Faculty Bibliography

BACKGROUND:Computed tomography (CT) and spirometry are the current standard methods for assessing lung anatomy and pulmonary ventilation, respectively. However, CT provides limited ventilation information and spirometry only provides global measures of lung ventilation. Thus, a method that can enable simultaneous examination of lung anatomy and ventilation is of clinical interest. PURPOSE/OBJECTIVE:To develop and test a 4D respiratory-resolved sparse lung MRI (XD-UTE: eXtra-Dimensional Ultrashort TE imaging) approach for simultaneous evaluation of lung anatomy and pulmonary ventilation. STUDY TYPE/METHODS:Prospective. POPULATION/METHODS:In all, 23 subjects (11 volunteers and 12 patients, mean age = 63.6 ± 8.4). FIELD STRENGTH/SEQUENCE/UNASSIGNED:3T MR; a prototype 3D golden-angle radial UTE sequence, a Cartesian breath-hold volumetric-interpolated examination (BH-VIBE) sequence. ASSESSMENT/RESULTS:All subjects were scanned using the 3D golden-angle radial UTE sequence during normal breathing. Ten subjects underwent an additional scan during alternating normal and deep breathing. Respiratory-motion-resolved sparse reconstruction was performed for all the acquired data to generate dynamic normal-breathing or deep-breathing image series. For comparison, BH-VIBE was performed in 12 subjects. Lung images were visually scored by three experienced chest radiologists and were analyzed by two observers who segmented the left and right lung to derive ventilation parameters in comparison with spirometry. STATISTICAL TESTS/UNASSIGNED:Nonparametric paired two-tailed Wilcoxon signed-rank test; intraclass correlation coefficient, Pearson correlation coefficient. RESULTS:XD-UTE achieved significantly improved image quality compared both with Cartesian BH-VIBE and radial reconstruction without motion compensation (P < 0.05). The global ventilation parameters (a sum of the left and right lung measures) were in good correlation with spirometry in the same subjects (correlation coefficient = 0.724). There were excellent correlations between the results obtained by two observers (intraclass correlation coefficient ranged from 0.8855-0.9995). DATA CONCLUSION/UNASSIGNED:Simultaneous evaluation of lung anatomy and ventilation using XD-UTE is demonstrated, which have shown good potential for improved diagnosis and management of patients with heterogeneous lung diseases. LEVEL OF EVIDENCE/METHODS:2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018.

PURPOSE/OBJECTIVE:Although deep learning has shown great promise for MR image reconstruction, an open question regarding the success of this approach is the robustness in the case of deviations between training and test data. The goal of this study is to assess the influence of image contrast, SNR, and image content on the generalization of learned image reconstruction, and to demonstrate the potential for transfer learning. METHODS:Reconstructions were trained from undersampled data using data sets with varying SNR, sampling pattern, image contrast, and synthetic data generated from a public image database. The performance of the trained reconstructions was evaluated on 10 in vivo patient knee MRI acquisitions from 2 different pulse sequences that were not used during training. Transfer learning was evaluated by fine-tuning baseline trainings from synthetic data with a small subset of in vivo MR training data. RESULTS:Deviations in SNR between training and testing led to substantial decreases in reconstruction image quality, whereas image contrast was less relevant. Trainings from heterogeneous training data generalized well toward the test data with a range of acquisition parameters. Trainings from synthetic, non-MR image data showed residual aliasing artifacts, which could be removed by transfer learning-inspired fine-tuning. CONCLUSION/CONCLUSIONS:This study presents insights into the generalization ability of learned image reconstruction with respect to deviations in the acquisition settings between training and testing. It also provides an outlook for the potential of transfer learning to fine-tune trainings to a particular target application using only a small number of training cases.

Following the success of deep learning in a wide range of applications, neural network-based machine learning techniques have received interest as a means of accelerating magnetic resonance imaging (MRI). A number of ideas inspired by deep learning techniques from computer vision and image processing have been successfully applied to non-linear image reconstruction in the spirit of compressed sensing for both low dose computed tomography and accelerated MRI. The additional integration of multi-coil information to recover missing k-space lines in the MRI reconstruction process, is still studied less frequently, even though it is the de-facto standard for currently used accelerated MR acquisitions. This manuscript provides an overview of the recent machine learning approaches that have been proposed specifically for improving parallel imaging. A general background introduction to parallel MRI is given that is structured around the classical view of image space and k-space based methods. Both linear and non-linear methods are covered, followed by a discussion of recent efforts to further improve parallel imaging using machine learning, and specifically using artificial neural networks. Image-domain based techniques that introduce improved regularizers are covered as well as k-space based methods, where the focus is on better interpolation strategies using neural networks. Issues and open problems are discussed as well as recent efforts for producing open datasets and benchmarks for the community.

We develop and evaluate a neural network-based method for Gibbs artifact and noise removal. A convolutional neural network (CNN) was designed for artifact removal in diffusion-weighted imaging data. Two implementations were considered: one for magnitude images and one for complex images. Both models were based on the same encoder-decoder structure and were trained by simulating MRI acquisitions on synthetic non-MRI images. Both machine learning methods were able to mitigate artifacts in diffusion-weighted images and diffusion parameter maps. The CNN for complex images was also able to reduce artifacts in partial Fourier acquisitions. The proposed CNNs extend the ability of artifact correction in diffusion MRI. The machine learning method described here can be applied on each imaging slice independently, allowing it to be used flexibly in clinical applications

The dynamics of large spin-1/2 ensembles are commonly described by the Bloch equation, which is characterized by the magnetization's non-linear response to the driving magnetic field. Consequently, most magnetic field variations result in non-intuitive spin dynamics, which are sensitive to small calibration errors. Although simplistic field variations result in robust spin dynamics, they do not explore the richness of the system's phase space. Here, we identify adiabaticity conditions that span a large experiment design space with tractable dynamics. All dynamics are trapped in a one-dimensional subspace, namely in the magnetization's absolute value, which is in a transient state, while its direction adiabatically follows the steady state. In this hybrid state, the polar angle is the effective drive of the spin dynamics. As an example, we optimize this drive for robust and efficient quantification of spin relaxation times and utilize it for magnetic resonance imaging of the human brain.

Accelerating Magnetic Resonance Imaging (MRI) by taking fewer measurements has the potential to reduce medical costs, minimize stress to patients and make MRI possible in applications where it is currently prohibitively slow or expensive. We introduce the fastMRI dataset, a large-scale collection of both raw MR measurements and clinical MR images, that can be used for training and evaluation of machine-learning approaches to MR image reconstruction. By introducing standardized evaluation criteria and a freely-accessible dataset, our goal is to help the community make rapid advances in the state of the art for MR image reconstruction. We also provide a self-contained introduction to MRI for machine learning researchers with no medical imaging background

PURPOSE/OBJECTIVE:To explore the motion sensitivity of magnetic resonance fingerprinting (MRF), we performed experiments with different types of motion at various time intervals during multiple scans. Additionally, we investigated the possibility to correct the motion artifacts based on redundancy in MRF data. METHODS:A radial version of the FISP-MRF sequence was used to acquire one transverse slice through the brain. Three subjects were instructed to move in different patterns (in-plane rotation, through-plane wiggle, complex movements, adjust head position, and pretend itch) during different time intervals. The potential to correct motion artifacts in MRF by removing motion-corrupted data points from the fingerprints and dictionary was evaluated. RESULTS:values (-10% on average). CONCLUSION/CONCLUSIONS:Our experimental results showed that different kinds of motion have distinct effects on the precision and effective resolution of the parametric maps measured with MRF. Although MRF-based acquisitions can be relatively robust to motion effects occurring at the beginning or end of the sequence, relying on redundancy in the data alone is not sufficient to assure the accuracy of the multi-parametric maps in all cases.

PURPOSE/OBJECTIVE:The ultimate intrinsic signal-to-noise ratio (UISNR) is normally calculated using electrodynamic simulations with a complete basis of modes. Here, we provide an exact solution for the UISNR at the center of a dielectric sphere and assess how accurately this solution approximates UISNR away from the center. METHODS:We performed a mode analysis to determine which modes contribute to central UISNR - ζ(r→0). We then derived an analytic expression to calculate ζ(r→0) and analyzed its dependence on main magnetic field strength, sample geometry, and electrical properties. We validated the proposed solution against an established method based on dyadic Green's function simulations. RESULTS:Only one divergence-free mode contributes to ζ(r→0). The UISNR given by the exact solution matched the full simulation results for various parameter settings, whereas calculation speed was approximately 1000 times faster. We showed that the analytic expression can approximate the UISNR with <5% error at positions as much as 10-20% of the radius away from the center. CONCLUSION/CONCLUSIONS:The proposed formula enables rapid and direct calculation of UISNR in the central region of a sphere. The resulting UISNR value may be used, for example, as an absolute reference to assess the performance of head coils with spherical phantoms.

PURPOSE/OBJECTIVE:-weighted images from a single scan and allows for free-breathing acquisition. THEORY AND METHODS/UNASSIGNED:-weighted gradient-echo (GRE) data. Improved robustness is achieved by extracting a respiratory signal from the GRE data and using it for motion-weighted reconstruction. RESULTS:-weighted Dixon acquisition is possible. CONCLUSION/CONCLUSIONS:-weighted imaging in a single scan. In addition to free-breathing abdominal examination, it promises value for clinical applications that are frequently affected by motion artifacts.

PURPOSE/OBJECTIVE:profiles in the longitudinal (z) direction and allow for next-nearest neighbor decoupling. METHODS:Two cylindrical 8-channel arrays having the same length and diameter, 1 of triangular coils and the other of rectangular coils, were constructed and compared in phantom imaging experiments using measures of excitation distribution for a variety of RF shim settings and geometry factor maps for different accelerations on different planes. RESULTS:Coupling between elements was -20 dB or better for all triangular coil pairs, but worse than -12 dB for several of the rectangular coil pairs. Both coils could produce adequate shims on a central transverse plane, but the same shim produced worse results off center for the triangular coil array than for the rectangular coil array. Compared to the rectangular coil array, the maximum geometry factor for the triangular coil array was reduced by a factor of 13.1 when using a 2-fold acceleration in the z-direction. CONCLUSION/CONCLUSIONS:profiles along the z-direction, although this also means that individual slices must be shimmed separately. This design is well suited for parallel transmit applications while also having high receive sensitivity.