotropic diffusion filtering, the wavelets adaptive multiscale products threshold method, and the traditional wavelet thresholding method-is also presented. The evaluation was performed by two groups of expert observers-18 signal/image processing experts, and 9 clinicians (8 radiographers and 1 radiologist)-using real DCE-MRI data. The results of the evaluation provide evidence, at the alpha=0.05 level of significance, that both groups of observers deem the DNLM algorithm to perform visually better than all of the other algorithms.

Error analysis of tumor blood flow measurement using dynamic contrast-enhanced data and model-independent deconvolution analysis

We performed error analysis of tumor blood flow (TBF) measurement using dynamic contrast-enhanced data and model-independent deconvolution analysis, based on computer simulations. For analysis, we generated a time-dependent concentration of the contrast agent in the volume of interest (VOI) from the arterial input function (AIF) consisting of gamma-variate functions using an adiabatic approximation to the tissue homogeneity model under various plasma flow (F(p)), mean capillary transit time (T(c)), permeability-surface area product (PS) and signal-to-noise ratio (SNR) values. Deconvolution analyses based on truncated singular value decomposition with a fixed threshold value (TSVD-F), with an adaptive threshold value (TSVD-A) and with the threshold value determined by generalized cross validation (TSVD-G) were used to estimate F(p) values from the simulated concentration-time curves in the VOI and AIF. First, we investigated the relationship between the optimal threshold value and SNR in TSVD-F, and then derived the equation describing the relationship between the threshold value and SNR for TSVD-A. Second, we investigated the dependences of the estimated F(p) values on T(c), PS, the total duration for data acquisition and the shape of AIF. Although TSVD-F with a threshold value of 0.025, TSVD-A with the threshold value determined by the equation derived in this study and TSVD-G could estimate the F(p) values in a similar manner, the standard deviation of the estimates was the smallest and largest for TSVD-A and TSVD-G, respectively. PS did not largely affect the estimates, while T(c) did in all methods. Increasing the total duration significantly improved the variations in the estimates in all methods. TSVD-G was most sensitive to the shape of AIF, especially when the total duration was short. In conclusion, this study will be useful for understanding the reliability and limitation of model-independent deconvolution analysis when applied to TBF measurement using an extravascular contrast agent.

The impact of partial-volume effects in dynamic susceptibility contrast magnetic resonance perfusion imaging

PURPOSE

To demonstrate the degree of the cerebral blood flow (CBF) estimation bias that could arise from distortion of the arterial input function (AIF) as a result of partial-volume effects (PVEs) in dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI).

MATERIALS AND METHODS

A model of the volume fraction an artery occupies in a voxel was devised, and a mathematical relationship between the amount of PVE and the measured baseline MR signal intensity was derived. Based on this model, simulation studies were performed to assess the impact of PVE on CBF. Furthermore, the effectiveness of linear PVE compensation approaches on the concentration function was investigated.

RESULTS

Simulation results showed a nonlinear relationship between PVE and the resulting CBF measurement error. In addition to AIF underestimation, PVE also causes distortions of AIF frequency characteristics, leading to CBF errors varying with mean transit time (MTT). An uncorrected AIF measured at a voxel with a partial-volume fraction of <or=50% could produce a CBF overestimation of more than fourfold. Linear compensation of the concentration curves did not produce correct CBF estimates.

CONCLUSION

PVE can induce significant CBF estimation biases. In addition, the MTT dependence of CBF accuracy raises doubts of the validity of adopting a single cross-calibration factor (i.e., setting normal white matter to 22 mL minute(-1) (100 g)(-1)) to obtain CBF values with absolute units. The impact of PVE may be reduced by decreasing the maximum arterial signal drop in the perfusion images. To correct the AIF distortions introduced by PVE, the nonlinear relationship between the impact of PVE on MR signal intensity and contrast concentration function must be considered.

Deconvolution Analysis of Dynamic Contrast-Enhanced Data Based on Singular Value Decomposition Optimized by Generalized Cross Validation

PURPOSE

To present an implementation of generalized cross validation (GCV) for automatically determining the regularization parameter--i.e., the threshold value in deconvolution analysis based on truncated singular value decomposition (TSVD) of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) data--and to investigate the usefulness of this approach in comparison with TSVD with a fixed threshold value (TSVD-F).

METHODS

Using computer simulations, we generated a time-dependent concentration of the contrast agent in the volume of interest (VOI) from the arterial input function (AIF) modeled as a gamma-variate function under various cerebral blood flows (CBFs), cerebral blood volumes (CBVs), and signal-to-noise ratios (SNRs) for three different types of residue functions (exponential, triangular, and box-shaped). We also considered the effects of delay and dispersion in AIF. The TSVD with GCV (TSVD-G) and TSVD-F with a fixed threshold value of 0.2 were used to estimate CBF values from the simulated concentration-time curves in the VOI and AIF, and the estimated values were compared with the assumed values. Additionally, the optimal threshold value was determined from the threshold value in TSVD-F giving the mean CBF value closest to the assumed value and was compared with the threshold value determined with TSVD-G.

RESULTS

With TSVD-G, the CBF estimation was substantially improved over a wide range of CBFs for all types of residue functions at the cost of more noise than was seen with TSVD-F. The dependency of the threshold value determined with TSVD-G on the CBF, CBV, and SNR was similar to that of the optimal threshold value, with some discrepancy being observed for the box-shaped residue function, although they did not always agree in terms of absolute value.

CONCLUSION

Given an improved SNR, TSVD-G is useful for quantification of CBF with deconvolution analysis of DCE-MRI data.

Diffusion weighted magnetic resonance imaging in stroke

Diffusion weighted magnetic resonance imaging has evolved from an esoteric laboratory experiment to a critical aspect of routine clinical care of the patient presenting with symptoms suspicious of acute ischemic stroke. The purpose of this article is to review the basis of diffusion weighted imaging (DWI), to consider its application in acute stroke and to recognize potential pitfalls and stroke mimics that might be encountered. Included in the discussion are comments on the elimination of 'T2 shine through' phenomena as well as construction of pixel-by-pixel maps of the apparent diffusion coefficient (ADC). Furthermore, discussion of techniques such as parallel imaging (using SENSE) and PROPELLER sequences will be introduced as methods potentially allowing DWI to be utilized in areas usually associated with prohibitive susceptibility artifact (e.g. the base of the brain). Finally, the concept of interventional therapeutic approaches to salvaging ischemic tissue is introduced, both in terms of the ischemic penumbra (defined by a diffusion/perfusion mismatch) and also in terms of the potential reversibility of the diffusion-weighted hyperintensity, associated with the lesion core.