High-Resolution Self-Gated Dynamic Abdominal MRI Using Manifold Alignment

Embedding-Alignment Fusion-Based Graph Convolution Network With Mixed Learning Strategy for 4D Medical Image Reconstruction

In recent years, 4D medical image involving structural and motion information of tissue has attracted increasing attention. The key to the 4D image reconstruction is to stack the 2D slices based on matching the aligned motion states. In this study, the distribution of the 2D slices with the different motion states is modeled as a manifold graph, and the reconstruction is turned to be the graph alignment. An embedding-alignment fusion-based graph convolution network (GCN) with a mixed-learning strategy is proposed to align the graphs. Herein, the embedding and alignment processes of graphs interact with each other to realize a precise alignment with retaining the manifold distribution. The mixed strategy of self- and semi-supervised learning makes the alignment sparse to avoid the mismatching caused by outliers in the graph. In the experiment, the proposed 4D reconstruction approach is validated on the different modalities including Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and Ultrasound (US). We evaluate the reconstruction accuracy and compare it with those of state-of-the-art methods. The experiment results demonstrate that our approach can reconstruct a more accurate 4D image.

Clinical feasibility of deep learning reconstruction in liver diffusion-weighted imaging: Improvement of image quality and impact on apparent diffusion coefficient value

PURPOSE

Diffusion-weighted imaging (DWI) of the liver suffers from low resolution, noise, and artifacts. This study aimed to investigate the effect of deep learning reconstruction (DLR) on image quality and apparent diffusion coefficient (ADC) quantification of liver DWI at 3 Tesla.

METHOD

In this prospective study, images of the liver obtained at DWI with b-values of 0 (DWI0), 50 (DWI50) and 800 s/mm2 (DWI800) from consecutive patients with liver lesions from February 2022 to February 2023 were reconstructed with and without DLR (non-DLR). Image quality was assessed qualitatively using Likert scoring system and quantitatively using signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and liver/parenchyma boundary sharpness from region-of-interest (ROI) analysis. ADC value of lesion were measured. Phantom experiment was also performed to investigate the factors that determine the effect of DLR on ADC value. Qualitative score, SNR, CNR, boundary sharpness, and apparent diffusion coefficients (ADCs) for DWI were compared using paired t-test and Wilcoxon signed rank test. P < 0.05 was considered statistically significant.

RESULTS

A total of 85 patients with 170 lesions were included. DLR group showed a higher qualitative score than the non-DLR group. for example, with DWI800 the score was 4.77 ± 0.52 versus 4.30 ± 0.63 (P < 0.001). DLR group also showed higher SNRs, CNRs and boundary sharpness than the non-DLR group. DLR reduced the ADC of malignant tumors (1.105[0.904, 1.340] versus 1.114[0.904, 1.320]) (P < 0.001), but there was no significant difference in the diagnostic value of malignancy for DLR and non-DLR groups (P = 57.3). The phantom study confirmed a reduction of ADC in images with low resolution, and a stronger reduction of ADC in heterogeneous structures than in homogeneous ones (P < 0.001).

CONCLUSIONS

DLR improved image quality of liver DWI. DLR reduced the ADC value of lesions, but did not affect the diagnostic performance of ADC in distinguishing malignant tumors on a 3.0-T MRI system.

Manifold Learning via Linear Tangent Space Alignment (LTSA) for Accelerated Dynamic MRI With Sparse Sampling

The spatial resolution and temporal frame-rate of dynamic magnetic resonance imaging (MRI) can be improved by reconstructing images from sparsely sampled k -space data with mathematical modeling of the underlying spatiotemporal signals. These models include sparsity models, linear subspace models, and non-linear manifold models. This work presents a novel linear tangent space alignment (LTSA) model-based framework that exploits the intrinsic low-dimensional manifold structure of dynamic images for accelerated dynamic MRI. The performance of the proposed method was evaluated and compared to state-of-the-art methods using numerical simulation studies as well as 2D and 3D in vivo cardiac imaging experiments. The proposed method achieved the best performance in image reconstruction among all the compared methods. The proposed method could prove useful for accelerating many MRI applications, including dynamic MRI, multi-parametric MRI, and MR spectroscopic imaging.

Variational Manifold Learning From Incomplete Data: Application to Multislice Dynamic MRI

Current deep learning-based manifold learning algorithms such as the variational autoencoder (VAE) require fully sampled data to learn the probability density of real-world datasets. However, fully sampled data is often unavailable in a variety of problems, including the recovery of dynamic and high-resolution magnetic resonance imaging (MRI). We introduce a novel variational approach to learn a manifold from undersampled data. The VAE uses a decoder fed by latent vectors, drawn from a conditional density estimated from the fully sampled images using an encoder. Since fully sampled images are not available in our setting, we approximate the conditional density of the latent vectors by a parametric model whose parameters are estimated from the undersampled measurements using back-propagation. We use the framework for the joint alignment and recovery of multi-slice free breathing and ungated cardiac MRI data from highly undersampled measurements. Experimental results demonstrate the utility of the proposed scheme in dynamic imaging alignment and reconstructions.

Manifold-based respiratory phase estimation enables motion and distortion correction of free-breathing cardiac diffusion tensor MRI

PURPOSE

For in vivo cardiac DTI, breathing motion and B₀ field inhomogeneities produce misalignment and geometric distortion in diffusion-weighted (DW) images acquired with conventional single-shot EPI. We propose using a dimensionality reduction method to retrospectively estimate the respiratory phase of DW images and facilitate both distortion correction (DisCo) and motion compensation.

METHODS

Free-breathing electrocardiogram-triggered whole left-ventricular cardiac DTI using a second-order motion-compensated spin echo EPI sequence and alternating directionality of phase encoding blips was performed on 11 healthy volunteers. The respiratory phase of each DW image was estimated after projecting the DW images into a 2D space with Laplacian eigenmaps. DisCo and motion compensation were applied to the respiratory sorted DW images. The results were compared against conventional breath-held T₂ half-Fourier single shot turbo spin echo. Cardiac DTI parameters including fractional anisotropy, mean diffusivity, and helix angle transmurality were compared with and without DisCo.

RESULTS

The left-ventricular geometries after DisCo and motion compensation resulted in significantly improved alignment of DW images with T₂ reference. DisCo reduced the distance between the left-ventricular contours by 13.2% ± 19.2%, P < .05 (2.0 ± 0.4 for DisCo and 2.4 ± 0.5 mm for uncorrected). DisCo DTI parameter maps yielded no significant differences (mean diffusivity: 1.55 ± 0.13 × 10^-3 mm² /s and 1.53 ± 0.13 × 10^-3 mm² /s, P = .09; fractional anisotropy: 0.375 ± 0.041 and 0.379 ± 0.045, P = .11; helix angle transmurality: 1.00% ± 0.10°/% and 0.99% ± 0.12°/%, P = .44), although the orientation of individual tensors differed.

CONCLUSION

Retrospective respiratory phase estimation with LE-based DisCo and motion compensation in free-breathing cardiac DTI resulting in significantly reduced geometric distortion and improved alignment within and across slices.

Deep Learning-Based ECG-Free Cardiac Navigation for Multi-Dimensional and Motion-Resolved Continuous Magnetic Resonance Imaging

For the clinical assessment of cardiac vitality, time-continuous tomographic imaging of the heart is used. To further detect e.g., pathological tissue, multiple imaging contrasts enable a thorough diagnosis using magnetic resonance imaging (MRI). For this purpose, time-continous and multi-contrast imaging protocols were proposed. The acquired signals are binned using navigation approaches for a motion-resolved reconstruction. Mostly, external sensors such as electrocardiograms (ECG) are used for navigation, leading to additional workflow efforts. Recent sensor-free approaches are based on pipelines requiring prior knowledge, e.g., typical heart rates. We present a sensor-free, deep learning-based navigation that diminishes the need for manual feature engineering or the necessity of prior knowledge compared to previous works. A classifier is trained to estimate the R-wave timepoints in the scan directly from the imaging data. Our approach is evaluated on 3-D protocols for continuous cardiac MRI, acquired in-vivo and free-breathing with single or multiple imaging contrasts. We achieve an accuracy of > 98% on previously unseen subjects, and a well comparable image quality with the state-of-the-art ECG-based reconstruction. Our method enables an ECG-free workflow for continuous cardiac scans with simultaneous anatomic and functional imaging with multiple contrasts. It can be potentially integrated without adapting the sampling scheme to other continuous sequences by using the imaging data for navigation and reconstruction.

OFx: A method of 4D image construction from free-breathing non-gated MRI slice acquisitions of the thorax via optical flux

PURPOSE

Since real-time 4D dynamic magnetic resonance imaging (dMRI) methods with adequate spatial and temporal resolution for imaging the pediatric thorax are currently not available, free-breathing slice acquisitions followed by appropriate 4D construction methods are currently employed. Self-gating methods, which extract breathing signals only from image information without any external gating technology, have much potential for this purpose, such as for use in studying pediatric thoracic insufficiency syndrome (TIS). Patients with TIS frequently suffer from extreme malformations of the chest wall, diaphragm, and spine, leading to breathing that is very complex, including deep or shallow respiratory cycles. Existing 4D construction methods cannot perform satisfactorily in this scenario, and most are not fully automatic, requiring manual interactive operations. In this paper, we propose a novel fully automatic 4D image construction method based on an image-derived concept called flux to address these challenges.

METHODS

We utilized 25 dMRI data sets from 25 pediatric subjects with no known thoracic anomalies and 58 dMRI data sets from 29 patients with TIS where each patient had a dMRI scan before and after surgery. A time sequence of 80 slices are acquired at each sagittal location continuously at a rate of ~480 ms per slice under free-breathing conditions, with 30-40 sagittal locations across the chest for each subject depending on the thoracic size. In our approach, we first extract the breathing signal for each sagittal location based on the flux of the optical flow vector field of the body region from the image time series. Here, for each time point of respiratory phase, the net flux of the body region can be regarded as the flux going into or out of the body region, which we term Optical Flux (OFx). OFx provides a very robust representation of the real breathing motion of the thorax. OFx allows us to perform a full analysis of all respiratory cycles, extract only normal cycles in a robust manner, and map all extracted normal cycles on to one cosine respiration model for each sagittal location. Subsequently, we re-sample one normal cycle from the respiration model for each location independently. The normal cycle models associated with the different sagittal locations are finally composited to form the final constructed 4D image.

RESULTS

We employ several metrics to evaluate the quality of the 4D construction results: E_ie - error in locating time instants corresponding to end inspiration and end expiration; E_to - deviation from correct temporal order in each detected normal cycle; E_ss - deviation in spatial smoothness; and E_sc - deviation from spatial continuity as scored by a reader. The means and standard deviations of these metrics for normal subjects and TIS patients are found to be, respectively: E_ie: 0.25 ± 0.05 and 0.38 ± 0.16 in units of time instance (ideal value = 0); E_to: 2.7% ± 2.3% and 1.8% ± 2% (ideal value = 0%); E_ss: 0.5 ± 0.17 and 0.54 ± 0.25 in pixel units (ideal value = 0); E_sc: 4.6 ± 0.48 and 4.56 ± 0.98 (score range: best = 5, worst = 1). The results show that the OFx method achieves excellent spatial and temporal continuity and its yield was 100% meaning that it successfully performed 4D construction on every data set tested. Compared to a recently published method, OFx is fully automatic requiring about 5 min of computational time per study starting from acquired dMRI scans. The method achieves high temporal and spatial continuity even on complex TIS data sets that include many abnormal respiratory cycles.

CONCLUSIONS

A new 4D dMRI construction method based on the concept of optical flux is presented which is fully automatic and very robust in deriving respiratory signals purely from dynamic image sequences even when presented with complex breathing patterns due to severe disease conditions like TIS. Evaluations show that its accuracy is comparable to the variations found in manual annotations. An important characteristic of the method is that it is independent of the number of sagittal locations used in the construction process, which suggests that it is applicable to imaging techniques where data are acquired at only a few sagittal locations instead of the full width of the thorax. The method is not tied to any specific imaging modality, as demonstrated in this paper on not just dMRI but dynamic computed tomography (CT) as well.

Weighted Manifold Alignment using Wave Kernel Signatures for Aligning Medical Image Datasets

Manifold alignment (MA) is a technique to map many high-dimensional datasets to one shared low-dimensional space. Here we develop a pipeline for using MA to reconstruct high-resolution medical images. We present two key contributions. First, we develop a novel MA scheme in which each high-dimensional dataset can be differently weighted preventing noisier or less informative data from corrupting the aligned embedding. We find that this generalisation improves performance in our experiments in both supervised and unsupervised MA problems. Second, we use the wave kernel signature as a graph descriptor for the unsupervised MA case finding that it significantly outperforms the current state-of-the-art methods and provides higher quality reconstructed magnetic resonance volumes than existing methods.

Deep Learning for Fast and Spatially Constrained Tissue Quantification From Highly Accelerated Data in Magnetic Resonance Fingerprinting

Magnetic resonance fingerprinting (MRF) is a quantitative imaging technique that can simultaneously measure multiple important tissue properties of human body. Although MRF has demonstrated improved scan efficiency as compared to conventional techniques, further acceleration is still desired for translation into routine clinical practice. The purpose of this paper is to accelerate MRF acquisition by developing a new tissue quantification method for MRF that allows accurate quantification with fewer sampling data. Most of the existing approaches use the MRF signal evolution at each individual pixel to estimate tissue properties, without considering the spatial association among neighboring pixels. In this paper, we propose a spatially constrained quantification method that uses the signals at multiple neighboring pixels to better estimate tissue properties at the central pixel. Specifically, we design a unique two-step deep learning model that learns the mapping from the observed signals to the desired properties for tissue quantification, i.e.: 1) with a feature extraction module for reducing the dimension of signals by extracting a low-dimensional feature vector from the high-dimensional signal evolution and 2) a spatially constrained quantification module for exploiting the spatial information from the extracted feature maps to generate the final tissue property map. A corresponding two-step training strategy is developed for network training. The proposed method is tested on highly undersampled MRF data acquired from human brains. Experimental results demonstrate that our method can achieve accurate quantification for T1 and T2 relaxation times by using only 1/4 time points of the original sequence (i.e., four times of acceleration for MRF acquisition).

Reconstruction techniques for cardiac cine MRI

The present survey describes the state-of-the-art techniques for dynamic cardiac magnetic resonance image reconstruction. Additionally, clinical relevance, main challenges, and future trends of this image modality are outlined. Thus, this paper aims to provide a general vision about cine MRI as the standard procedure in functional evaluation of the heart, focusing on technical methodologies.

Manifold recovery using kernel low-rank regularization: application to dynamic imaging

We introduce a novel kernel low-rank algorithm to recover free-breathing and ungated dynamic MRI data from highly undersampled measurements. The image frames in the free breathing and ungated dataset are assumed to be points on a bandlimited manifold. We show that the non-linear features of these images satisfy annihilation conditions, which implies that the kernel matrix derived from the dataset is low-rank. We penalize the nuclear norm of the feature matrix to recover the images from highly undersampled measurements. The regularized optimization problem is solved using an iterative reweighted least squares (IRLS) algorithm, which alternates between the update of the Laplacian matrix of the manifold and the recovery of the signals from the noisy measurements. To improve computational efficiency, we use a two step algorithm using navigator measurements. Specifically, the Laplacian matrix is estimated from the navigators using the IRLS scheme, followed by the recovery of the images using a quadratic optimization. We show the relation of this two step algorithm with our recent SToRM approach, thus reconciling SToRM and manifold regularization methods with algorithms that rely on explicit lifting of data to a high dimensional space. The IRLS based estimation of the Laplacian matrix is a systematic and noise-robust alternative to current heuristic strategies based on exponential maps. We also approximate the Laplacian matrix using a few eigen vectors, which results in a fast and memory efficient algorithm. The proposed scheme is demonstrated on several patients with different breathing patterns and cardiac rates.

Automatic self-gated 4D-MRI construction from free-breathing 2D acquisitions applied on liver images

PURPOSE

MRI slice reordering is a necessary step when three-dimensional (3D) motion of an anatomical region of interest has to be extracted from multiple two-dimensional (2D) dynamic acquisition planes, e.g., for the construction of motion models used for image-guided radiotherapy. Existing reordering methods focus on obtaining a spatially coherent reconstructed volume for each time. However, little attention has been paid to the temporal coherence of the reconstructed volumes, which is of primary importance for accurate 3D motion extraction. This paper proposes a fully automatic self-sorting four-dimensional MR volume construction method that ensures the temporal coherence of the results.

METHODS

First, a pseudo-navigator signal is extracted for each 2D dynamic slice acquisition series. Then, a weighted graph is created using both spatial and motion information provided by the pseudo-navigator. The volume at a given time point is reconstructed following the shortest paths in the graph starting that time point of a reference slice chosen based on its pseudo-navigator signal.

RESULTS

The proposed method is evaluated against two state-of-the-art slice reordering algorithms on a prospective dataset of 12 volunteers using both spatial and temporal quality metrics. The automated end-exhale extraction showed results closed to the median value of the manual operators. Furthermore, the results of the validation metrics show that the proposed method outperforms state-of-the-art methods in terms of both spatial and temporal quality.

CONCLUSION

Our approach is able to automatically detect the end-exhale phases within one given anatomical position and cope with irregular breathing.