Magnetic Resonance Imaging Reconstruction Research Articles

Comparison of Electrophysiological and Radiological Subthalamic Nucleus Length and Volume.

To investigate the surgical value of MER recordings and improve surgical technique by demonstrating the consistency between preoperative radiological STN volume and intraoperative neurophysiological STN length. Sixty-one patients with PD were enrolled. The volumes of the STN were measured using magnetic resonance images 3-dimensional volume reconstructions of stereotactic magnetic resonance images. MER were performed in all patient and the maximal electrophysiologic length of the STN was recorded each patient. In the postoperative period, the permanent electrode was modeled and reconstructed in 3D, and the longest distance traveled in the STN was calculated. A total of 61 patients who underwent surgery between 2012-2022 were included in the study. Thirty-six (59%) of the patients were male, and 25 (41%) were female. A total of 122 STNs were performed with 166 electrodes. The most common end alignment used was center with 86. STN length averaged 4.9 mm (0-10.5 mm). The mean STN volume was 0.11 cm3. The STN Volume of men were significantly higher than women. The STN Length, Volume, and the target MER length showed a positive correlation significantly. With radiological advances, it is possible to better visualize the target points and define the boundaries better, and direct methods can be used more in making targeting plans. MER records obtained during surgery and STN dimensions in presurgical planning show compatibility, and it is seen that there may be differences between the right and left sides because of brain shifting. Although radiology is increasingly providing better support, electrophysiological recordings provides real-time information on the electrodes? locations and give the opportunity to surgical team choosing alternative target.

MRI Distortion Correction and Robot-to-MRI Scanner Registration for an MRI-Guided Robotic System.

Magnetic resonance imaging (MRI) guided robotic procedures require safe robotic instrument navigation and precise target localization. This depends on reliable tracking of the instrument from MR images, which requires accurate registration of the robot to the scanner. A novel differential image based robot-to-MRI scanner registration approach is proposed that utilizes a set of active fiducial coils, where background subtraction method is employed for coil detection. In order to use the presented preoperative registration approach jointly with the real-time high speed MRI image acquisition and reconstruction methods in real-time interventional procedures, the effects of the geometric MRI distortion in robot to scanner registration is analyzed using a custom distortion mapping algorithm. The proposed approach is validated by a set of target coils placed within the workspace, employing multi-planar capabilities of the scanner. Registration and validation errors are respectively 2.05 mm and 2.63 mm after the distortion correction showing an improvement of respectively 1.08 mm and 0.14 mm compared to the results without distortion correction.

Comparison of a Cryopneumatic Compression Device and Ice Packs for Cryotherapy Following Anterior Cruciate Ligament Reconstruction.

The purpose of the current study was to evaluate and compare the effectiveness of a cryopneumatic compression device with that of standard ice packs following arthroscopic anterior cruciate ligament (ACL) reconstruction, with a primary focus on early postoperative pain. Participants were divided into two groups: cryopneumatic compression device group (CC group) and standard ice pack group (IP group). Patients in the CC Group (28 patients) received a cryopneumatic compression device (CTC-7, Daesung Maref) treatment, while patients in the IP group (28 patients) received standard ice pack cryotherapy postoperatively. All cryotherapy was applied three times (every 8 hours) per day for 20 minutes until discharge (postoperative day 7). Pain scores were assessed preoperatively and at 4, 7, and 14 days after surgery, and the primary outcome for analysis was pain at postoperative day 4 assessed using a visual analog scale (VAS). Other variables were opioid and rescue medication use, knee and thigh circumferences, postoperative drainage, and joint effusion quantified by a three-dimensional magnetic resonance imaging (MRI) reconstruction model. The mean pain VAS score and difference in VAS relative to the preoperative measurements for postoperative day 4 were significantly lower in the CC group than in the IP group (p = 0.001 and p = 0.007, respectively). The sum of postoperative drainage and effusion quantified by MRI showed a significant reduction of postoperative effusion in the CC group compared to the IP group (p = 0.015). The average total rescue medication consumption was comparable between the two groups. Circumferential measurements at days 7 and 14 postoperatively relative to those at day 4 (index day) demonstrated no significant differences between the groups. Compared to standard ice packs, application of cryopneumatic compression was associated with a significant reduction in VAS pain scores and joint effusion during the early postoperative period following ACL reconstruction.

Undersampled magnetic resonance image reconstruction based on support prior and deep image prior without pre-training

Magnetic resonance imaging (MRI) method based on deep learning needs large-quantity and high-quality patient-based datasets for pre-training. However, this is a challenge to the clinical applications because it is difficult to obtain a sufficient quantity of patient-based MR datasets due to the limitation of equipment and patient privacy concerns. In this paper, we propose a novel undersampled MRI reconstruction method based on deep learning. This method does not require any pre-training procedures and does not depend on training datasets. The proposed method is inspired by the traditional deep image prior (DIP) framework, and integrates the structure prior and support prior of the target MR image to improve the efficiency of learning. Based on the similarity between the reference image and the target image, the high-resolution reference image obtained in advance is used as the network input, thereby incorporating the structural prior information into network. By taking the coefficient index set of the reference image with large amplitude in the wavelet domain as the known support of the target image, the regularization constraint term is constructed, and the network training is transformed into the optimization process of network parameters. Experimental results show that the proposed method can obtain more accurate reconstructions from undersampled <i>k</i>-space data, and has obvious advantages in preserving tissue features and detailed texture.

Decentralised Agent-Based Medical Image Reconstruction

In this paper we propose a solution to a challenging task of magnetic resonance image reconstruction in a scalable and flexible way. The proposal elaborates task slicing and its distribution using market-based approach, where participating nodes perform task processing using their own spare computing capacities. The distributed task execution is orchestrated within multi-agent environment where agents offer the execution services by bidding for the offered tasks. Proposed negotiation protocol includes local and cloud-based resources where the offered services are exchanged for an award based on the negotiated award value function. The proposed solution considers simplicity and expressiveness of SDLMAS language and the advantages of corresponding platform for rapid design and development of multi-agent systems.

MARFit: An Integrated Software for Real-Time MR Guided Focused Ultrasound Neuromodulation System

MR guided focused ultrasound (MRgFUS) therapy has been a promising treatment modality for many neurological disorders. However, the lack of real-time image processing software platform sets barriers for relevant pre-clinical researches. This work intends to develop an integrated software for MRgFUS therapy. The software contains three functional modules: a communication module, an image post-processing module, and a visualization module. The communication module provides a data interface with an open-source MR image reconstruction platform (Gadgetron) to receive the reconstructed MR images in real-time. The post-processing module contains the algorithms of image coordinate registration, focus localization by MR acoustic radiation force imaging (MR-ARFI), temperature and thermal dose calculations, motion correction, and temperature feedback control. The visualization module displays monitoring information and provides a user-machine interface. The software was tested to be compatible with systems from two different vendors and validated in multiple scenarios for MRgFUS. The software was tested in many ex vivo and in vivo experiments to validate its functions. The in vivo transcranial focus localization experiments were carried out for targeting the focused ultrasound in neuromodulation.

Evaluation on the generalization of a learned convolutional neural network for MRI reconstruction

Recently, deep learning approaches with various network architectures have drawn significant attention from the magnetic resonance imaging (MRI) community because of their great potential for image reconstruction from undersampled k-space data in fast MRI. However, the robustness of a trained network when applied to test data deviated from training data is still an important open question. In this work, we focus on quantitatively evaluating the influence of image contrast, human anatomy, sampling pattern, undersampling factor, and noise level on the generalization of a trained network composed by a cascade of several CNNs and a data consistency layer, called a deep cascade of convolutional neural network (DC-CNN). The DC-CNN is trained from datasets with different image contrast, human anatomy, sampling pattern, undersampling factor, and noise level, and then applied to test datasets consistent or inconsistent with the training datasets to assess the generalizability of the learned DC-CNN network. The results of our experiments show that reconstruction quality from the DC-CNN network is highly sensitive to sampling pattern, undersampling factor, and noise level, which are closely related to signal-to-noise ratio (SNR), and is relatively less sensitive to the image contrast. We also show that a deviation of human anatomy between training and test data leads to a substantial reduction of image quality for the brain dataset, whereas comparable performance for the chest and knee dataset having fewer anatomy details than brain images. This work further provides some empirical understanding of the generalizability of trained networks when there are deviations between training and test data. It also demonstrates the potential of transfer learning for image reconstruction from datasets different from those used in training the network.

POCS-Augmented CycleGAN for MR Image Reconstruction.

Recent years have seen increased research interest in replacing the computationally intensive Magnetic resonance (MR) image reconstruction process with deep neural networks. We claim in this paper that the traditional image reconstruction methods and deep learning (DL) are mutually complementary and can be combined to achieve better image reconstruction quality. To test this hypothesis, a hybrid DL image reconstruction method was proposed by combining a state-of-the-art deep learning network, namely a generative adversarial network with cycle loss (CycleGAN), with a traditional data reconstruction algorithm: Projection Onto Convex Set (POCS). The output of the first iteration's training results of the CycleGAN was updated by POCS and used as the extra training data for the second training iteration of the CycleGAN. The method was validated using sub-sampled Magnetic resonance imaging data. Compared with other state-of-the-art, DL-based methods (e.g., U-Net, GAN, and RefineGAN) and a traditional method (compressed sensing), our method showed the best reconstruction results.

Adjustable shrinkage-thresholding projection algorithm for compressed sensing magnetic resonance imaging

Compressed sensing (CS) aims to reconstruct a high quality images with as little sample data as possible. Magnetic resonance imaging (MRI) plays an important role in medical imaging tools but has a slower data acquisition process. Applying CS to MRI offers significant scan time reductions. In this paper, we proposed a fast and efficient algorithm for compressed sensing magnetic resonance imaging (CS-MRI) reconstruction, denoted as adjustable shrinkage-thresholding projection algorithm (ASTP). It is designed to use adjustable shrinkage rules for lp-norm based CS-MRI model. This algorithm is established by using an iterative projection and acceleration scheme. In each iteration, the proposed adjustable shrinkage-thresholding rules are employed to ensure global convergence to accurate solution. Furthermore, the parameter p can be selected flexibly according to different practical application situations, and the orthogonal projection operation is used to reduce the dimension of the solution space to accelerate the convergence speed and improve the reconstruction quality. Numerical experiments show that proposed ASTP algorithm provides a higher accuracy, convergence speed and ability to suppress noise compared with some certain state-of-the-art algorithms.

Three-dimensional reconstruction of magnetic resonance images of carp brain for brain control technology

BackgroundIn the field of animal robot control, brain control technology is currently used to achieve control. It is usually necessary to accurately implant brain electrodes into the animal's brain movement area with the help of a brain stereotaxic apparatus, and apply electrical stimulation to achieve control of the animal. The prerequisite for accurate electrode implantation is to study the internal tissues of the carp skull. New MethodWith the help of 3.0 T magnetic resonance imaging (MRI) instrument and 8_CH MRI scanning coil, carp brain magnetic resonance images was obtained. The visualization tool package VTK and the marching cube algorithm were used for surface rendering, the ray casting algorithm was used for volume rendering and reconstruction. ResultsThe three-dimensional reconstruction results could show the carp skull surface contour and internal tissue details, and the measured coordinates after three-dimensional reconstruction of magnetic resonance images could be transformed into three-dimensional positioning coordinates suitable for brain stereotaxic apparatus. Comparison with existing methodsThe three-dimensional reconstruction images based on magnetic resonance could analyze the relative spatial position relationship between the surface structure of the carp's brain and the internal tissue at any angle, and the three-dimensional positioning coordinates of the brain could be obtained quickly and accurately. ConclusionsThe visualization of carp brain magnetic resonance images based on marching cubes algorithm and ray projection algorithm could obtain ideal reconstruction effects, which could be used in the brain control technology of carp robot.

Open-Source Software in Medical Imaging: Case-Based Study with Interdisciplinary Innovative Product Design

Purpose: Recently, comprehensive open-source software (OSS) solutions for medical images have been developed. They offer powerful tools for the management, visualization, storage and analysis of images. The purpose of the present study was to present the features and use of OSS in comparison with commercial software solutions when evaluating medical images.
 Methods: A case study of the development of a tool handle of optimal size and shape using the free OSS “3D Slicer” is presented. Many authors recommend cylindrical-shaped handles to increase comfort, performance and to avoid cumulative trauma, but none have considered the optimal shape of the handle. Hence, we derived a method for obtainingan optimally shaped tool handle based on the power grip hand posture and non-deformed soft tissue. During magnetic resonance imaging (MRI), an outer hand mold was used to maintain the optimal hand posture. 3D Slicer was used for segmentation and 3D reconstruction of MRI. A three-dimensional (3D) hand model was exported to a commercial program for computer-aided design for the optimal tool handle.
 Results: Measurements of the diameters on 3D reconstruction showed that the position of fingers wasere withheld within small deviations, which maximized maximum finger force exertion. An optimally shaped handle this way provided 25% higher contact area compared with the optimal cylindrical handle. It therefore reduced contact pressure, increased performance and comfort and could probably prevent disorders due to cumulative trauma.
 Conclusion: This case study supported the use of OSS for medical imaging as a powerful and (mostly) free tool comparable with (usually expensive) commercial solutions.

Deep Learning Applications in Magnetic Resonance Imaging: Has the Future Become Present?

Deep learning technologies and applications demonstrate one of the most important upcoming developments in radiology. The impact and influence of these technologies on image acquisition and reporting might change daily clinical practice. The aim of this review was to present current deep learning technologies, with a focus on magnetic resonance image reconstruction. The first part of this manuscript concentrates on the basic technical principles that are necessary for deep learning image reconstruction. The second part highlights the translation of these techniques into clinical practice. The third part outlines the different aspects of image reconstruction techniques, and presents a review of the current literature regarding image reconstruction and image post-processing in MRI. The promising results of the most recent studies indicate that deep learning will be a major player in radiology in the upcoming years. Apart from decision and diagnosis support, the major advantages of deep learning magnetic resonance imaging reconstruction techniques are related to acquisition time reduction and the improvement of image quality. The implementation of these techniques may be the solution for the alleviation of limited scanner availability via workflow acceleration. It can be assumed that this disruptive technology will change daily routines and workflows permanently.

Tic Disorder of Children Analyzed and Diagnosed by Magnetic Resonance Imaging Features under Convolutional Neural Network.

This work aimed to explore the analysis and diagnosis of children with tic disorder by magnetic resonance imaging (MRI) features under convolutional neural network (CNN), to provide a certain reference basis for clinical identification. A total of 45 children diagnosed with tic disorder in hospital from January 2018 to June 2020 were selected as the research subjects. A total of 30 normal children were selected as the control group. MRI images were collected, and CNN was constructed for image processing. The results showed that the convolutional neural network could significantly improve the speed of MRI reconstruction and can improve the diagnostic accuracy. Compared with normal children, the metabolites in children with tic disorder were slightly increased, but there was no statistical significance (P > 0.05). The results of the Yale score showed that the proportion of children with moderate illness was significantly greater than that of children with mild and severe illness. In short, the pathological changes of tic disorder were effectively discovered by MRI based on CNN algorithms, which can provide a reference for clinical identification.

MR image reconstruction using densely connected residual convolutional networks

MR image reconstruction techniques based on deep learning have shown their capacity for reducing MRI acquisition time and performance improvement compared to analytical methods. Despite the many challenges in training these rather large networks, novel methodologies have enhanced the capability for having clinical-grade MR image reconstruction in real-time. In recent literature, novel developments have facilitated the utilization of deep networks in various image processing inverse problems. In particular, it has been reported multiple times that the performance of deep networks can be improved by using short connections between layers. In this study, we introduce a novel MRI reconstruction method that utilizes such short connections. The dense connections are used inside densely connected residual blocks. Inside these blocks, the feature maps are concatenated to the subsequent layers. In this way, the extracted information is propagated until the last stage of the block. We have evaluated this densely connected residual block's efficiency in MRI reconstruction settings, by augmenting different types of effective deep network models with these blocks in novel structures. The quantitative and qualitative results indicate that this original introduction of the densely connected blocks to the MR image reconstruction problem improves the reconstruction performance significantly.

An Efficient Light-weight Network for Fast Reconstruction on MR Images.

In recent years, Deep Learning (DL) algorithms have emerged endlessly and achieved impressive performance, which makes it possible to accelerate Magnetic Resonance (MR) image reconstruction with DL instead of Compressed Sensing (CS) methods. However, a DL-based MR image reconstruction method has always suffered from its heavy learning parameters and poor generalization ability so far. Therefore, an efficient, light-weight network is still in desperate need of fast MR image reconstruction. We propose an efficient and light-weight MR reconstruction network (named RecNet) that uses a Convolutional Neural Network (CNN) to fast reconstruct high-quality MR images. Specifically, the network is composed of cascade modules, and each cascade module is further divided into feature extraction blocks and a data consistency layer. The feature extraction block can not only effectively extract the features of MR images, but also do not introduce too many parameters for the whole network. To stabilize the training procedure, the correction information of image frequency is adopted in the Data Consistency (DC) layer. We have evaluated RecNet on a public dataset and the results show that the image quality reconstructed by RecNet is the best on the Peak Signal-To-Noise Ratio (PSNR) and structural similarity index (SSIM) evaluation standards. In addition, the pre-trained RecNet can also reconstruct high-quality MR images on an unseen dataset. The results demonstrate that the RecNet has superior reconstruction ability in various metrics than comparative methods. The RecNet can quickly generate high-quality MR images in fewer parameters. Furthermore, the RecNet has an excellent generalization ability on pathological images and different sampling rates data.

Blind Primed Supervised (BLIPS) Learning for MR Image Reconstruction.

This paper examines a combined supervised-unsupervised framework involving dictionary-based blind learning and deep supervised learning for MR image reconstruction from under-sampled k-space data. A major focus of the work is to investigate the possible synergy of learned features in traditional shallow reconstruction using adaptive sparsity-based priors and deep prior-based reconstruction. Specifically, we propose a framework that uses an unrolled network to refine a blind dictionary learning-based reconstruction. We compare the proposed method with strictly supervised deep learning-based reconstruction approaches on several datasets of varying sizes and anatomies. We also compare the proposed method to alternative approaches for combining dictionary-based methods with supervised learning in MR image reconstruction. The improvements yielded by the proposed framework suggest that the blind dictionary-based approach preserves fine image details that the supervised approach can iteratively refine, suggesting that the features learned using the two methods are complementary.

Learning Data Consistency and its Application to Dynamic MR Imaging.

Magnetic resonance (MR) image reconstruction from undersampled k-space data can be formulated as a minimization problem involving data consistency and image prior. Existing deep learning (DL)-based methods for MR reconstruction employ deep networks to exploit the prior information and integrate the prior knowledge into the reconstruction under the explicit constraint of data consistency, without considering the real distribution of the noise. In this work, we propose a new DL-based approach termed Learned DC that implicitly learns the data consistency with deep networks, corresponding to the actual probability distribution of system noise. The data consistency term and the prior knowledge are both embedded in the weights of the networks, which provides an utterly implicit manner of learning reconstruction model. We evaluated the proposed approach with highly undersampled dynamic data, including the dynamic cardiac cine data with up to 24-fold acceleration and dynamic rectum data with the acceleration factor equal to the number of phases. Experimental results demonstrate the superior performance of the Learned DC both quantitatively and qualitatively than the state-of-the-art.

Parallel MRI Reconstruction Using Broad Learning System.

As an inverse problem, parallel magnetic resonance imaging (pMRI) reconstruction accelerates imaging speed by interpolating missing k-space data from a group of phased-array coils. Deep learning methods have been used for improving pMRI reconstruction quality in recent years. However, deep learning approaches need a large amount of training data that are acquired from different hardware configurations and anatomical areas. Data distributions may be different between training data and testing data, and a long-time training is needed. In this work, we proposed a broad learning system based parallel MRI reconstruction that exploits approximation capability of one-layer neural network through broadening network structure with expanded nodes. Experimental results show that the proposed method is able to suppress noise in compared to the conventional pMRI reconstruction.

An Optimization-Based Meta-Learning Model for MRI Reconstruction with Diverse Dataset

This work aims at developing a generalizable Magnetic Resonance Imaging (MRI) reconstruction method in the meta-learning framework. Specifically, we develop a deep reconstruction network induced by a learnable optimization algorithm (LOA) to solve the nonconvex nonsmooth variational model of MRI image reconstruction. In this model, the nonconvex nonsmooth regularization term is parameterized as a structured deep network where the network parameters can be learned from data. We partition these network parameters into two parts: a task-invariant part for the common feature encoder component of the regularization, and a task-specific part to account for the variations in the heterogeneous training and testing data. We train the regularization parameters in a bilevel optimization framework which significantly improves the robustness of the training process and the generalization ability of the network. We conduct a series of numerical experiments using heterogeneous MRI data sets with various undersampling patterns, ratios, and acquisition settings. The experimental results show that our network yields greatly improved reconstruction quality over existing methods and can generalize well to new reconstruction problems whose undersampling patterns/trajectories are not present during training.

Deep learning-based thin-section MRI reconstruction improves tumour detection and delineation in pre- and post-treatment pituitary adenoma

Even a tiny functioning pituitary adenoma could cause symptoms; hence, accurate diagnosis and treatment are crucial for management. However, it is difficult to diagnose a small pituitary adenoma using conventional MR sequence. Deep learning-based reconstruction (DLR) using magnetic resonance imaging (MRI) enables high-resolution thin-section imaging with noise reduction. In the present single-institution retrospective study of 201 patients, conducted between August 2019 and October 2020, we compared the performance of 1 mm DLR MRI with that of 3 mm routine MRI, using a combined imaging protocol to detect and delineate pituitary adenoma. Four readers assessed the adenomas in a pairwise fashion, and diagnostic performance and image preferences were compared between inexperienced and experienced readers. The signal-to-noise ratio (SNR) was quantitatively assessed. New detection of adenoma, achieved using 1 mm DLR MRI, was not visualised using 3 mm routine MRI (overall: 6.5% [13/201]). There was no significant difference depending on the experience of the readers in new detections. Readers preferred 1 mm DLR MRI over 3 mm routine MRI (overall superiority 56%) to delineate normal pituitary stalk and gland, with inexperienced readers more preferred 1 mm DLR MRI than experienced readers. The SNR of 1 mm DLR MRI was 1.25-fold higher than that of the 3 mm routine MRI. In conclusion, the 1 mm DLR MRI achieved higher sensitivity in the detection of pituitary adenoma and provided better delineation of normal pituitary gland than 3 mm routine MRI.