arXiv subjects

Martin Uecker

Publications and source records attributed to Martin Uecker.

At least 19 recordsLinked to original sources

arXiv2026

BART Online Open-Source Sequence Toolbox for Computational MRI

Purpose In advanced computational MRI techniques, acquisition and reconstruction techniques are jointly designed. For reproducibility, it is therefore important to provide an open implementation of both. At the same time, any use in a clinical environment usually requires a close integration with the MRI scanner. Ensuring long-time reproducibility and maintenance then poses additional challenges. In this work, we aim to provide a fully integrated open-source framework that can meet these demands. Methods A software framework to develop pulse sequences is added to the BART toolbox. In addition, a vendor-specific driver sequence is developed that can be used to run the sequence on a clinical MRI scanner enabling online adjustment of all relevant sequence parameters. Using the Pulseq format, the exact same sequence can also be reproduced offline. As proof-of-concept, quantitative MRI methods for T1 and joint water/fat R2*, B0 mapping using radial FLASH and model-based reconstruction are implemented in the proposed framework. Consistency between online and offline acquisition is validated in phantom and in vivo experiments. Results Quantitative MRI methods consisting of acquisition and reconstruction were successfully implemented in BART. Acquisition parameters and FOV can be adapted online on a clinical MRI system. Quantitative parameter maps from model-based reconstruction agree for online and offline regenerated Pulseq acquisitions. Conclusion This work enables reproducibility of advanced computational MRI methods within a comprehensive end-to-end open-source framework.

physics.med-ph↗

arXiv2026

Group-kernel auto-calibration and group-patch k-space reconstruction: Fast MRI with time-variant B0 kernels partitioned into time-invariant subsets

Purpose: Pushing MRI speed further demands more spatially-encoded information captured per unit time, e.g., by superimposing additional field modulations during oversampled readout. However, this can introduce calibration errors and increase reconstruction time. Thus, we propose a continuous field calibration approach and an efficient k-space reconstruction technique. Theory and Methods: Our auto-calibration generalizes GRAPPA kernels to explicitly extract continuous B0 modulation kernels, solving interpolation relationships between two ACS regions differing only in the extra field modulation. The k-space locations sharing the same instantaneous image-space modulation are grouped, so that subsets of time-invariant kernels can be separately estimated, as a generalized solution for Wave-CAIPI/FRONSAC-type scans. This view further inspires a k-space subregion-wise reconstruction technique, as an efficient alternative to conventional hybrid-space reconstruction. At 9.4T, FLASH accelerated by a local B0 coil array and Wave-CAIPI were tested with retrospective undersampling. Results: Artifact-free images were reconstructed, under diverse rapid B0 modulation schemes, reaching maximum acceleration factors of 8-fold in 2D and 14.6-fold in 3D. Some nonlinear gradients modulation schemes reach similar sampling efficiency as linear gradients modulation. The proposed reconstruction shows potentials in reducing reconstruction time. Conclusion: Rapid B0 modulations and widely-adopted parallel imaging can share a common mathematical framework, and consequently, achieve similarly-robust reconstructions. Specifically, for scans using dynamic B0 and static RF kernels, not only signal encoding, but also auto-calibration and reconstruction can be performed in k-space. This paves the way to robustly remove eddy currents, and explore more complex B0 modulation strategies towards ultimate MRI speed.

physics.med-ph↗

arXiv2025

Fast and Robust Diffusion Posterior Sampling for MR Image Reconstruction Using the Preconditioned Unadjusted Langevin Algorithm

Purpose: The Unadjusted Langevin Algorithm (ULA) in combination with diffusion models can generate high quality MRI reconstructions with uncertainty estimation from highly undersampled k-space data. However, sampling methods such as diffusion posterior sampling (DPS) or likelihood annealing suffer from long reconstruction times and the need for parameter tuning. The purpose of this work is to develop a robust sampling algorithm with fast convergence. Theory and Methods: In the reverse diffusion process used for sampling the posterior, the exact likelihood is multiplied with the diffused prior at all noise scales. To overcome the issue of slow convergence, preconditioning is used. The method is trained on fastMRI data and tested on retrospectively undersampled brain data of a healthy volunteer. Results: For posterior sampling in Cartesian and non-Cartesian accelerated MRI the new approach outperforms annealed sampling and DPS in terms of reconstruction speed and sample quality. Conclusion: The proposed exact likelihood with preconditioning enables rapid and reliable posterior sampling across various MRI reconstruction tasks without the need for parameter tuning.

physics.med-ph↗

arXiv2025

FENCE: Flexible Electric Noise reduCtion Endo-shield for the Suppression of Electromagnetic Interference in Low-Field MRI

Electromagnetic interference (EMI) is a significant challenge for low-field MRI systems operating without conventional Faraday-shielded rooms. Traditional EMI mitigation approaches include external shields, subject grounding via electrodes, or active noise cancellation requiring synchronized receive channels. These methods either limit portability, introduce patient discomfort, or demand advanced hardware. In this work, we start from the hypothesis that EMI primarily couples capacitively from the body to the RF coil. We investigated two methods of blocking capacitive coupling while preserving inductive MRI signal detection: First, we employed capacitive segmentation of the RF coil and studied its effect on EMI coupling. Second, we present FENCE (Flexible Electromagnetic Noise reduCtion Endo-shield), a novel approach blocking capacitive coupling using flexible PCB shields placed inside the RF coil. FENCE can be retrofitted to existing RF coils. Finite element (FE) simulations were used to estimate the expected shielding performance and the impact on RF coil losses prior to practical implementation. Testing in various realistic scenarios then demonstrated that the combination of FENCE with segmented solenoid coils is effective against both environmental noise sources and controlled EMI. In phantom experiments, FENCE significantly improved imaging performance and reduced EMI levels to near-baseline levels with 9% reduction in coil quality factor (Q factor), showing good agreement with the predictions from the FE simulations. In-vivo head imaging confirmed these results across diverse electromagnetic environments significantly improving imaging performance while showing an ~18% decrease in Q factor. FENCE provides a simple method for EMI mitigation in low-field MRI, enhancing image quality while maintaining system portability and accessibility.

physics.med-ph↗

arXiv2025

BART Streams: Real-time Reconstruction Using a Modular Framework for Pipeline Processing

Purpose: To create modular solutions for interactive real-time MRI using reconstruction algorithms implemented in BART. Methods: A new protocol for streaming of multidimensional arrays is presented and integrated into BART. The new functionality is demonstrated using examples for cardiac interactive real-time MRI based on radial FLASH, where iterative reconstruction is combined with advanced features such as dynamic coil compression and gradient-delay orrection. We analyze the latency of the reconstruction and measure end-to-end latency of the full imaging process. Results: Reconstruction pipelines with iterative reconstruction and advanced functionality were built in a modular way using scripting. Latency measurements demonstrate latency sufficient for interactive real-time MRI, on the order of 30 ms for BART processing and network transfer time, or 200 ms for end-to-end latency including acquisition, vendor processing, and display. Conclusion: With the new streaming capabilities, real-time reconstruction pipelines can be assembled using BART in a flexible way, enabling rapid prototyping of advanced applications such as interactive real-time MRI.

physics.med-ph↗

arXiv2025

Dynamic transitions for fast joint acquisition and reconstruction of CEST-Rex and T1

Purpose: This work proposes a method for the simultaneous estimation of the exchange-dependent relaxation rate Rex and the longitudinal relaxation time T1 from a single acquisition. Methods: A novel acquisition scheme was developed that combines CEST saturation with an inversion pulse and a Look-Locker readout to capture the magnetization evolution starting from the inverse transient Z-spectrum. The corresponding signal model, derived from the Bloch-McConnell equations, describes both the transient Z-spectrum and the Look-Locker dynamics. A model-based reconstruction approach is employed to jointly estimate Rex and T1. The proposed method was validated using a numerical phantom and benchmarked against conventional CEST and Look-Locker T1 mapping in phantom and in vivo on a clinical 3T scanner. Results: The joint estimation approach demonstrated strong agreement with ground truth and conventional methods across a wide range of T1 and CEST parameters. The acquisition time was reduced by 20-30% compared to standard CEST protocols, while providing higher signal-to-noise ratio (SNR) in parameter maps. Conclusion: The proposed technique enables robust and efficient simultaneous quantification of CEST Rex and T1 in a single acquisition. It improves parameter map quality and reduces scan time, making it suitable for both phantom and in vivo imaging across a wide range of physiological conditions.

physics.med-ph↗

arXiv2025

Phase-Pole-Free Images and Smooth Coil Sensitivity Maps by Regularized Nonlinear Inversion

Purpose: Phase singularities are a common problem in image reconstruction with auto-calibrated sensitivities due to an inherent ambiguity of the estimation problem. The purpose of this work is to develop a method for detecting and correcting phase poles in non-linear inverse (NLINV) reconstruction of MR images and coil sensitivity maps. Methods: Phase poles are detected in individual coil sensitivity maps by computing the curl in each pixel. A weighted average of the curl in each coil is computed to detect phase poles. Phase pole detection and correction is then integrated into the iteratively regularized Gauss-Newton method of the NLINV algorithm, which then avoid phase singularities in the reconstructed images. The method is evaluated for reconstruction of accelerated Cartesian MPRAGE data of the brain and interactive radial real-time MRI of the human heart. Results: Phase poles are reliably removed in NLINV reconstructions for both applications. NLINV with phase pole correction can reliably and efficiently estimate coil sensitivity profiles free from singularities even from very small ($7\times7$) auto-calibration (AC) regions. Conclusion: NLINV emerges as an efficient and reliable tool for image reconstruction and coil sensitivity estimation in challenging MRI applications.

physics.med-ph↗

arXiv2024

Rapid, High-resolution and Distortion-free $R_{2}^{*}$ Mapping of Fetal Brain using Multi-echo Radial FLASH and Model-based Reconstruction

Purpose: To develop a rapid, high-resolution and distortion-free technique for simultaneous water-fat separation, $R_{2}^{*}$ and $B_{0}$ mapping of the fetal brain at 3T. Methods: A 2D multi-echo radial FLASH sequence with blip gradients is adapted for data acquisition during maternal free breathing. A calibrationless model-based reconstruction with sparsity constraints is developed to jointly estimate water, fat, $R_{2}^{*}$ and $B_{0}$ field maps directly from k-space. This approach was validated and compared to reference methods using numerical and NIST phantoms and data from nine fetuses between 26 and 36 weeks of gestation age. Results: Both numerical and experimental phantom studies confirm good accuracy and precision. In fetal studies, model-based reconstruction yields quantitative $R_{2}^{*}$ values in close agreement with those from a parallel imaging compressed sensing (PICS) technique using Graph Cut (intra-class correlation coefficient [ICC] = 0.9601), while providing enhanced image detail. Repeated scans confirm good reproducibility (ICC = 0.9213). Compared to multi-echo EPI, the proposed radial technique produces higher-resolution (1.1 $\times$ 1.1 $\times$ 3 mm$^{3}$ vs. 2-3 $\times$ 2-3 $\times$ 3 mm$^{3}$) $R_{2}^{*}$ maps with reduced distortion. Despite of differences in motion, resolution and distortion, $R_{2}^{*}$ values are comparable between the two acquisition strategies (ICC = 0.8049). Additionally, the proposed approach enables synthesis of high-resolution and distortion-free $R_{2}^{*}$-weighted images. Conclusion: This study demonstrates the feasibility of using multi-echo radial FLASH combined with calibrationless model-based reconstruction for motion-robust, distortion-free $R_{2}^{*}$ mapping of the fetal brain at 3T, achieving a nominal resolution of $1.1 \times 1.1 \times 3$ mm$^{3}$ within 2 seconds per slice.

physics.med-ph↗

arXiv2024

Autoregressive Image Diffusion: Generation of Image Sequence and Application in MRI

Magnetic resonance imaging (MRI) is a widely used non-invasive imaging modality. However, a persistent challenge lies in balancing image quality with imaging speed. This trade-off is primarily constrained by k-space measurements, which traverse specific trajectories in the spatial Fourier domain (k-space). These measurements are often undersampled to shorten acquisition times, resulting in image artifacts and compromised quality. Generative models learn image distributions and can be used to reconstruct high-quality images from undersampled k-space data. In this work, we present the autoregressive image diffusion (AID) model for image sequences and use it to sample the posterior for accelerated MRI reconstruction. The algorithm incorporates both undersampled k-space and pre-existing information. Models trained with fastMRI dataset are evaluated comprehensively. The results show that the AID model can robustly generate sequentially coherent image sequences. In MRI applications, the AID can outperform the standard diffusion model and reduce hallucinations, due to the learned inter-image dependencies. The project code is available at https://github.com/mrirecon/aid.

eess.IV↗

arXiv2024

LiveDocs: Crafting Interactive Development Environments From Research Findings

Open Science is a recurrent topic in scientific discussion, and there is a current effort to make research more accessible to a broader audience. A focus on delivering research findings that are reproducible, or even re-usable has been proposed as one way of achieving such accessibility goals. In this work, we present the LiveDocs initiative, an effort of the ``Collaborative Research Center 1456 - Mathematics of Experiment'' on tackling common issues of reproducibility and re-usability in scientific publications. The LiveDocs initiative is proposed as a concept alongside a collection of methods that enable scientists to provide research findings under an interactive development environment. This environment allows users from a broader audience to easily reproduce research findings by re-running scripts, for instance, those that generate figures, tables, and other elements from scientific publications. Moreover, LiveDocs also allow the audience to interact with code and data in such environments, thus allowing users to explore algorithms, datasets and software interfaces. This directly lowers the barriers to access and comprehend research methods and findings, which facilitates more scientific exchange and fosters knowledge advancement.

cs.HC↗

arXiv2024

Self-Supervised Learning for Improved Calibrationless Radial MRI with NLINV-Net

Purpose: To develop a neural network architecture for improved calibrationless reconstruction of radial data when no ground truth is available for training. Methods: NLINV-Net is a model-based neural network architecture that directly estimates images and coil sensitivities from (radial) k-space data via non-linear inversion (NLINV). Combined with a training strategy using self-supervision via data undersampling (SSDU), it can be used for imaging problems where no ground truth reconstructions are available. We validated the method for (1) real-time cardiac imaging and (2) single-shot subspace-based quantitative T1 mapping. Furthermore, region-optimized virtual (ROVir) coils were used to suppress artifacts stemming from outside the FoV and to focus the k-space based SSDU loss on the region of interest. NLINV-Net based reconstructions were compared with conventional NLINV and PI-CS (parallel imaging + compressed sensing) reconstruction and the effect of the region-optimized virtual coils and the type of training loss was evaluated qualitatively. Results: NLINV-Net based reconstructions contain significantly less noise than the NLINV-based counterpart. ROVir coils effectively suppress streakings which are not suppressed by the neural networks while the ROVir-based focussed loss leads to visually sharper time series for the movement of the myocardial wall in cardiac real-time imaging. For quantitative imaging, T1-maps reconstructed using NLINV-Net show similar quality as PI-CS reconstructions, but NLINV-Net does not require slice-specific tuning of the regularization parameter. Conclusion: NLINV-Net is a versatile tool for calibrationless imaging which can be used in challenging imaging scenarios where a ground truth is not available.

physics.med-ph↗

arXiv2024

Model-Based Reconstruction for Joint Estimation of $T_{1}$, $R_{2}^{*}$ and $B_{0}$ Field Maps Using Single-Shot Inversion-Recovery Multi-Echo Radial FLASH

Purpose: To develop a model-based nonlinear reconstruction for simultaneous water-specific $T_{1}$, $R_{2}^{*}$, $B_{0}$ field and/or fat fraction (FF) mapping using single-shot inversion-recovery (IR) multi-echo radial FLASH. Methods: The proposed model-based reconstruction jointly estimates water-specific $T_{1}$, $R_{2}^{*}$, $B_{0}$ field and/or FF maps, as well as a set of coil sensitivities directly from $k$-space obtained with a single-shot IR multi-echo radial FLASH sequence using blip gradients across echoes. Joint sparsity constraints are exploited on multiple quantitative maps to improve precision. Validations are performed on numerical and NIST phantoms and with in vivo studies of the human brain and liver at 3 T. Results: Numerical phantom studies demonstrate the effects of fat signals in $T_{1}$ estimation and confirm good quantitative accuracy of the proposed method for all parameter maps. NIST phantom results confirm good quantitative $T_{1}$ and $R_{2}^{*}$ accuracy in comparison to Cartesian references. Apart from good quantitative accuracy and precision for multiple parameter maps, in vivo studies show improved image details utilizing the proposed joint estimation. The proposed method can achieve simultaneous water-specific $T_{1}$, $R_{2}^{*}$, $B_{0}$ field and/or FF mapping for brain (0.81 $\times$ 0.81 $\times$ 5 mm$^{3}$) and liver (1.6 $\times$ 1.6 $\times$ 6 mm$^{3}$) imaging within four seconds. Conclusion: The proposed model-based nonlinear reconstruction, in combination with a single-shot IR multi-echo radial FLASH acquisition, enables joint estimation of accurate water-specific $T_{1}$, $R_{2}^{*}$, $B_{0}$ field and/or FF maps within four seconds. The present work is of potential value for specific clinical applications.

physics.med-ph↗

arXiv2024

Rational Approximation of Golden Angles: Accelerated Reconstructions for Radial MRI

Purpose: To develop a generic radial sampling scheme that combines the advantages of golden ratio sampling with simplicity of equidistant angular patterns. The irrational angle between consecutive spokes in golden ratio based sampling schemes enables a flexible retrospective choice of temporal resolution, while preserving good coverage of k-space for each individual bin. Nevertheless, irrational increments prohibit precomputation of the point-spread function (PSF), can lead to numerical problems, and require more complex processing steps. To avoid these problems, a new sampling scheme based on a rational approximation of golden angles (RAGA) is developed. Methods: The theoretical properties of RAGA sampling are mathematically derived. Sidelobe-to-peak ratios (SPR) are numerically computed and compared to the corresponding golden ratio sampling schemes. The sampling scheme is implemented in the BART toolbox and in a radial gradient-echo sequence. Feasibility is shown for quantitative imaging in a phantom and a cardiac scan of a healthy volunteer. Results: RAGA sampling can accurately approximate golden ratio sampling and has almost identical PSF and SPR. In contrast to golden ratio sampling, each frame can be reconstructed with the same equidistant trajectory using different sampling masks, and the angle of each acquired spoke can be encoded as a small index, which simplifies processing of the acquired data. Conclusion: RAGA sampling provides the advantages of golden ratio sampling while simplifying data processing, rendering it a valuable tool for dynamic and quantitative MRI.

physics.med-ph↗

arXiv2023

Assessment of Deep Learning Segmentation for Real-Time Free-Breathing Cardiac Magnetic Resonance Imaging at Rest and Under Exercise Stress

In recent years, a variety of deep learning networks for cardiac MRI (CMR) segmentation have been developed and analyzed. However, nearly all of them are focused on cine CMR under breathold. In this work, accuracy of deep learning methods is assessed for volumetric analysis (via segmentation) of the left ventricle in real-time free-breathing CMR at rest and under exercise stress. Data from healthy volunteers (n=15) for cine and real-time free-breathing CMR at rest and under exercise stress were analyzed retrospectively. Segmentations of a commercial software (comDL) and a freely available neural network (nnU-Net), were compared to a reference created via the manual correction of comDL segmentation. Segmentation of left ventricular endocardium (LV), left ventricular myocardium (MYO), and right ventricle (RV) is evaluated for both end-systolic and end-diastolic phases and analyzed with Dice's coefficient (DC). The volumetric analysis includes LV end-diastolic volume (EDV), LV end-systolic volume (ESV), and LV ejection fraction (EF). For cine CMR, nnU-Net and comDL achieve a DC above 0.95 for LV and 0.9 for MYO, and RV. For real-time CMR, the accuracy of nnU-Net exceeds that of comDL overall. For real-time CMR at rest, nnU-Net achieves a DC of 0.94 for LV, 0.89 for MYO, and 0.90 for RV; mean absolute differences between nnU-Net and reference are 2.9mL for EDV, 3.5mL for ESV and 2.6% for EF. For real-time CMR under exercise stress, nnU-Net achieves a DC of 0.92 for LV, 0.85 for MYO, and 0.83 for RV; mean absolute differences between nnU-Net and reference are 11.4mL for EDV, 2.9mL for ESV and 3.6% for EF. Deep learning methods designed or trained for cine CMR segmentation can perform well on real-time CMR. For real-time free-breathing CMR at rest, the performance of deep learning methods is comparable to inter-observer variability in cine CMR and is usable or fully automatic segmentation.

eess.IV↗

arXiv2023

Generative Priors for MRI Reconstruction Trained from Magnitude-Only Images Using Phase Augmentation

Purpose: In this work, we present a workflow to construct generic and robust generative image priors from magnitude-only images. The priors can then be used for regularization in reconstruction to improve image quality. Methods: The workflow begins with the preparation of training datasets from magnitude-only MR images. This dataset is then augmented with phase information and used to train generative priors of complex images. Finally, trained priors are evaluated using both linear and nonlinear reconstruction for compressed sensing parallel imaging with various undersampling schemes. Results: The results of our experiments demonstrate that priors trained on complex images outperform priors trained only on magnitude images. Additionally, a prior trained on a larger dataset exhibits higher robustness. Finally, we show that the generative priors are superior to L1 -wavelet regularization for compressed sensing parallel imaging with high undersampling. Conclusion: These findings stress the importance of incorporating phase information and leveraging large datasets to raise the performance and reliability of the generative priors for MRI reconstruction. Phase augmentation makes it possible to use existing image databases for training.

eess.IV↗

arXiv2022

Quantitative Magnetic Resonance Imaging by Nonlinear Inversion of the Bloch Equations

Purpose: Development of a generic model-based reconstruction framework for multi-parametric quantitative MRI that can be used with data from different pulse sequences. Methods: Generic nonlinear model-based reconstruction for quantitative MRI estimates parametric maps directly from the acquired k-space by numerical optimization. This requires numerically accurate and efficient methods to solve the Bloch equations and their partial derivatives. In this work, we combine direct sensitivity analysis and pre-computed state-transition matrices into a generic framework for calibrationless model-based reconstruction that can be applied to different pulse sequences. As a proof-of-concept, the method is implemented and validated for quantitative $T_1$ and $T_2$ mapping with single-shot inversion-recovery (IR) FLASH and IR bSSFP sequences in simulations, phantoms, and the human brain. Results: The direct sensitivity analysis enables a highly accurate and numerically stable calculation of the derivatives. The state-transition matrices efficiently exploit repeating patterns in pulse sequences, speeding up the calculation by a factor of 10 for the examples considered in this work, while preserving the accuracy of native ODE solvers. The generic model-based method reproduces quantitative results of previous model-based reconstructions based on the known analytical solutions for radial IR FLASH. For IR bSFFP it produces accurate $T_1$ and $T_2$ maps for the NIST phantom in numerical simulations and experiments. Feasibility is also shown for human brain, although results are affected by magnetization transfer effects. Conclusion: By developing efficient tools for numerical optimizations using the Bloch equations as forward model, this work enables generic model-based reconstruction for quantitative MRI.

physics.med-ph↗

arXiv2022

In silico modeling for personalized stenting in aortic coarctation

Stent intervention is a recommended therapy to reduce the pressure gradient and restore blood flow for patients with coarctation of the aorta (CoA). In this work, we developed a framework for personalized stent intervention in CoA using in silico modeling, combining computational fluid dynamics (CFD) and image-based prediction of the geometry of the aorta after stent intervention. Firstly, the blood flow in the aorta, whose geometry was reconstructed from magnetic resonance imaging (MRI) data, was numerically modeled using the lattice Boltzmann method (LBM). Both large eddy simulation (LES) and direct numerical simulation (DNS) were considered to adequately resolve the turbulent hemodynamics, with boundary conditions extracted from phase-contrast flow MRI. By comparing the results from CFD and 4D-Flow MRI in 3D-printed flow phantoms, we concluded that the LBM based LES is capable of obtaining accurate aortic flow with acceptable computational cost. In silico stent implantation for a patient with CoA was then performed by predicting the deformed geometry after stent intervention and predicting the blood flow. By evaluating the pressure drop and maximum wall shear stress, an optimal stent can be selected.

physics.med-ph↗

arXiv2022

Deep, Deep Learning with BART

Purpose: To develop a deep-learning-based image reconstruction framework for reproducible research in MRI. Methods: The BART toolbox offers a rich set of implementations of calibration and reconstruction algorithms for parallel imaging and compressed sensing. In this work, BART was extended by a non-linear operator framework that provides automatic differentiation to allow computation of gradients. Existing MRI-specific operators of BART, such as the non-uniform fast Fourier transform, are directly integrated into this framework and are complemented by common building blocks used in neural networks. To evaluate the use of the framework for advanced deep-learning-based reconstruction, two state-of-the-art unrolled reconstruction networks, namely the Variational Network [1] and MoDL [2], were implemented. Results: State-of-the-art deep image-reconstruction networks can be constructed and trained using BART's gradient based optimization algorithms. The BART implementation achieves a similar performance in terms of training time and reconstruction quality compared to the original implementations based on TensorFlow. Conclusion: By integrating non-linear operators and neural networks into BART, we provide a general framework for deep-learning-based reconstruction in MRI.

cs.CV↗