Efficient pulse sequence design framework for high-dimensional MR fingerprinting scans using systematic error index.

PURPOSE: For effective optimization of MR fingerprinting (MRF) pulse sequences, estimating and minimizing errors from actual scan conditions are crucial. Although virtual-scan simulations offer an approximation to these errors, their computational demands become expensive for high-dimensional MRF frameworks, where interactions between more than two tissue properties are considered. This complexity makes sequence optimization impractical. We introduce a new mathematical model, the systematic error index (SEI), to address the scalability challenges for high-dimensional MRF sequence design. METHODS: By eliminating the need to perform dictionary matching, the SEI model approximates quantification errors with low computational costs. The SEI model was validated in comparison with virtual-scan simulations. The SEI model was further applied to optimize three high-dimensional MRF sequences that quantify two to four tissue properties. The optimized scans were examined in simulations and healthy subjects. RESULTS: The proposed SEI model closely approximated the virtual-scan simulation outcomes while achieving hundred- to thousand-times acceleration in the computational speed. In both simulation and in vivo experiments, the optimized MRF sequences yield higher measurement accuracy with fewer undersampling artifacts at shorter scan times than the heuristically designed sequences. CONCLUSION: We developed an efficient method for estimating real-world errors in MRF scans with high computational efficiency. Our results illustrate that the SEI model could approximate errors both qualitatively and quantitatively. We also proved the practicality of the SEI model of optimizing sequences for high-dimensional MRF frameworks with manageable computational power. The optimized high-dimensional MRF scans exhibited enhanced robustness against undersampling and system imperfections with faster scan times.

Quantifying 3D MR fingerprinting (3D-MRF) reproducibility across subjects, sessions, and scanners automatically using MNI atlases.

PURPOSE: Quantitative MRI techniques such as MR fingerprinting (MRF) promise more objective and comparable measurements of tissue properties at the point-of-care than weighted imaging. However, few direct cross-modal comparisons of MRF's repeatability and reproducibility versus weighted acquisitions have been performed. This work proposes a novel fully automated pipeline for quantitatively comparing cross-modal imaging performance in vivo via atlas-based sampling. METHODS: We acquire whole-brain 3D-MRF, turbo spin echo, and MPRAGE sequences three times each on two scanners across 10 subjects, for a total of 60 multimodal datasets. The proposed automated registration and analysis pipeline uses linear and nonlinear registration to align all qualitative and quantitative DICOM stacks to Montreal Neurological Institute (MNI) 152 space, then samples each dataset's native space through transformation inversion to compare performance within atlas regions across subjects, scanners, and repetitions. RESULTS: Voxel values within MRF-derived maps were found to be more repeatable (σT1 = 1.90, σT2 = 3.20) across sessions than vendor-reconstructed MPRAGE (σT1w = 6.04) or turbo spin echo (σT2w = 5.66) images. Additionally, MRF was found to be more reproducible across scanners (σT1 = 2.21, σT2 = 3.89) than either qualitative modality (σT1w = 7.84, σT2w = 7.76). Notably, differences between repeatability and reproducibility of in vivo MRF were insignificant, unlike the weighted images. CONCLUSION: MRF data from many sessions and scanners can potentially be treated as a single dataset for harmonized analysis or longitudinal comparisons without the additional regularization steps needed for qualitative modalities.

Automated design of pulse sequences for magnetic resonance fingerprinting using physics-inspired optimization.

Magnetic resonance fingerprinting (MRF) is a method to extract quantitative tissue properties such as [Formula: see text] and [Formula: see text] relaxation rates from arbitrary pulse sequences using conventional MRI hardware. MRF pulse sequences have thousands of tunable parameters, which can be chosen to maximize precision and minimize scan time. Here, we perform de novo automated design of MRF pulse sequences by applying physics-inspired optimization heuristics. Our experimental data suggest that systematic errors dominate over random errors in MRF scans under clinically relevant conditions of high undersampling. Thus, in contrast to prior optimization efforts, which focused on statistical error models, we use a cost function based on explicit first-principles simulation of systematic errors arising from Fourier undersampling and phase variation. The resulting pulse sequences display features qualitatively different from previously used MRF pulse sequences and achieve fourfold shorter scan time than prior human-designed sequences of equivalent precision in [Formula: see text] and [Formula: see text] Furthermore, the optimization algorithm has discovered the existence of MRF pulse sequences with intrinsic robustness against shading artifacts due to phase variation.

Novel 3D magnetic resonance fingerprinting radiomics in adult brain tumors: a feasibility study.

OBJECTIVES: To test the feasibility of using 3D MRF maps with radiomics analysis and machine learning in the characterization of adult brain intra-axial neoplasms. METHODS: 3D MRF acquisition was performed on 78 patients with newly diagnosed brain tumors including 33 glioblastomas (grade IV), 6 grade III gliomas, 12 grade II gliomas, and 27 patients with brain metastases. Regions of enhancing tumor, non-enhancing tumor, and peritumoral edema were segmented and radiomics analysis with gray-level co-occurrence matrices and gray-level run-length matrices was performed. Statistical analysis was performed to identify features capable of differentiating tumors based on type, grade, and isocitrate dehydrogenase (IDH1) status. Receiver operating curve analysis was performed and the area under the curve (AUC) was calculated for tumor classification and grading. For gliomas, Kaplan-Meier analysis for overall survival was performed using MRF T1 features from enhancing tumor region. RESULTS: Multiple MRF T1 and T2 features from enhancing tumor region were capable of differentiating glioblastomas from brain metastases. Although no differences were identified between grade 2 and grade 3 gliomas, differentiation between grade 2 and grade 4 gliomas as well as between grade 3 and grade 4 gliomas was achieved. MRF radiomics features were also able to differentiate IDH1 mutant from the wild-type gliomas. Radiomics T1 features for enhancing tumor region in gliomas correlated to overall survival (p < 0.05). CONCLUSION: Radiomics analysis of 3D MRF maps allows differentiating glioblastomas from metastases and is capable of differentiating glioblastomas from metastases and characterizing gliomas based on grade, IDH1 status, and survival. KEY POINTS: • 3D MRF data analysis using radiomics offers novel tissue characterization of brain tumors. • 3D MRF with radiomics offers glioma characterization based on grade, IDH1 status, and overall patient survival.

3D magnetic resonance fingerprinting with quadratic RF phase.

PURPOSE: To implement 3D magnetic resonance fingerprinting (MRF) with quadratic RF phase (qRF-MRF) for simultaneous quantification of T1 , T2 , ΔB0 , and T2∗ . METHODS: 3D MRF data with effective undersampling factor of 3 in the slice direction were acquired with quadratic RF phase patterns for T1 , T2 , and T2∗ sensitivity. Quadratic RF phase encodes the off-resonance by modulating the on-resonance frequency linearly in time. Transition to 3D brings practical limitations for reconstruction and dictionary matching because of increased data and dictionary sizes. Randomized singular value decomposition (rSVD)-based compression in time and reduction in dictionary size with a quadratic interpolation method are combined to be able to process prohibitively large data sets in feasible reconstruction and matching times. RESULTS: Accuracy of 3D qRF-MRF maps in various resolutions and orientations are compared to 3D fast imaging with steady-state precession (FISP) for T1 and T2 contrast and to 2D qRF-MRF for T2∗ contrast and ΔB0 . The precision of 3D qRF-MRF was 1.5-2 times higher than routine clinical scans. 3D qRF-MRF ΔB0 maps were further processed to highlight the susceptibility contrast. CONCLUSION: Natively co-registered 3D whole brain T1 , T2 , T2∗ , ΔB0 , and QSM maps can be acquired in as short as 5 min with 3D qRF-MRF.

Free-Breathing Abdominal Magnetic Resonance Fingerprinting Using a Pilot Tone Navigator.

BACKGROUND: Quantitative T1 and T2 mapping in the abdomen provides valuable information in tissue characterization but is technically challenging due to respiratory motions. The proposed technique integrates magnetic resonance fingerprinting (MRF) and pilot tone (PT) navigator with retrospective gating to provide simultaneous quantification of multiple tissue properties in a single acquisition without breath-holding or patient set-up. PURPOSE: To develop a free-breathing abdominal MRF technique for quantitative mapping in the abdomen. STUDY TYPE: Prospective. POPULATION: Twelve healthy volunteers. FIELD STRENGTH/SEQUENCE: A 3 T, two-dimensional (2D) and three-dimensional (3D) spiral MRF sequence with fast imaging with steady-state free precession (FISP) readout. ASSESSMENT: The PT navigator was compared to standard respiratory belt performance. The T1 and T2 values acquired using 2D and 3D MRF with and without PT were obtained in a phantom and compared to reference values. Digital phantom simulation was performed to evaluate PT MRF reconstruction with varying breathing patterns. In the in vivo studies, T1 and T2 values derived from PT 2D MRF were compared to 2D breath-hold MRF. T1 and T2 values derived from PT 3D MRF were compared to published values. STATISTICAL TESTS: Principal component analysis (PCA), linear regression, relative error, Pearson correlation, paired Student's t-test, Bland-Altman Analysis. RESULTS: The phantom study showed PT MRF T1 values had a mean difference of 0.2% ± 0.1%, and T2 values had a mean difference of 0.1% ± 0.4% when compared to no-PT MRF values. The digital phantom experiment suggested the T1 and T2 maps at both end-exhalation and end-inhalation states resemble the corresponding ground-truth maps. DATA CONCLUSION: The phantom study showed good agreement between MRF T1 and T2 values and with reference values. In vivo studies demonstrated that 2D and 3D quantitative imaging in the abdomen could be achieved with integration of PT navigation with MRF reconstruction using retrospective gating of respiratory motion. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 1.

Magnetic resonance fingerprinting review part 2: Technique and directions.

Magnetic resonance fingerprinting (MRF) is a general framework to quantify multiple MR-sensitive tissue properties with a single acquisition. There have been numerous advances in MRF in the years since its inception. In this work we highlight some of the recent technical developments in MRF, focusing on sequence optimization, modifications for reconstruction and pattern matching, new methods for partial volume analysis, and applications of machine and deep learning. Level of Evidence: 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2020;51:993-1007.

Multiband echo-shifted echo planar imaging.

PURPOSE: To propose the technique multiband echo-shifted (MESH) echo planar imaging (EPI), which combines the principles of echo-shifted acquisition for two-dimensional multislice EPI, with both in-plane and multiband acceleration by means of partial parallel imaging techniques. METHODS: MESH EPI is suitable for functional MRI (fMRI) in situations where there is sufficient time to insert an additional EPI readout in the dead time between slice selection and the standard EPI readout. In such situations, MESH EPI can further accelerate data acquisition compared with standard multiband techniques. The method is particularly well suited for low static magnetic field strengths and lower spatial resolutions. We compared MESH with multiband and standard EPI with temporal signal-to-noise ratio (tSNR) measurements and resting state fMRI data. RESULTS: Results obtained at 1.5 T from healthy subjects revealed that the additional gradient switching did not additionally affect time course SNR over and above the reduction inherent to multiband imaging. Functional results were qualitatively similar between methods. MESH was not affected by the tSNR reduction and echo shifting gradients. The MESH data were acquired at a factor 2 or 3 faster than corresponding multiband acquisitions for echo shift factors of 1 and 2, respectively. CONCLUSION: MESH can offer further acceleration of image acquisition for fMRI at no loss in sensitivity. Magn Reson Med 77:1981-1986, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Multiband multislab 3D time-of-flight magnetic resonance angiography for reduced acquisition time and improved sensitivity.

PURPOSE: To explore the use of multiband (MB) imaging in multislab (MS) 3D time-of-flight-magnetic resonance angiography (TOF-MRA) and to improve acquisition time efficiency (TA), inflow contrast and sensitivity in vessel detection. THEORY AND METHODS: TOF-MRA is commonly used for imaging intracranial vessels. A MB-MS 3D-TOF-MRA sequence was implemented to excite and acquire multiple slabs simultaneously. Controlled aliasing in parallel imaging results in higher acceleration was used in addition to improve the quality of image reconstruction. Compared to a standard protocol which acquired three slabs in total the MB-MS protocol reduced the thickness by 3 while simultaneously acquiring data from 3 slabs. The total TA was also reduced by a factor 3. RESULTS: This technique maintains contrast-to-noise ratio while reducing TA, compared to standard single-band/MOTSA acquisitions, leading to an increase in CNR/TA of 1.65 compared to the standard protocol. Furthermore, the strong inflow contrast and increased magnetization transfer contrast caused by the MB excitation pulses improves the sharpness of the vessel borders which is reflected by a 5% higher full width at half maximum of the vessel size and a 17% higher slope of the vessel borders compared to the standard single-band acquisition. CONCLUSION: MB-MS 3D-TOF-MRA can appreciably accelerate image acquisition and combines the high spatial resolution of 3D imaging with the additional inflow contrast advantage of thinner slab acquisitions without introducing excessive noise arising from the MB reconstruction.

Improved sensitivity and specificity for resting state and task fMRI with multiband multi-echo EPI compared to multi-echo EPI at 7 T.

A multiband multi-echo (MBME) sequence is implemented and compared to a matched standard multi-echo (ME) protocol to investigate the potential improvement in sensitivity and spatial specificity at 7 T for resting state and task fMRI. ME acquisition is attractive because BOLD sensitivity is less affected by variation in T2*, and because of the potential for separating BOLD and non-BOLD signal components. MBME further reduces TR thus increasing the potential reduction in physiological noise. In this study we used FSL-FIX to clean ME and MBME resting state and task fMRI data (both 3.5mm isotropic). After noise correction, the detection of resting state networks improves with more non-artifactual independent components being observed. Additional activation clusters for task data are discovered for MBME data (increased sensitivity) whereas existing clusters become more localized for resting state (improved spatial specificity). The results obtained indicate that MBME is superior to ME at high field strengths.

Whole brain, high resolution multiband spin-echo EPI fMRI at 7 T: a comparison with gradient-echo EPI using a color-word Stroop task.

A whole brain, multiband spin-echo (SE) echo planar imaging (EPI) sequence employing a high spatial (1.5 mm isotropic) and temporal (TR of 2 s) resolution was implemented at 7 T. Its overall performance (tSNR, sensitivity and CNR) was assessed and compared to a geometrically matched gradient-echo (GE) EPI multiband sequence (TR of 1.4 s) using a color-word Stroop task. PINS RF pulses were used for refocusing to reduce RF amplitude requirements and SAR, summed and phase-optimized standard pulses were used for excitation enabling a transverse or oblique slice orientation. The distortions were minimized with the use of parallel imaging in the phase encoding direction and a post-acquisition distortion correction. In general, GE-EPI shows higher efficiency and higher CNR in most brain areas except in some parts of the visual cortex and superior frontal pole at both the group and individual-subject levels. Gradient-echo EPI was able to detect robust activation near the air/tissue interfaces such as the orbito-frontal and subcortical regions due to reduced intra-voxel dephasing because of the thin slices used and high in-plane resolution.

Application of PINS radiofrequency pulses to reduce power deposition in RARE/turbo spin echo imaging of the human head.

PURPOSE: To explore the use of PINS radiofrequency (RF) pulses to reduce RF power deposition in multiband/simultaneous multislice imaging with the RARE/turbo spin echo (TSE) sequence at 3T and 7T. METHODS: A PINS-TSE sequence was implemented and combined with blipped CAIPI to improve the reconstruction of superposed slices. Whole brain imaging of healthy volunteers was performed at both 3T and 7T using a 32-channel coil for signal reception. RESULTS: A considerable reduction in power deposition was achieved compared with a standard sequence of the manufacturer. At 3T, the reduction in specific absorption rate (SAR) made short pulse repetition times (TRs) possible, however, in order to obtain a good T2 contrast, it is advisable to maintain TR while extending the echo train length. At 7T, whole brain coverage with a spatial resolution of 1 × 1 × 2 mm(3) was achieved in an acquisition time of 150 s. Furthermore, it could be shown that pulse sequences that use PINS pulses do not suffer from having additional magnetization transfer contrast. CONCLUSION: PINS RF pulses combined with multiband imaging reduce SAR sufficiently to enable routine TSE imaging at 7T within clinically acceptable acquisition times. In general, the combination of multiband imaging with PINS RF pulses represents a method to reduce total RF power deposition.

Generalized INverse imaging (GIN): ultrafast fMRI with physiological noise correction.

An ultrafast functional magnetic resonance imaging (fMRI) technique, called generalized inverse imaging (GIN), is proposed, which combines inverse imaging with a phase constraint-leading to a less underdetermined reconstruction-and physiological noise correction. A single 3D echo planar imaging (EPI) prescan is sufficient to obtain the necessary coil sensitivity information and reference images that are used to reconstruct standard images, so that standard analysis methods are applicable. A moving dots stimulus paradigm was chosen to assess the performance of GIN. We find that the spatial localization of activation for GIN is comparable to an EPI protocol and that maximum z-scores increase significantly. The high temporal resolution of GIN (50 ms) and the acquisition of the phase information enable unaliased sampling and regression of physiological signals. Using the phase time courses obtained from the 32 channels of the receiver coils as nuisance regressors in a general linear model results in significant improvement of the functional activation, rendering the acquisition of external physiological signals unnecessary. The proposed physiological noise correction can in principle be used for other fMRI protocols, such as simultaneous multislice acquisitions, which acquire the phase information sufficiently fast and sample physiological signals unaliased.

Simultaneous multislice inversion contrast imaging using power independent of the number of slices (PINS) and delays alternating with nutation for tailored excitation (DANTE) radio frequency pulses.

A method for simultaneous multislice (SMS) inversion contrast imaging is presented using a combination of the delays alternating with nutation for tailored excitation (DANTE) and the power independent of the number of slices (PINS) techniques. In SMS imaging, simultaneously excited slices result in an aliased image that is disentangled using parallel imaging reconstruction techniques. At high-magnetic field strengths, the peak amplitude and specific absorption rate of conventional (summed) SMS radio frequency pulses can be prohibitively high. Using the PINS approach, specific absorption rate is independent of the number of slices allowing high SMS acceleration factors even at high fields. Using DANTE, adiabatic SMS radio frequency pulses can be created to be combined with PINS. This allows 2D imaging protocols that employ adiabatic pulses to also reap the benefits of low specific absorption rate SMS acceleration. As a proof-of-concept, simulations and measurements using hyperbolic secant inversion pulses are shown.

A fast MR fingerprinting simulator for direct error estimation and sequence optimization.

Magnetic resonance fingerprinting (MRF) is a novel quantitative MR technique that simultaneously provides multiple tissue property maps. When optimizing MRF scans, modeling undersampling errors and field imperfections in cost functions for direct measurement of quantitative errors will make the optimization results more practical and robust. However, optimizing such cost function is computationally expensive and impractical for MRF optimization with tens of thousands of iterations. Here, we introduce a fast MRF simulator to simulate aliased images from actual scan scenarios including undersampling and system imperfections, which substantially reduces computational time and allows for direct error estimation of the quantitative maps and efficient sequence optimization. We evaluate the performance and computational speed of the proposed approach by simulations and in vivo experiments. The simulations from the proposed method closely approximate the signals and MRF maps from in vivo scans, with 158 times shorter processing time than the conventional simulation method using Non-uniform Fourier transform. We also demonstrate the power of applying the fast MRF simulator in MRF sequence optimization. The optimized sequences are validated with in vivo scans to assess the image quality and accuracy. The optimized sequences produce artifact-free T1 and T2 maps in 2D and 3D scans with equivalent mapping accuracy as the human-designed sequence but at shorter scan times. Incorporating the proposed simulator in the MRF optimization framework makes direct estimation of undersampling errors during the optimization process feasible, and provide optimized MRF sequences that are robust against undersampling artifacts and field inhomogeneity.

Whole brain, high resolution spin-echo resting state fMRI using PINS multiplexing at 7 T.

This article demonstrates the application of spin-echo EPI for resting state fMRI at 7 T. A short repetition time of 1860 ms was made possible by the use of slice multiplexing which permitted whole brain coverage at high spatial resolution (84 slices of 1.6 mm thickness). Radiofrequency power deposition was kept within regulatory limits by use of the power independent of number of slices (PINS) technique. A high in-plane spatial resolution of 1.5 mm was obtained, while image distortion was ameliorated by the use of in-plane parallel imaging techniques. Data from six subjects were obtained with a measurement time of just over 15 min per subject. A group level independent component (IC) analysis revealed 24 non-artefactual resting state networks, including those commonly found in standard acquisitions, as well as plausible networks for a broad range of regions. Signal was measured from regions commonly rendered inaccessible due to signal voids in gradient echo acquisitions. Dual regression was used to obtain spatial IC maps at the single subject level revealing exquisite localisation to grey matter that is consistent with a high degree of T(2)-weighting in the acquisition sequence. This technique hence holds great promise for both resting state and activation studies at 7 T.

Power Independent of Number of Slices (PINS) radiofrequency pulses for low-power simultaneous multislice excitation.

This communication describes radiofrequency pulses capable of performing spatially periodic excitation, inversion, and refocusing. The generation of such pulses either by multiplication of existing radiofrequency pulses by a Dirac comb function or by means of Fourier series expansion is described. Practical schemes for the implementation of such pulses are given, and strategies for optimizing the pulse profile at fixed pulse duration are outlined. The pulses are implemented using a spin-echo sequence. The power deposition is independent of the number of slices acquired, and hence the power deposition per slice is considerably reduced compared to multislice imaging. Excellent image quality is obtained both in phantoms and in images of the human head. These pulses should find widespread application for multiplexed imaging, in particular at high static magnetic field strengths and for pulse sequences that have a high radiofrequency power deposition and could lead to dramatic increases in scanning efficiency.

An Investigation of RSN Frequency Spectra Using Ultra-Fast Generalized Inverse Imaging.

With the advancements in MRI hardware, pulse sequences and reconstruction techniques, many low TR sequences are becoming more and more popular within the functional MRI (fMRI) community. In this study, we have investigated the spectral characteristics of resting state networks (RSNs) with a newly introduced ultra fast fMRI technique, called generalized inverse imaging (GIN). The high temporal resolution of GIN (TR = 50 ms) enables to sample cardiac signals without aliasing into a separate frequency band from the BOLD fluctuations. Respiration related signal changes are, on the other hand, removed from the data without the need for external physiological recordings. We have observed that the variance over the subjects is higher than the variance over RSNs.