Beat Pilot Tone (BPT): Simultaneous MRI and RF motion sensing at arbitrary frequencies.

PURPOSE: To introduce a simple system exploitation with the potential to turn MRI scanners into general-purpose radiofrequency (RF) motion monitoring systems. METHODS: Inspired by Pilot Tone (PT), this work proposes Beat Pilot Tone (BPT), in which two or more RF tones at arbitrary frequencies are transmitted continuously during the scan. These tones create motion-modulated standing wave patterns that are sensed by the receiver coil array, incidentally mixed by intermodulation in the receiver chain, and digitized simultaneously with the MRI data. BPT can operate at almost any frequency as long as the intermodulation products lie within the bandwidth of the receivers. BPT's mechanism is explained in electromagnetic simulations and validated experimentally. RESULTS: Phantom and volunteer experiments over a range of transmit frequencies suggest that BPT may offer frequency-dependent sensitivity to motion. Using a semi-flexible anterior receiver array, BPT appears to sense cardiac-induced body vibrations at microwave frequencies ( ≥ $$ \ge $$ 1.2 GHz). At lower frequencies, it exhibits a similar cardiac signal shape to PT, likely due to blood volume changes. Other volunteer experiments with respiratory, bulk, and head motion show that BPT can achieve greater sensitivity to motion than PT and greater separability between motion types. Basic multiple-input multiple-output ( 4 × 22 $$ 4\times 22 $$ MIMO) operation with simultaneous PT and BPT in head motion is demonstrated using two transmit antennas and a 22-channel head-neck coil. CONCLUSION: BPT may offer a rich source of motion information that is frequency-dependent, simultaneous, and complementary to PT and the MRI exam.

Multiphoton simultaneous multislice imaging.

PURPOSE: To develop multiphoton excitation techniques for simultaneous multislice (SMS) imaging and evaluate their performance and specific absorption rate (SAR) benefit. To improve multiphoton SMS reconstruction quality with a novel CAIPIRINHA (controlled aliasing in parallel imaging results in higher acceleration) design. THEORY AND METHODS: When a conventional single-slice RF field is applied together with an oscillating gradient field, the two can combine to generate multiphoton excitation at multiple discrete spatial locations. Because the conventional RF is reused at multiple spatial locations, multiphoton excitation offers reduced SAR for SMS applications. CAIPIRINHA shifts are often used to improve parallel-imaging acceleration. Interestingly, CAIPIRINHA-type shifts can be obtained for multiphoton SMS by updating the oscillating gradient phase at every phase encode. In this work, both a gradient-echo and a spin-echo sequence with multiphoton CAIPIRINHA-SMS excitation pulses are implemented for in vivo human imaging at 3 T. RESULTS: For three slices, multiphoton SMS provides a 51% reduction in SAR compared with conventional superposition SMS, whereas for five slices, SAR is reduced by 66%. Multiphoton SMS outperforms PINS (power independent of number of slices) and MultiPINS in terms of SAR reduction especially when the pulse duration is short, slices are thin, and/or the slice spacing is large. A custom CAIPIRINHA phase-encoding design for multiphoton SMS significantly improves reconstruction quality. CONCLUSION: Multiphoton SMS excitation can be obtained by combining conventional single-slice RF pulses with an oscillating gradient and offers significant SAR benefits compared with conventional superposition SMS. A novel CAIPIRINHA design allows higher multiband factors for multiphoton SMS imaging.

K-band: Self-supervised MRI Reconstruction via Stochastic Gradient Descent over K-space Subsets.

Although deep learning (DL) methods are powerful for solving inverse problems, their reliance on high-quality training data is a major hurdle. This is significant in high-dimensional (dynamic/volumetric) magnetic resonance imaging (MRI), where acquisition of high-resolution fully sampled k-space data is impractical. We introduce a novel mathematical framework, dubbed k-band, that enables training DL models using only partial, limited-resolution k-space data. Specifically, we introduce training with stochastic gradient descent (SGD) over k-space subsets. In each training iteration, rather than using the fully sampled k-space for computing gradients, we use only a small k-space portion. This concept is compatible with different sampling strategies; here we demonstrate the method for k-space "bands", which have limited resolution in one dimension and can hence be acquired rapidly. We prove analytically that our method stochastically approximates the gradients computed in a fully-supervised setup, when two simple conditions are met: (i) the limited-resolution axis is chosen randomly-uniformly for every new scan, hence k-space is fully covered across the entire training set, and (ii) the loss function is weighed with a mask, derived here analytically, which facilitates accurate reconstruction of high-resolution details. Numerical experiments with raw MRI data indicate that k-band outperforms two other methods trained on limited-resolution data and performs comparably to state-of-the-art (SoTA) methods trained on high-resolution data. k-band hence obtains SoTA performance, with the advantage of training using only limited-resolution data. This work hence introduces a practical, easy-to-implement, self-supervised training framework, which involves fast acquisition and self-supervised reconstruction and offers theoretical guarantees.

Pulmonary Ventilation Analysis Using 1H Ultra-Short Echo Time (UTE) Lung MRI: A Reproducibility Study.

Purpose: To evaluate methods for quantification of pulmonary ventilation with ultrashort echo time (UTE) MRI. Methods: We performed a reproducibility study, acquiring two free-breathing 1H UTE lung MRIs on the same day for six healthy volunteers. The 1) 3D + t cyclic b-spline and 2) symmetric image normalization (SyN) methods for image registration were applied after respiratory phase-resolved image reconstruction. Ventilation maps were calculated using 1) Jacobian determinant of the deformation fields minus one, termed regional ventilation, and 2) intensity percentage difference between the registered and fixed image, termed specific ventilation. We compared the reproducibility of all four method combinations via statistical analysis. Results: Split violin plots and Bland-Altman plots are shown for whole lungs and lung sections. The cyclic b-spline registration and Jacobian determinant regional ventilation quantification provide total ventilation volumes that match the segmentation tidal volume, smooth and uniform ventilation maps. The cyclic b-spline registration and specific ventilation combination yields the smallest standard deviation in the Bland-Altman plot. Conclusion: Cyclic registration performs better than SyN for respiratory phase-resolved 1H UTE MRI ventilation quantification. Regional ventilation correlates better with segmentation lung volume, while specific ventilation is more reproducible.

Metal-pad-enhanced resistive pulse sensor reveals complex-valued Braess paradox.

A resistive pulse sensor measures the electrical impedance of an electrolyte-filled channel as particles flow through it. Ordinarily, the presence of a nonconductive particle increases the impedance of the channel. Here we report a surprising experimental result in which a microfluidic resistive pulse sensor experiences the opposite effect: The presence of a nonconductive particle decreases the channel impedance. We explain the counterintuitive phenomenon by relating to the Braess paradox from traffic network theory, and we call it the complex-valued Braess paradox (CVBP). We develop theoretical models to study the CVBP and corroborate the experimental data using finite element simulations and lumped-element circuit modeling. We then discuss implications and potential applications of the CVBP in resistive pulse sensing and beyond.

High-fidelity direct contrast synthesis from magnetic resonance fingerprinting.

PURPOSE: This work was aimed at proposing a supervised learning-based method that directly synthesizes contrast-weighted images from the Magnetic Resonance Fingerprinting (MRF) data without performing quantitative mapping and spin-dynamics simulations. METHODS: To implement our direct contrast synthesis (DCS) method, we deploy a conditional generative adversarial network (GAN) framework with a multi-branch U-Net as the generator and a multilayer CNN (PatchGAN) as the discriminator. We refer to our proposed approach as N-DCSNet. The input MRF data are used to directly synthesize T1-weighted, T2-weighted, and fluid-attenuated inversion recovery (FLAIR) images through supervised training on paired MRF and target spin echo-based contrast-weighted scans. The performance of our proposed method is demonstrated on in vivo MRF scans from healthy volunteers. Quantitative metrics, including normalized root mean square error (nRMSE), peak signal-to-noise ratio (PSNR), structural similarity (SSIM), learned perceptual image patch similarity (LPIPS), and Fréchet inception distance (FID), were used to evaluate the performance of the proposed method and compare it with others. RESULTS: In-vivo experiments demonstrated excellent image quality with respect to that of simulation-based contrast synthesis and previous DCS methods, both visually and according to quantitative metrics. We also demonstrate cases in which our trained model is able to mitigate the in-flow and spiral off-resonance artifacts typically seen in MRF reconstructions, and thus more faithfully represent conventional spin echo-based contrast-weighted images. CONCLUSION: We present N-DCSNet to directly synthesize high-fidelity multicontrast MR images from a single MRF acquisition. This method can significantly decrease examination time. By directly training a network to generate contrast-weighted images, our method does not require any model-based simulation and therefore can avoid reconstruction errors due to dictionary matching and contrast simulation (code available at:https://github.com/mikgroup/DCSNet).

Direct 16S/18S rRNA Gene PCR Followed by Sanger Sequencing as a Clinical Diagnostic Tool for Detection of Bacterial and Fungal Infections: a Systematic Review and Meta-Analysis.

rRNA gene Sanger sequencing is being used for the identification of cultured pathogens. A new diagnostic approach is sequencing of uncultured samples by using the commercial DNA extraction and sequencing platform SepsiTest (ST). The goal was to analyze the clinical performance of ST with a focus on nongrowing pathogens and the impact on antibiotic therapy. A literature search used PubMed/Medline, Cochrane, Science Direct, and Google Scholar. Eligibility followed PRISMA-P criteria. Quality and risk of bias were assessed drawing on QUADAS-2 (quality assessment of diagnostic accuracy studies, revised) criteria. Meta-analyses were performed regarding accuracy metrics compared to standard references and the added value of ST in terms of extra found pathogens. We identified 25 studies on sepsis, infectious endocarditis, bacterial meningitis, joint infections, pyomyositis, and various diseases from routine diagnosis. Patients with suspected infections of purportedly sterile body sites originated from various hospital wards. The overall sensitivity (79%; 95% confidence interval [CI], 73 to 84%) and specificity (83%; 95% CI, 72 to 90%) were accompanied by large effect sizes. ST-related positivity was 32% (95% CI, 30 to 34%), which was significantly higher than the culture positivity (20%; 95% CI, 18 to 22%). The overall added value of ST was 14% (95% CI, 10 to 20%) for all samples. With 130 relevant taxa, ST uncovered high microbial richness. Four studies demonstrated changes of antibiotic treatment at 12% (95% CI, 9 to 15%) of all patients upon availability of ST results. ST appears to be an approach for the diagnosis of nongrowing pathogens. The potential clinical role of this agnostic molecular diagnostic tool is discussed regarding changes of antibiotic treatment in cases where culture stays negative.

Motion-compensated low-rank reconstruction for simultaneous structural and functional UTE lung MRI.

PURPOSE: Three-dimensional UTE MRI has shown the ability to provide simultaneous structural and functional lung imaging, but it is limited by respiratory motion and relatively low lung parenchyma SNR. The purpose of this paper is to improve this imaging by using a respiratory phase-resolved reconstruction approach, named motion-compensated low-rank reconstruction (MoCoLoR), which directly incorporates motion compensation into a low-rank constrained reconstruction model for highly efficient use of the acquired data. THEORY AND METHODS: The MoCoLoR reconstruction is formulated as an optimization problem that includes a low-rank constraint using estimated motion fields to reduce the rank, optimizing over both the motion fields and reconstructed images. The proposed reconstruction along with XD and motion state-weighted motion-compensation (MostMoCo) methods were applied to 18 lung MRI scans of pediatric and young adult patients. The data sets were acquired under free-breathing and without sedation with 3D radial UTE sequences in approximately 5 min. After reconstruction, they went through ventilation analyses. Performance across reconstruction regularization and motion-state parameters were also investigated. RESULTS: The in vivo experiments results showed that MoCoLoR made efficient use of the data, provided higher apparent SNR compared with state-of-the-art XD reconstruction and MostMoCo reconstructions, and yielded high-quality respiratory phase-resolved images for ventilation mapping. The method was effective across the range of patients scanned. CONCLUSION: The motion-compensated low-rank regularized reconstruction approach makes efficient use of acquired data and can improve simultaneous structural and functional lung imaging with 3D-UTE MRI. It enables the scanning of pediatric patients under free-breathing and without sedation.

SLfRank: Shinnar-Le-Roux Pulse Design With Reduced Energy and Accurate Phase Profiles Using Rank Factorization.

The Shinnar-Le-Roux (SLR) algorithm is widely used to design frequency selective pulses with large flip angles. We improve its design process to generate pulses with lower energy (by as much as 26%) and more accurate phase profiles. Concretely, the SLR algorithm consists of two steps: (1) an invertible transform between frequency selective pulses and polynomial pairs that represent Cayley-Klein (CK) parameters and (2) the design of the CK polynomial pair to match the desired magnetization profiles. Because the CK polynomial pair is bi-linearly coupled, the original algorithm sequentially solves for each polynomial instead of jointly. This results in sub-optimal pulses. Instead, we leverage a convex relaxation technique, commonly used for low rank matrix recovery, to address the bi-linearity. Our numerical experiments show that the resulting pulses are almost always globally optimal in practice. For slice excitation, the proposed algorithm results in more accurate linear phase profiles. And in general the improved pulses have lower energy than the original SLR pulses.

Caterpillar traps: A highly flexible, distributed system of toroidal cable traps.

PURPOSE: Coil arrays are connected to the main MRI system with long, shielded coaxial cables. RF coupling of these cables to the main transmit coil can cause high shield currents, which pose risks of heating and RF burns. High-blocking resonant RF traps are placed at distinct positions along cables to mitigate these currents. Traditional traps are designed to be stiff to avoid changes in their resonant frequency, hindering the overall system flexibility. Instead of using a few high-blocking traps, we propose the use of caterpillar traps-a distributed system of small, elastic traps that cover the full length of cables. METHODS: We leverage an array of resonant toroids as traps, forming a caterpillar-like structure whereby bending only impacts individual traps minimally. Benchtop measurements are used to determine the blocking of caterpillar traps and show their robustness to bending. We also compare an anterior array system cable covered with caterpillar traps to a commercial cable with B1 + and heating measurements. RESULTS: Benchtop experiments with caterpillar traps demonstrate high robustness to bending. B1 + mapping experiments of an anterior array cable show improved blocking and flexibility compared to a commercial cable. CONCLUSION: Caterpillar traps provide sufficient attenuation to shield currents while allowing cable flexibility. Our distributed design can provide high blocking efficiency at different positions and orientations, even in cases where commercial cable traps cannot.

Vacuum formed coils for MRI.

PURPOSE: To describe a digital fabrication method used for custom MRI receive coils with vacuum forming and electroless copper plating. METHODS: Our process produces intricate copper traces on curved surfaces. A three-dimensional scan of a desired anatomy is obtained and used to design coil elements. The layout is predistorted with a self built simulation of the vacuum forming process and the geometric overlaps are tested with electromagnetic simulation software. The desired coil geometry is patterned onto a polycarbonate sheet by sandblasting through a tape mask. The sandblasted areas are then catalyzed with a palladium-tin solution and vacuum formed. The catalyzed, three-dimensional part is placed into a custom built plating tank and copper plated. Electronic components are attached to the copper traces to form resonant receive coils. The methods described here are demonstrated and tested with an 8 channel visual cortex coil array. RESULTS: The prototype coils exhibit quality factor ratios higher than three, indicating body noise dominance. The coil array shows high signal-to-noise ratio (SNR) near the periphery of a head shaped phantom. In vivo images with up to 0 . 37 × 0 . 37 × 0 . 67 mm 3 $$ 0.37\times 0.37\times 0.67\;{\mathrm{mm}}^3 $$ spatial resolution were acquired on a human volunteer. CONCLUSION: This work presents the first example of vacuum formed coils with direct electroless copper plating. Our fabrication method results in coil arrays that are in close proximity to the body. This methods described here may enable the rapid development of a set of coils of different sizes for applications including longitudinal fMRI studies and MR-guided therapies.

Mechano-Node-Pore Sensing: A Rapid, Label-Free Platform for Multi-Parameter Single-Cell Viscoelastic Measurements.

Cellular mechanical properties are involved in a wide variety of biological processes and diseases, ranging from stem cell differentiation to cancer metastasis. Conventional methods for measuring these properties, such as atomic force microscopy (AFM) and micropipette aspiration (MA), capture rich information, reflecting a cell's full viscoelastic response; however, these methods are limited by very low throughput. High-throughput approaches, such as real-time deformability cytometry (RT-DC), can only measure limited mechanical information, as they are often restricted to single-parameter readouts that only reflect a cell's elastic properties. In contrast to these methods, mechano-node-pore sensing (mechano-NPS) is a flexible, label-free microfluidic platform that bridges the gap in achieving multi-parameter viscoelastic measurements of a cell with moderate throughput. A direct current (DC) measurement is used to monitor cells as they transit a microfluidic channel, tracking their size and velocity before, during, and after they are forced through a narrow constriction. This information (i.e., size and velocity) is used to quantify each cell's transverse deformation, resistance to deformation, and recovery from deformation. In general, this electronics-based microfluidic platform provides multiple viscoelastic cell properties, and thus a more complete picture of a cell's mechanical state. Because it requires minimal sample preparation, utilizes a straightforward electronic measurement (in contrast to a high-speed camera), and takes advantage of standard soft lithography fabrication, the implementation of this platform is simple, accessible, and adaptable to downstream analysis. This platform's flexibility, utility, and sensitivity have provided unique mechanical information on a diverse range of cells, with the potential for many more applications in basic science and clinical diagnostics.

Deep Learning-Based Water-Fat Separation from Dual-Echo Chemical Shift-Encoded Imaging.

Conventional water-fat separation approaches suffer long computational times and are prone to water/fat swaps. To solve these problems, we propose a deep learning-based dual-echo water-fat separation method. With IRB approval, raw data from 68 pediatric clinically indicated dual echo scans were analyzed, corresponding to 19382 contrast-enhanced images. A densely connected hierarchical convolutional network was constructed, in which dual-echo images and corresponding echo times were used as input and water/fat images obtained using the projected power method were regarded as references. Models were trained and tested using knee images with 8-fold cross validation and validated on out-of-distribution data from the ankle, foot, and arm. Using the proposed method, the average computational time for a volumetric dataset with ~400 slices was reduced from 10 min to under one minute. High fidelity was achieved (correlation coefficient of 0.9969, l1 error of 0.0381, SSIM of 0.9740, pSNR of 58.6876) and water/fat swaps were mitigated. I is of particular interest that metal artifacts were substantially reduced, even when the training set contained no images with metallic implants. Using the models trained with only contrast-enhanced images, water/fat images were predicted from non-contrast-enhanced images with high fidelity. The proposed water-fat separation method has been demonstrated to be fast, robust, and has the added capability to compensate for metal artifacts.

High fidelity deep learning-based MRI reconstruction with instance-wise discriminative feature matching loss.

PURPOSE: To improve reconstruction fidelity of fine structures and textures in deep learning- (DL) based reconstructions. METHODS: A novel patch-based Unsupervised Feature Loss (UFLoss) is proposed and incorporated into the training of DL-based reconstruction frameworks in order to preserve perceptual similarity and high-order statistics. The UFLoss provides instance-level discrimination by mapping similar instances to similar low-dimensional feature vectors and is trained without any human annotation. By adding an additional loss function on the low-dimensional feature space during training, the reconstruction frameworks from under-sampled or corrupted data can reproduce more realistic images that are closer to the original with finer textures, sharper edges, and improved overall image quality. The performance of the proposed UFLoss is demonstrated on unrolled networks for accelerated two- (2D) and three-dimensional (3D) knee MRI reconstruction with retrospective under-sampling. Quantitative metrics including normalized root mean squared error (NRMSE), structural similarity index (SSIM), and our proposed UFLoss were used to evaluate the performance of the proposed method and compare it with others. RESULTS: In vivo experiments indicate that adding the UFLoss encourages sharper edges and more faithful contrasts compared to traditional and learning-based methods with pure ℓ2$$ {\ell}_2 $$ loss. More detailed textures can be seen in both 2D and 3D knee MR images. Quantitative results indicate that reconstruction with UFLoss can provide comparable NRMSE and a higher SSIM while achieving a much lower UFLoss value. CONCLUSION: We present UFLoss, a patch-based unsupervised learned feature loss, which allows the training of DL-based reconstruction to obtain more detailed texture, finer features, and sharper edges with higher overall image quality under DL-based reconstruction frameworks. (Code available at: https://github.com/mikgroup/UFLoss).

Implicit data crimes: Machine learning bias arising from misuse of public data.

SignificancePublic databases are an important resource for machine learning research, but their growing availability sometimes leads to "off-label" usage, where data published for one task are used for another. This work reveals that such off-label usage could lead to biased, overly optimistic results of machine-learning algorithms. The underlying cause is that public data are processed with hidden processing pipelines that alter the data features. Here we study three well-known algorithms developed for image reconstruction from magnetic resonance imaging measurements and show they could produce biased results with up to 48% artificial improvement when applied to public databases. We relate to the publication of such results as implicit "data crimes" to raise community awareness of this growing big data problem.

DiSpect: Displacement spectrum imaging of flow and tissue perfusion using spin-labeling and stimulated echoes.

PURPOSE: We propose a new method, displacement spectrum (DiSpect) imaging, for probing in vivo complex tissue dynamics such as motion, flow, diffusion, and perfusion. Based on stimulated echoes and image phase, our flexible approach enables observations of the spin dynamics over short (milliseconds) to long (seconds) evolution times. METHODS: The DiSpect method is a Fourier-encoded variant of displacement encoding with stimulated echoes, which encodes bulk displacement of spins that occurs between tagging and imaging in the image phase. However, this method fails to capture partial volume effects as well as blood flow. The DiSpect variant mitigates this by performing multiple scans with increasing displacement-encoding steps. Fourier analysis can then resolve the multidimensional spectrum of displacements that spins exhibit over the mixing time. In addition, repeated imaging following tagging can capture dynamic displacement spectra with increasing mixing times. RESULTS: We demonstrate properties of DiSpect MRI using flow phantom experiments as well as in vivo brain scans. Specifically, the ability of DiSpect to perform retrospective vessel-selective perfusion imaging at multiple mixing times is highlighted. CONCLUSION: The DiSpect variant is a new tool in the arsenal of MRI techniques for probing complex tissue dynamics. The flexibility and the rich information it provides open the possibility of alternative ways to quantitatively measure numerous complex spin dynamics, such as flow and perfusion within a single exam.

Quantitative anatomy mimicking slice phantoms.

PURPOSE: To present a reproducible methodology for building an anatomy mimicking phantom with targeted T1 and T2 contrast for use in quantitative magnetic resonance imaging. METHODS: We propose a reproducible method for creating high-resolution, quantitative slice phantoms. The phantoms are created using gels with different concentrations of NiCl2 and MnCl2 to achieve targeted T1 and T2 values. We describe a calibration method for accurately targeting anatomically realistic relaxation pairs. In addition, we developed a method of fabricating slice phantoms by extruding 3D printed walls on acrylic sheets. These procedures are combined to create a physical analog of the Brainweb digital phantom. RESULTS: With our method, we are able to target specific T1 /T2 values with less than 10% error. Additionally, our slice phantoms look realistic since their geometries are derived from anatomical data. CONCLUSION: Standardized and accurate tools for validating new techniques across sequences, platforms, and different imaging sites are important. Anatomy mimicking, multi-contrast phantoms designed with our procedures could be used for evaluating, testing, and verifying model-based methods.

SURE-based automatic parameter selection for ESPIRiT calibration.

PURPOSE: ESPIRiT is a parallel imaging method that estimates coil sensitivity maps from the auto-calibration region (ACS). This requires choosing several parameters for the optimal map estimation. While fairly robust to these parameter choices, occasionally, poor selection can result in reduced performance. The purpose of this work is to automatically select parameters in ESPIRiT for more robust and consistent performance across a variety of exams. METHODS: By viewing ESPIRiT as a denoiser, Stein's unbiased risk estimate (SURE) is leveraged to automatically optimize parameter selection in a data-driven manner. The optimum parameters corresponding to the minimum true squared error, minimum SURE as derived from densely sampled, high-resolution, and non-accelerated data and minimum SURE as derived from ACS are compared using simulation experiments. To avoid optimizing the rank of ESPIRiT's auto-calibrating matrix (one of the parameters), a heuristic derived from SURE-based singular value thresholding is also proposed. RESULTS: Simulations show SURE derived from the densely sampled, high-resolution, and non-accelerated data to be an accurate estimator of the true mean squared error, enabling automatic parameter selection. The parameters that minimize SURE as derived from ACS correspond well to the optimal parameters. The soft-threshold heuristic improves computational efficiency while providing similar results to an exhaustive search. In-vivo experiments verify the reliability of this method. CONCLUSIONS: Using SURE to determine ESPIRiT parameters allows for automatic parameter selections. In-vivo results are consistent with simulation and theoretical results.

Extreme MRI: Large-scale volumetric dynamic imaging from continuous non-gated acquisitions.

PURPOSE: To develop a framework to reconstruct large-scale volumetric dynamic MRI from rapid continuous and non-gated acquisitions, with applications to pulmonary and dynamic contrast-enhanced (DCE) imaging. THEORY AND METHODS: The problem considered here requires recovering 100 gigabytes of dynamic volumetric image data from a few gigabytes of k-space data, acquired continuously over several minutes. This reconstruction is vastly under-determined, heavily stressing computing resources as well as memory management and storage. To overcome these challenges, we leverage intrinsic three-dimensional (3D) trajectories, such as 3D radial and 3D cones, with ordering that incoherently cover time and k-space over the entire acquisition. We then propose two innovations: (a) A compressed representation using multiscale low-rank matrix factorization that constrains the reconstruction problem, and reduces its memory footprint. (b) Stochastic optimization to reduce computation, improve memory locality, and minimize communications between threads and processors. We demonstrate the feasibility of the proposed method on DCE imaging acquired with a golden-angle ordered 3D cones trajectory and pulmonary imaging acquired with a bit-reversed ordered 3D radial trajectory. We compare it with "soft-gated" dynamic reconstruction for DCE and respiratory-resolved reconstruction for pulmonary imaging. RESULTS: The proposed technique shows transient dynamics that are not seen in gating-based methods. When applied to datasets with irregular, or non-repetitive motions, the proposed method displays sharper image features. CONCLUSIONS: We demonstrated a method that can reconstruct massive 3D dynamic image series in the extreme undersampling and extreme computation setting.