Real Magnetic Resonance Imaging Data Research Articles

Synthetic Brain Images: Bridging the Gap in Brain Mapping With Generative Adversarial Model

Magnetic Resonance Imaging (MRI) is a vital modality for gaining precise anatomical information, and it plays a significant role in medical imaging for diagnosis and therapy planning. Image synthesis problems have seen a revolution in recent years due to the introduction of deep learning techniques, specifically Generative Adversarial Networks (GANs). This work investigates the use of Deep Convolutional Generative Adversarial Networks (DCGAN) for producing high-fidelity and realistic MRI image slices. The suggested approach uses a dataset with a variety of brain MRI scans to train a DCGAN architecture. While the discriminator network discerns between created and real slices, the generator network learns to synthesise realistic MRI image slices. The generator refines its capacity to generate slices that closely mimic real MRI data through an adversarial training approach. The outcomes demonstrate that the DCGAN promise for a range of uses in medical imaging research, since they show that it can effectively produce MRI image slices if we train them for a consequent number of epochs. This work adds to the expanding corpus of research on the application of deep learning techniques for medical image synthesis. The slices that are could be produced possess the capability to enhance datasets, provide data augmentation in the training of deep learning models, as well as a number of functions are made available to make MRI data cleaning easier, and a three ready to use and clean dataset on the major anatomical plans.

Deep Kronecker network

Summary We develop a novel framework for the analysis of medical imaging data, including magnetic resonance imaging, functional magnetic resonance imaging, computed tomography and more. Medical imaging data differ from general images in two main aspects: (i) the sample size is often considerably smaller and (ii) the interpretation of the model is usually more crucial than predicting the outcome. As a result, standard methods such as convolutional neural networks cannot be directly applied to medical imaging analysis. Therefore, we propose the deep Kronecker network, which can adapt to the low sample size constraint and offer the desired model interpretation. Our approach is versatile, as it works for both matrix- and tensor-represented image data and can be applied to discrete and continuous outcomes. The deep Kronecker network is built upon a Kronecker product structure, which implicitly enforces a piecewise smooth property on coefficients. Moreover, our approach resembles a fully convolutional network as the Kronecker structure can be expressed in a convolutional form. Interestingly, our approach also has strong connections to the tensor regression framework proposed by Zhou et al. (2013), which imposes a canonical low-rank structure on tensor coefficients. We conduct both classification and regression analyses using real magnetic resonance imaging data from the Alzheimer’s Disease Neuroimaging Initiative to demonstrate the effectiveness of our approach.

Multimodality-based super-resolution reconstruction for routine brain magnetic resonance images

To propose a multi-modality-based super-resolution synthesis model for reconstruction of routine brain magnetic resonance images (MRI) with a low resolution and a high thickness into high-resolution images. Based on real paired low-high resolution MRI data (2D T1, 2D T2 FLAIR and 3D T1), a structure-constrained image mapping network was used to extract important features from the images with different modalities including the whole T1 and subcortical regions of T2 FLAIR to reconstruct T1 images with higher resolutions. The gray scale intensity and structural similarities between the super-resolution images and high-resolution images were used to enhance the reconstruction performance. We used the anatomical information acquired from segment maps of the super-resolution T1 image and the ground truth by a segmentation tool as a significant constraint for adaptive learning of the intrinsic tissue structure characteristics of the brain to improve the reconstruction performance of the model. Our method showed the performance on the testing dataset than other methods with an average PSNR of 33.11 and SSIM of 0.996. The anatomical structure of the brain including the sulcus, gyrus, and subcortex were all reconstructed clearly using the proposed method, which also greatly enhanced the precision of MSCSR for brain volume measurement. The proposed MSCSR model shows excellent performance for reconstructing super-resolution brain MR images based on the information of brain tissue structure and multimodality MR images.

Intensity inhomogeneity correction of MRI images using InhomoNet

Intensity inhomogeneity is one of the major artifacts in magnetic resonance imaging (MRI). Bias field present in MRI images alters true pixel value and produces spurious varying pixel intensities. This artifact affects the diagnosis by radiologists in a detrimental manner and also degrades the performance of computer-aided diagnosis algorithms such as segmentation. The present work proposes a novel network called InhomoNet for intensity inhomogeneity correction of MRI image. The generator architecture of InhomoNet consists of a new multi-scale local information module at each encoder block that helps to capture features at multiple scales. The horizontal and vertical kernels help to reduce the problems like loss of neighborhood information, gridding issues caused due to large dilated convolution operations. The attention-driven skip connections in the generator network are utilized to transfer optimal semantic and spatial localization information from the encoder to decoder blocks. Further, the present work proposes two new losses functions, i.e. histogram correlation and 3D pixel loss. These losses help to realize pixel consistency across different regions of brain MRI. The inculcation of the L1 loss provides guidance to the upsampling process as it compares the prediction from each decoder block with the ground truth. The proposed method is evaluated on simulated and real MRI data. The comparative analysis with popular state-of-the-art methods depicts the ability of the proposed method to perform intensity inhomogeneity correction accurately.

Gaussian process emulation to accelerate parameter estimation in a mechanical model of the left ventricle: a critical step towards clinical end-user relevance

In recent years, we have witnessed substantial advances in the mathematical modelling of the biomechanical processes underlying the dynamics of the cardiac soft-tissue. Gao et al. (Gao et al. 2017 J. R. Soc. Interface 14, 20170203 ( doi:10.1098/rsif.2017.0203 )) demonstrated that the parameters underlying the biomechanical model have diagnostic value for prognosticating the risk of myocardial infarction. However, the computational costs of parameter estimation are prohibitive when the goal lies in building real-time clinical decision support systems. This is due to the need to repeatedly solve the mathematical equations numerically using finite-element discretization during an iterative optimization routine. The present article presents a method for accelerating the inference of the constitutive parameters by using statistical emulation with Gaussian processes. We demonstrate how the computational costs can be reduced by about three orders of magnitude, with hardly any loss in accuracy, and we assess various alternative techniques in a comparative evaluation study based on simulated data obtained by solving the left ventricular model with the finite-element method, and real magnetic resonance images data for a human volunteer.

Rician noise and intensity nonuniformity correction (NNC) model for MRI data

Abstract Rician noise and intensity nonuniformity are two common artifacts and usually coexist in magnetic resonance imaging (MRI) data. Many methods have been proposed in the literature dealing with either Rician noise or intensity nonuniformity individually. We numerically verify that the existence of intensity nonuniformity may lead to the underestimation of noise, which means intensity nonuniformity influences the performance of denoising and vice versa. Thus, we propose a novel restoration model via a maximum a posteriori (MAP) estimator by regarding MRI data as a combination of two multiplicative components, namely, the true intensity and the bias field, and a noise followed a Rician distribution. We also guarantee that the proposed model has at least one positive nontrivial solution theoretically. An efficient algorithm based on alternating minimization method is developed, all subproblems of which can be solved effectively by either Newton's method or closed-form solutions. Intensive numerical results on synthetic and real MRI data confirm the robustness of the method and its better performance for MRI data restoration.

An Optimized Cascaded Stochastic Resonance for the Enhancement of Brain MRI

BackgroundThe contrast enhancement of Magnetic Resonance Imaging (MRI) data is quite challenging as the noise present in this data also get amplified in this process. Dynamic Stochastic Resonance (DSR) is the technique that utilizes the noise to enhance the contrast of MRI data. MethodThe present study proposes the cascaded stochastic resonance, which exploits the properties of modified potential neuron model and quartic bistable model of DSR. The Multi-objective Particle Swarm Optimization (MOPSO) tunes the DSR parameters associated with the cascading of both the models. The MOPSO produces a set of the solution called Pareto front for the maximization of two image quality measures, i.e., contrast enhancement factor and universal image quality index. Further, the maximization of another image quality measure, i.e., anisotropy helps to obtain the optimum enhanced image from the Pareto fronts solution. ResultsThe present study included the simulated and real MRI data. The results show that the proposed method obtained mean contrast enhancement factor, universal image quality index and anisotropy equal to 1.79, 0.78 and 0.021 respectively. These values are better than those obtained for classical bistable DSR and other conventional contrast enhancement techniques. The proposed algorithm has been tested on real MRI data as well and found valuable in the diagnosis of lacunar infarct and mesial temporal sclerosis. ConclusionThe cascaded DSR based on MOPSO has shown promising results and may be highly beneficial to the diagnosis of different brain pathology.

Tucker Tensor Regression and Neuroimaging Analysis.

Neuroimaging data often take the form of high dimensional arrays, also known as tensors. Addressing scientific questions arising from such data demands new regression models that take multidimensional arrays as covariates. Simply turning an image array into a vector would both cause extremely high dimensionality and destroy the inherent spatial structure of the array. In a recent work, Zhou et al. (2013) proposed a family of generalized linear tensor regression models based upon the CP (CANDECOMP/PARAFAC) decomposition of regression coefficient array. Low rank approximation brings the ultrahigh dimensionality to a manageable level and leads to efficient estimation. In this article, we propose a tensor regression model based on the more flexible Tucker decomposition. Compared to the CP model, Tucker regression model allows different number of factors along each mode. Such flexibility leads to several advantages that are particularly suited to neuroimaging analysis, including further reduction of the number of free parameters, accommodation of images with skewed dimensions, explicit modeling of interactions, and a principled way of image downsizing. We also compare the Tucker model with CP numerically on both simulated data and a real magnetic resonance imaging data, and demonstrate its effectiveness in finite sample performance.

3-D Reconstruction in Canonical Co-Ordinate Space From Arbitrarily Oriented 2-D Images

Limited capture range, and the requirement to provide high quality initialization for optimization-based 2-D/3-D image registration methods, can significantly degrade the performance of 3-D image reconstruction and motion compensation pipelines. Challenging clinical imaging scenarios, which contain significant subject motion, such as fetal in-utero imaging, complicate the 3-D image and volume reconstruction process. In this paper, we present a learning-based image registration method capable of predicting 3-D rigid transformations of arbitrarily oriented 2-D image slices, with respect to a learned canonical atlas co-ordinate system. Only image slice intensity information is used to perform registration and canonical alignment, no spatial transform initialization is required. To find image transformations, we utilize a convolutional neural network architecture to learn the regression function capable of mapping 2-D image slices to a 3-D canonical atlas space. We extensively evaluate the effectiveness of our approach quantitatively on simulated magnetic resonance imaging (MRI), fetal brain imagery with synthetic motion and further demonstrate qualitative results on real fetal MRI data where our method is integrated into a full reconstruction and motion compensation pipeline. Our learning based registration achieves an average spatial prediction error of 7 mm on simulated data and produces qualitatively improved reconstructions for heavily moving fetuses with gestational ages of approximately 20 weeks. Our model provides a general and computationally efficient solution to the 2-D/3-D registration initialization problem and is suitable for real-time scenarios.

PVR: Patch-to-Volume Reconstruction for Large Area Motion Correction of Fetal MRI.

In this paper, we present a novel method for the correction of motion artifacts that are present in fetal magnetic resonance imaging (MRI) scans of the whole uterus. Contrary to current slice-to-volume registration (SVR) methods, requiring an inflexible anatomical enclosure of a single investigated organ, the proposed patch-to-volume reconstruction (PVR) approach is able to reconstruct a large field of view of non-rigidly deforming structures. It relaxes rigid motion assumptions by introducing a specific amount of redundant information that is exploited with parallelized patchwise optimization, super-resolution, and automatic outlier rejection. We further describe and provide an efficient parallel implementation of PVR allowing its execution within reasonable time on commercially available graphics processing units, enabling its use in the clinical practice. We evaluate PVR’s computational overhead compared with standard methods and observe improved reconstruction accuracy in the presence of affine motion artifacts compared with conventional SVR in synthetic experiments. Furthermore, we have evaluated our method qualitatively and quantitatively on real fetal MRI data subject to maternal breathing and sudden fetal movements. We evaluate peak-signal-to-noise ratio, structural similarity index, and cross correlation with respect to the originally acquired data and provide a method for visual inspection of reconstruction uncertainty. We further evaluate the distance error for selected anatomical landmarks in the fetal head, as well as calculating the mean and maximum displacements resulting from automatic non-rigid registration to a motion-free ground truth image. These experiments demonstrate a successful application of PVR motion compensation to the whole fetal body, uterus, and placenta.

Deducing magnetic resonance neuroimages based on knowledge from samples

PurposeBecause individual variance always exists, using the same set of predetermined parameters for magnetic resonance imaging (MRI) may not be exactly suitable for each participant. We propose a knowledge-based method that can repair MRI data of undesired contrast as if a new scan were acquired using imaging parameters that had been individually optimized. MethodsThe method employed a strategy called analogical reasoning to deduce voxel-wise relaxation properties using morphological and biological similarity. The proposed framework involves steps of intensity normalization, tissue segmentation, relaxation time deducing, and image deducing. ResultsThis approach has been preliminarily validated using conventional MRI data at 3T from several examples, including 5 normal and 9 clinical datasets. It can effectively improve the contrast of real MRI data by deducing imaging data using optimized imaging parameters based on deduced relaxation properties. The statistics of deduced images shows a high correlation with real data that were actually collected using the same set of imaging parameters. ConclusionThe proposed method of deducing MRI data using knowledge of relaxation times alternatively provides a way of repairing MRI data of less optimal contrast. The method is also capable of optimizing an MRI protocol for individual participants, thereby realizing personalized MR imaging.

Active Localization and Tracking of Needle and Target in Robotic Image-Guided Intervention Systems.

This paper describes a framework of algorithms for the active localization and tracking of flexible needles and targets during image-guided percutaneous interventions. The needle and target configurations are tracked by Bayesian filters employing models of the needle and target motions and measurements of the current system state obtained from an intra-operative imaging system which is controlled by an entropy-minimizing active localization algorithm. Versions of the system were built using particle and unscented Kalman filters and their performance was measured using both simulations and hardware experiments with real magnetic resonance imaging data of needle insertions into gel phantoms. Performance of the localization algorithms is given in terms of accuracy of the predictions and computational efficiency is discussed.

Combining interpolation and 3D level set method (I+3DLSM) for medical image segmentation

A combined interpolation - 3D Level Set Method (I+3DLSM) based segmentation process is presented. The performance in terms of accuracy of the 3-dimensional (3D) level set method (LSM) in the segmentation of throat regions from highly anisotropic magnetic resonance imaging (MRI) volumes, with and without an interpolation step is evaluated. Qualitative and quantitative results from real MRI data suggest that performing interpolation, to reconstruct isotropic MRI volumes, prior to 3D LSM improves the accuracy of the segmentation results, compared to interpolation post 3D LSM and no interpolation at all.

A two-step optimization approach for nonlocal total variation-based Rician noise reduction in magnetic resonance images.

Magnetic resonance imaging (MRI) often suffers from apparent noise during image acquisition and transmission. The degraded data can easily result in nonrobust postprocessing steps in medical image analysis. The purpose of this study is to eliminate noise effects and improve image quality using a nonlocal feature-preserving denoising method. From a statistical point of view, the magnitude MR images in the presence of noise are usually modeled using a Rician distribution. In the maximum a posteriori framework, a nonlocal total variation (NLTV)-based feature-preserving MRI Rician denoising model is proposed by taking full advantage of high degree of selfsimilarity and redundancy within MR images. However, the nonconvex data-fidelity term and nonsmooth NLTV regularizer make the denoising problem difficult to solve. To guarantee solution stability, a piecewise convex function is first introduced to approximate the nonconvex version. In what follows, a two-step optimization approach is developed to solve the resulting convex denoising model. In each step of this approach, the subproblem can be efficiently solved using existing optimization algorithms. The method performance is evaluated using synthetic and clinical MRI data sets as well as one diffusion tensor MRI (DTI) data set. Extensive experiments are conducted to compare the proposed method with several state-of-the-art denoising methods. For the synthetic and clinical MRI data sets, the proposed method considerably outperformed other competing denoising methods in terms of both quantitative and visual quality evaluations. It was capable of effectively removing noise in MR images and enhancing tissue characterization. The advantage of the proposed method became more significant as the noise level increased. For the DTI data set, compared with other denoising methods, the proposed method not only preserved the apparent diffusion coefficient but also generated more regular fractional anisotropy (FA) and color-coded FA without obvious visual artifacts. This study describes and validates a nonlocal feature-preserving method for Rician noise reduction on synthetic and real MRI data sets. By exploiting the feature-preserving capability of NLTV regularizer, the proposed method maintains a good balance between noise reduction and fine detail preservation. The experiments have demonstrated a huge potential of the proposed method for routine clinical practice.

3-dimensional throat region segmentation from MRI data based on Fourier interpolation and 3-dimensional level set methods.

A new algorithm for 3D throat region segmentation from magnetic resonance imaging (MRI) is presented. The proposed algorithm initially pre-processes the MRI data to increase the contrast between the throat region and its surrounding tissues and to reduce artifacts. Isotropic 3D volume is reconstructed using the Fourier interpolation. Furthermore, a cube encompassing the throat region is evolved using level set method to form a smooth 3D boundary of the throat region. The results of the proposed algorithm on real and synthetic MRI data are used to validate the robustness and accuracy of the algorithm.

Anisotropic fractional diffusion tensor imaging

Traditional diffusion tensor imaging (DTI) maps brain structure by fitting a diffusion model to the magnitude of the electrical signal acquired in magnetic resonance imaging (MRI). Fractional DTI employs anomalous diffusion models to obtain a better fit to real MRI data, which can exhibit anomalous diffusion in both time and space. In this paper, we describe the challenge of developing and employing anisotropic fractional diffusion models for DTI. Since anisotropy is clearly present in the three-dimensional MRI signal response, such models hold great promise for improving brain imaging. We then propose some candidate models, based on stochastic theory.

Hybrid intensity- and phase-based optical flow tracking of tagged MRI.

Accurate tracking of the myocardium tissues in tagged Magnetic Resonance Images (MRI) is essential for evaluating the cardiac function. Current tracking methods utilize either the image intensity or the image phase as landmarks that can be tracked. In either case, the performance is vulnerable to the image quality and the fading of the tag lines. In this work, we propose a hybrid optical flow tracking method that combines both the intensity and the phase features of the image. The method is validated using numerical cardiac phantom as well as real MRI data experiments. Both experiments showed that the proposed method outperforms current intensity-based optical flow tracking and the phase-based HARP method with maximum error of 1 pixel at extreme conditions of tag fading.

Higher-order principal component pursuit via tensor approximation and convex optimization

Recovering the low-rank structure of data matrix from sparse errors arises in the principal component pursuit (PCP). This paper exploits the higher-order generalization of matrix recovery, named higher-order principal component pursuit (HOPCP), since it is critical in multi-way data analysis. Unlike the convexification (nuclear norm) for matrix rank function, the tensorial nuclear norm is still an open problem. While existing preliminary works on the tensor completion field provide a viable way to indicate the low complexity estimate of tensor, therefore, the paper focuses on the low multi-linear rank tensor and adopt its convex relaxation to formulate the convex optimization model of HOPCP. The paper further propose two algorithms for HOPCP based on alternative minimization scheme: the augmented Lagrangian alternating direction method (ALADM) and its truncated higher-order singular value decomposition (ALADM-THOSVD) version. The former can obtain a high accuracy solution while the latter is more efficient to handle the computationally intractable problems. Experimental results on both synthetic data and real magnetic resonance imaging data show the applicability of our algorithms in high-dimensional tensor data processing.

Characterization of Respiratory and Cardiac Motion from Electro-Anatomical Mapping Data for Improved Fusion of MRI to Left Ventricular Electrograms

Accurate fusion of late gadolinium enhancement magnetic resonance imaging (MRI) and electro-anatomical voltage mapping (EAM) is required to evaluate the potential of MRI to identify the substrate of ventricular tachycardia. However, both datasets are not acquired at the same cardiac phase and EAM data is corrupted with respiratory motion limiting the accuracy of current rigid fusion techniques. Knowledge of cardiac and respiratory motion during EAM is thus required to enhance the fusion process. In this study, we propose a novel approach to characterize both cardiac and respiratory motion from EAM data using the temporal evolution of the 3D catheter location recorded from clinical EAM systems. Cardiac and respiratory motion components are extracted from the recorded catheter location using multi-band filters. Filters are calibrated for each EAM point using estimates of heart rate and respiratory rate. The method was first evaluated in numerical simulations using 3D models of cardiac and respiratory motions of the heart generated from real time MRI data acquired in 5 healthy subjects. An accuracy of 0.6–0.7 mm was found for both cardiac and respiratory motion estimates in numerical simulations. Cardiac and respiratory motions were then characterized in 27 patients who underwent LV mapping for treatment of ventricular tachycardia. Mean maximum amplitude of cardiac and respiratory motion was 10.2±2.7 mm (min = 5.5, max = 16.9) and 8.8±2.3 mm (min = 4.3, max = 14.8), respectively. 3D Cardiac and respiratory motions could be estimated from the recorded catheter location and the method does not rely on additional imaging modality such as X-ray fluoroscopy and can be used in conventional electrophysiology laboratory setting.

Nonlocal denoising using anisotropic structure tensor for 3D MRI

Noise in magnetic resonance imaging (MRI) data is widely recognized to be harmful to image processing and subsequent quantitative analysis. To ameliorate the effects of image noise, the authors present a structure-tensor based nonlocal mean (NLM) denoising technique that can effectively reduce noise in MRI data and improve tissue characterization. The proposed 3D NLM algorithm uses a structure tensor to characterize information around tissue boundaries. The similarity weight of a pixel (or patch), which determines its contribution to the denoising process, is determined by the intensity and structure tensor simultaneously. Meanwhile, similarity of structure tensors is computed using an affine-invariant Riemannian metrics, which compares tensor properties more comprehensively and avoids orientation inaccuracy of structure subsequently. The proposed method is further extended for denoising high dimensional MRI data such as diffusion weighted MRI. It is also extended to handle Rician noise corruption so that denoising effects are further enhanced. The proposed method was implemented in both simulated datasets and multiply modalities of real 3D MRI datasets. Comparisons with related state-of-the-art algorithms demonstrated that this method improves denoising performance qualitatively and quantitatively. In this paper, high order structure information of 3D MRI was characterized by 3D structure tensor and compared for NLM denoising in a Riemannian space. Experiments with simulated and real human MRI data demonstrate a great potential of the proposed technique for routine clinical use.