Large FOV Research Articles

NEMA-2008 and In-Vivo Animal and Plant Imaging Performance of the Large FOV Preclinical Digital PET/CT System Discoverist 180

The RAYCAN Discoverist 180 (RAYCAN D180) is a novel preclinical positron emission tomography (PET) and computed tomography (CT) integrated system for the imaging of small- and medium-sized animals and plants. We measured a system resolution of (2.56±0.04) mm (radial), (2.46±0.05) mm (transverse), and (1.67 ± 0.03) mm (axial) at the center of the field of view using a filtered backprojection reconstruction algorithm. An OSEM-PSF-3D reconstruction algorithm improves the spatial resolution to (1.34±0.04) mm (radial), (1.35±0.05) mm (transverse), and (1.64±0.04) mm (axial). The peak noise equivalent count rate of the system is (713 ± 2) kcps, (207 ± 2) kcps, and (47 ± 1) kcps reached at approximately 97 MBq for the mouse-, rat- and monkey-like phantoms, respectively. This result is approximately $2\times $ higher and is reached at approximately $3\times $ higher activity in comparison with most of the existing commercial and research scanners. The peak absolute system sensitivity is (4.00±0.02)% and is competitive with state-of-the-art preclinical PET scanners. We found an image uniformity of (4.86±0.04)%. Finally, we report the static PET/CT scan of a mouse and of a leaf of Epipremnum aureum, and the time-activity curves of the [18F]-FDG metabolism in mouse left myocardial ventricle and kidneys obtained with a PET/CT scan in dynamic modality.

Multi-spin echo T2 relaxation imaging with compressed sensing (METRICS) for rapid myelin water imaging.

Myelin water imaging (MWI) provides a valuable biomarker for myelin, but clinical application has been restricted by long acquisition times. Accelerating the standard multi-echo T2 acquisition with gradient echoes (GRASE) or by 2D multi-slice data collection results in image blurring, contrast changes, and other issues. Compressed sensing (CS) can vastly accelerate conventional MRI. In this work, we assessed the use of CS for in vivo human MWI, using a 3D multi spin-echo sequence. We implemented multi-echo T2 relaxation imaging with compressed sensing (METRICS) and METRICS with partial Fourier acceleration (METRICS-PF). Scan-rescan data were acquired from 12 healthy controls for assessment of repeatability. MWI data were acquired for METRICS in 9m:58s and for METRICS-PF in 7m:25s, both with 1.5×2×3mm3 voxels, 56 echoes, 7ms ΔTE, and 240×240×170mm3 FOV. METRICS was compared with a novel multi-echo spin-echo gold-standard (MSE-GS) MWI acquisition, acquired for a single additional subject in 2h:2m:40s. METRICS/METRICS-PF myelin water fraction had mean: repeatability coefficient 1.5/1.1, coefficient of variation 6.2/4.5%, and intra-class correlation coefficient 0.79/0.84. Repeatability metrics comparing METRICS with METRICS-PF were similar, and both sequences agreed with reference values from literature. METRICS images and quantitative maps showed excellent qualitative agreement with those of MSE-GS. METRICS and METRICS-PF provided highly repeatable MWI data without the inherent disadvantages of GRASE or 2D multi-slice acquisition. CS acceleration allows MWI data to be acquired rapidly with larger FOV, higher estimated SNR, more isotropic voxels and more echoes than with previous techniques. The approach introduced here generalizes to any multi-component T2 mapping application.

Low-level fat fraction quantification at 3 T: comparative study of different tools for water–fat reconstruction and MR spectroscopy

Chemical Shift Encoded Magnetic Resonance Imaging (CSE-MRI)-based quantification of low-level (< 5% of proton density fat fraction-PDFF) fat infiltration requires highly accurate data reconstruction for the assessment of hepatic or pancreatic fat accumulation in diagnostics and biomedical research. We compare three software tools available for water/fat image reconstruction and PDFF quantification with MRS as the reference method. Based on the algorithm exploited in the tested software, the accuracy of fat fraction quantification varies. We evaluate them in phantom and in vivo MRS and MRI measurements. The signal model of Intralipid 20% emulsion used for phantoms was established for 3T and 9.4T fields. In all cases, we noticed a high coefficient of determination (R-squared) between MRS and MRI-PDFF measurements: in phantoms <0.9924-0.9990>; and in vivo <0.8069-0.9552>. Bland-Altman analysis was applied to phantom and in vivo measurements. Multi-echo MRI in combination with an advanced algorithm including multi-peak spectrum modeling appears as a valuable and accurate method for low-level PDFF quantification over large FOV in high resolution, and is much faster than MRS methods. The graph-cut algorithm (GC) showed the fewest water/fat swaps in the PDFF maps, and hence stands out as the most robust method of those tested.

Raster adaptive optics for video rate aberration correction and large FOV multiphoton imaging.

Removal of complex aberrations at millisecond time scales over millimeters in distance in multiphoton laser scanning microscopy limits the total spatiotemporal imaging throughput for deep tissue imaging. Using a single low resolution deformable mirror and time multiplexing (TM) adaptive optics, we demonstrate video rate aberration correction (5 ms update rate for a single wavefront mask) for a complex heterogeneous distribution of refractive index differences through a depth of up to 1.1 mm and an extended imaging FOV of up to 0.8 mm, with up to 167% recovery of fluorescence intensity 335 µm from the center of the FOV. The proposed approach, termed raster adaptive optics (RAO), integrates image-based aberration retrieval and video rate removal of arbitrarily defined regions of dominant, spatially varied wavefronts. The extended FOV was achieved by demonstrating rapid recovery of up to 50 distinct wavefront masks at 500 ms update rates that increased imaging throughput by 2.3-fold. Because RAO only requires a single deformable mirror with image-based aberration retrieval, it can be directly implemented on a standard laser scanning multiphoton microscope.

Three-dimensional accelerated acquisition for hyperpolarized 13 C MR with blipped stack-of-spirals and conjugate-gradient SENSE.

To test a new parallel imaging strategy for acceleration of hyperpolarized 13 C MR acquisitions based on a 3D blipped stack-of-spirals trajectory and conjugate-gradient SENSE reconstruction with precalibrated sensitivities. The blipped stack-of-spirals trajectory was developed for an acceleration factor of 4, based on an undersampled stack-of-spirals with gradient blips during spiral readout. The trajectory was developed with volumetric coverage of a large FOV and with high spatial resolution. High temporal resolution was attained through spectral-spatial excitation and 4 excitations per volume. The blipped stack-of-spirals was evaluated in simulations and phantom experiments. Next, the method was evaluated for kidney and cardiac imaging in 2 separate healthy pigs. Simulation and phantom results showed successful acquisition and reconstruction, but also revealed reconstruction challenges for certain locations and for wide signal sources. For the kidney experiment, the accelerated acquisition showed high similarity to 2 separately acquired fully sampled data sets with matched spatial and temporal resolution, respectively. For the cardiac experiment, the accelerated acquisition proved able to map each metabolite in 3 dimensions within a single cardiac cycle. The proposed method demonstrated effective mapping of metabolism in both kidneys and the heart of healthy pigs. Limitations seen in phantom experiments, may be irrelevant for most clinical applications, but should be kept in mind as well as reconstruction challenges related to residual aliasing. All in all, we show that the blipped stack-of-spirals is a relevant parallel imaging method for hyperpolarized human imaging, facilitating better insights into metabolism compared with nonaccelerated acquisition.

Feasibility of 31 P spectroscopic imaging at 7 T in lung carcinoma patients.

Currently, it is difficult to predict effective therapy response to molecular therapies for the treatment of lung cancer based solely on anatomical images. 31P MR spectroscopic imaging could provide as a non‐invasive method to monitor potential biomarkers for early therapy evaluation, a necessity to improve personalized care and reduce cost. However, surface coils limit the imaging volume in conventional 31P MRSI. High‐energetic adiabatic RF pulses are required to achieve flip angle homogeneity but lead to high SAR. Birdcage coils permit use of conventional amplitude modulated pulses, even over large FOV, potentially decreasing overall SAR massively. Here, we investigate the feasibility of 3D 31P MRSI at 7 T in lung carcinoma patients using an integrated 31P birdcage body coil in combination with either a dual‐coil or a 16‐channel receiver.Simulations showed a maximum decrease in SNR per unit of time of 8% for flip angle deviations in short TR low flip‐angle excitation 3D CSI. The minimal SNR loss allowed for fast 3D CSI without time‐consuming calibration steps (>10:00 min.). 31P spectra from four lung carcinoma patients were acquired within 29:00 minutes and with high SNR using both receivers. The latter allowed discrimination of individual phosphodiesters, inorganic phosphate, phosphocreatine and ATP. The receiver array allowed for an increased FOV compared to the dual‐coil receiver.3D 31P‐CSI were acquired successfully in four lung carcinoma patients using the integrated 31P body coil at ultra‐high field. The increased spectral resolution at 7 T allowed differentiation of multiple 31P metabolites related to phospholipid and energy metabolism. Simulations provide motivation to exclude 31P B1 calibrations, potentially decreasing total scan duration. Employing large receiver arrays improves the field of view allowing for full organ coverage. 31P MRSI is feasible in lung carcinoma patients and has potential as a non‐invasive method for monitoring personalized therapy response in lung tumors.

Resolution enhancement of near-eye displays by overlapping images

It has always been a hot topic to break the angular resolution/field-of-view (FOV) invariant and achieve a near-eye display with a large FOV and high-resolution. We present a super-resolution method for near-eye displays by overlapping images from two display panels using a half mirror. Key aspects including theoretical analysis, optimized method, system set-up, and calibration process are discussed detailly. A proof-of-concept prototype was developed to provide a super-resolution image capable of significantly higher sampling rate.

Controlling readability of head-fixed large field-of-view displays

Background. The information displayed on a Head Mounted Display (HMD) can only be read by making eye movements, since head movements have no effect on the ocular image position. Aniseikonia (a common visual deficit) is expected to cause eye strain and limit the readability of large FoV binocular HMDs. As the FoV increases, the screen layout needs to optimize overall display readability by preventing clutter while taking common optometric conditions into account. Methods. We measured the ability to quickly determine the orientation of a target ⊤ (⊤ vs ⊥) surrounded by 4 randomly oriented (up, down, left, right) flanker T’s as a function of target-flanker spacing and eccentricity, in conditions where the target had either the same or opposite opposite luminance polarity as the flankers. All 12 subjects scored normal on relevant optometric tests (stereopsis, visual acuity, Awaya aniseikonia test, phoria). An aniseikonic lens placed in front of one eye optically enlarged the image by 2½%, simulating a common optometric condition. The additional delay caused by the presence of the four flankers is adopted as the ‘Crowding component’ of the reaction time. Results. Compared to the Same polarity condition, Opposite polarity reduced the Crowding time by a factor of 2.3 (p< 0.001). The Crowding times can be described as an extension of Fitts’ law. Unexpectedly, the mild aniseikonia condition doubled the Crowding time (p< 0.001) and caused the highest level of eye strain (p< 0.001). Conclusion. For all eccentricities and target-flanker spacings, the Crowding time more than halved in the opposite polarity condition, while it doubled due to the addition of just 2½% aniseikonia. Practical implications. Even users with mild aniseikonia are likely to experience problems while reading a large FOV HMD. ‘Polarity decluttering’ can significantly enhance symbology legibility.

3D nonrigid motion correction for quantitative assessment of hepatic lesions in DCE‐MRI

PurposeTo provide nonrigid respiratory motion‐corrected DCE‐MRI images with isotropic resolution of 1.5 mm, full coverage of abdomen, and covering the entire uptake curve with a temporal resolution of 6 seconds, for the quantitative assessment of hepatic lesions.Methods3D DCE‐MRI data were acquired at 3 T during free breathing for 5 minutes using a 3D T1‐weighted golden‐angle radial phase‐encoding sequence. Nonrigid respiratory motion information was extracted and used in motion‐corrected image reconstruction to obtain high‐quality DCE‐MRI images with temporal resolution of 6 seconds and isotropic resolution of 1.5 mm. An extended Tofts model was fitted to the dynamic data sets, yielding quantitative parametric maps of endothelial permeability using the hepatic artery as input function. The proposed approach was evaluated in 11 patients (52 ± 17 years, 5 men) with and without known hepatic lesions, undergoing DCE‐MRI.ResultsRespiratory motion produced artifacts and misalignment between dynamic volumes (lesion average motion amplitude of 3.82 ± 1.11 mm). Motion correction minimized artifacts and improved average contrast‐to‐noise ratio of hepatic lesions in late phase by 47% (p < .01). Quantitative endothelial permeability maps of motion‐corrected data demonstrated enhanced visibility of different pathologies (e.g., metastases, hemangiomas, cysts, necrotic tumor substructure) and showed improved contrast‐to‐noise ratio by 62% (p < .01) compared with uncorrected data.Conclusion3D nonrigid motion correction in DCE‐MRI improves both visual and quantitative assessment of hepatic lesions by ensuring accurate alignment between 3D DCE images and reducing motion blurring. This approach does not require breath‐holds and minimizes scan planning by using a large FOV with isotropic resolution.

Influence of Small, Midi, Medium and Large Fields of View on Accuracy of Linear Measurements in CBCT Imaging: Diagnostic Accuracy Study.

AIM:This study aimed to assess the effect of changing the field of view on the dimensional accuracy of CBCT imaging.METHODS:The implant-bone models were randomly numbered from 1 to 13 by the principal researcher, and then on each model at the incisors region three positions were selected and marked on the model with a permanent blue marker. Then at each marked position three radio-opaque ‘RO’ markers “gutta-percha pieces” were glued on the model surfaces as following; two pieces on the facial surface one occlusally (at the alveolar crest) and one apically (at the inferior border of the model) both were on the same vertical line and perpendicular to the horizontal plane, while the third one was placed on the lingual surface opposing the occlusally placed buccal piece. CBCT examinations of each bone model were performed using Cranex3Dx CBCT (Helsinki, Finland) machine. Each model was scanned four times with standardised tube current and voltage of 12.5 mA and 90 kVp respectively at four different FOVs. The FOVs used were as following: Small FOV: 50 x 50 mm with voxel size 200 µm, Midi FOV: 61 x 78 mm with voxel size 300 µm, Medium FOV: 78 x 78 mm with voxel size 300 µm, Large FOV: 78 x 150 mm with voxel size 350 µm. The reference standard in this study was the real linear measurements that were obtained directly on the implant-bone models using high precision sliding electronic digital calliper with 0-150 mm internal and external measuring range and 0.01 mm resolution accuracy. The index test in the current study was the CBCT linear measurements obtained from CBCT images of implant-bone models using small, midi, medium and large FOVs.RESULTS:The results of this study showed that both medium and large FOVs showed a statistically significant difference, which could be translated into clinical relevance only in thickness measurements.CONCLUSION:The interpretation of these results leads to the assumption that increasing the FOV size together with voxel size could adversely affect the accuracy of CBCT linear measurements, especially when small distances are to be assessed.

Image quality optimization of narrow detector dental computed tomography for paediatric patients.

Dental CBCT exposure parameters should be optimized according to patient-specific indications, mainly for children that are most vulnerable to harmful effects of ionizing radiation. The aim of this study was to determine optimized kV settings for paediatric acquisitions for a dental CBCT device. Clinical and quantitative evaluations of image quality were performed using 5 and 10 years old (y/o) anthropomorphic phantoms. Technical evaluation was performed with the SEDENTEXCT-IQ phantom. Images were obtained using a PaX-i3D Green CBCT (Vatech, Korea) device, combining tube voltages ranging from 85 to 110 kV and 2 fields of view (FOVs: 21 × 19 and 12 × 9 cm), while maintaining the radiation dose fixed by adjusting the mA accordingly. Clinically, observers assessed images based on overall quality, sharpness, contrast, artefacts, and noise. For quantitative evaluation, mean grey value shift, % increase standard deviation, % beam-hardening and contrast-to-noise ratio were calculated. For technical evaluation, segmentation accuracy, contrast-to-noise ratio and full width at half maximum were measured. Biplot graphs were used to choose representative parameters, from which the best kV was selected for each protocol and evaluation. kV values that had no statistical differences (p > 0.05) with the best kV chosen were considered as having the same quality. Clinically, 95 kV was found as a cut-off value. From the quantitative aspect, 85 kV (p < 0.05) showed the worst quality, except in 12 × 9 cm 5 y/o. Technically, 85 and 110 kV in the large FOV showed significantly worse quality for the large FOV. For paediatric indications, 95 kV or higher (and correspondingly low mA values) was found as optimal.

Search for tau neutrinos with the MAGIC telescopes: improving selection criteria*

MAGIC, a system of two Cherenkov telescopes located at the Roque de los Muchachos Observatory (2200 a.s.l.) in the Canary Island of La Palma, has lately been engaged in an unconventional task: the search for a signature of particle showers induced by earth-skimming cosmic tau neutrinos arising from the ocean, in the PeV to EeV energy range. When pointing at the sea, the MAGIC telescopes can collect data in a range of about 5 deg in zenith and 80 deg in azimuth: the analysis of the shower images from ~30 hours of data, together with the simulations of upward-going tau neutrino showers, shows that the air showers induced by tau neutrinos can be discriminated from the hadronic background coming from a similar direction. We have calculated the point source acceptance and the expected event rates, assuming an incoming v flux consistent with IceCube measurements, and for a sample of generic neutrino fluxes from photohadronic interactions in AGNs and GRBs. A 90% C.L. upper limit on the tau-neutrino point source flux of 2.0 × 10–4 GeV cm–2 s–1 has been obtained. The presented results can also be important for future Cherenkov experiments such as the Cherenkov Telescope Array. This next generation ground-based observatory can have a much better possibility to detect v, given its larger FOV and much larger effective area.

Design of a reduced objective Lens fluorescence dPCR Gene chip detection system with high-throughput and large field of view

Achieving large FOV imaging and high resolution simultaneously for DNA detection poses an enormous conflict in the fluorescence microscope design field. As a result, this study sought to fill this gap by designing a system that is capable of producing the same quality imaging but of a simple, wieldy, cost-effective and unsophisticated nature. Therefore, we designed a dPCR gene chip detection system with high throughput, large field precision and reduced objective lens. This detection system can directly be used for imaging the fluorescence gene chip of 28 mm × 18 mm containing more than 20,000 channels. Three different kinds of fluorescence: FAM, HEX, and ROX can be used to capture images with a CMOS camera. The designed objective lens has a numerical aperture of 0.1 and a size reduction ratio of 1/2, capable of 20 mm working distance from objects, and with modulation transfer function (MTF) of more than 0.5 at 70lp/mm. Commercial optical design software ZEMAX was used to design the objective lens and control the aberrations by optimization operands. The Tolerance distribution and Monte Carlo tolerance analyses showed the design to be suitable for fabrication. According to the analyses, the detection system with the high-resolution ability of less than 10 μm (about 6um), thus, having low aberration and larger imaging scale is appropriate for DNA imaging. The designed gene fluorescent chip is expected to be adopted and utilized for the one-step imaging detection in clinical and bio application functions.

Design of a high-performance in-coupling grating using differential evolution algorithm for waveguide display.

Illuminance nonuniformity caused by natural vignetting can seriously affect the display quality of large-field-of-view (FOV) waveguide displays. In this paper, an optimization method based on the differential evolution algorithm is proposed for in-coupling grating design to improve coupling efficiency and compensate for natural vignetting. The in-coupling grating parameters are optimized to achieve efficiency distributions in which efficiency increases continuously with incidence angle, realizing uniform illuminance over a large FOV of 45°. The angular uniformity reaches 0.89. Additionally, average diffraction efficiency reaches 89.13% for transverse-electric polarization at 532 nm and 76% in the wavelength region between 450 and 700 nm.

P‐98: Incorporating Space‐variant Holographic Grating in Waveguide Display

Holographic waveguide display (HWD) is a promising solution supporting augmented reality, that can be applied in a variety of fields. Despite its advantage of lightweight and compact form factor, designing a HWD with large FOV and high efficiency simultaneously is still challenging. In this work, we revisit this FOV compromise issue and propose a HWD design incorporating space‐variant volume holographic gratings. The resulted prototype display achieves 20° FOV and 31.9% system efficiency.

P‐94: Free‐form Micro‐optical Design for Enhancing Image Quality (MTF) at Large FOV in Light Field Near Eye Display

Considering commercial virtual reality device now, most of them have the problem of visual fatigue which caused by accommodation and convergence (AC) conflict. Also, the bulky device makes user uncomfortable experience. This study aim to enhance the image quality by free‐form micro‐optical design based on light field technique using in near eye system. The results get the apparent improvement and show the importance of optical element in the light field system. Meanwhile, the depth perception of light field technique has been verified in the results.

A EUSO-like experiment as a precursor of a ultra-high energy neutrino Space Observatory

In last decade the technology for light mass, deployable and active mirrors for space applications reached a high Technological Readiness Level (TRL), and it is now ripe to be applied to an EUSO-like experiment for observing from space the particle showers produced by ultra-high energy charged CRs and neutrinos in the terrestrial atmosphere. The instrument should compromise between two conflicting requirements to the optical systems: (a) a wide as possible FoV for maximizing the volume of atmosphere observed from an Earth orbit in order to maximize the reachable energy of observed CRs; (b) a wide as possible pupil for maximizing the collected light for reaching down an enough low energy threshold for overlapping with the observations of ground based observatories. The high TRL reached in the realization of light mass, deployable active mirrors promises to improve the performance of the present EUSO project while reaching down an energy threshold of a few tens EeV. Therefore a new EUSO-like experiment should likely allows to observe some cosmogenic neutrino events, their number depending from the distribution and properties of the sources. Besides its main mission of observing ultra high energy CRs (UHECR), it can be considered a precursor or path-finder for addressing the problems to be afforded for the realization of an Observatory of ultra-high energy neutrinos, such as the realization of very large pupil and large FoV optical systems, the optimization of their number in a possible neutrino observatory, and the analysis machinery for identifying the neutrino events in the flood of the charged CR generated events.

A multi-wavelength study of the evolution of early-type galaxies in groups: the ultraviolet view

ABRIDGED- The UV-optical color magnitude diagram (CMD) of rich galaxy groups is characterised by a well developed Red Sequence (RS), a Blue Cloud (BC) and the so-called Green Valley (GV). Loose, less evolved groups of galaxies likely not virialized yet may lack a well defined RS. This is actually explained in the framework of galaxy evolution. We are focussing on understanding galaxy migration towards the RS, checking for signatures of such a transition in their photometric and morphological properties. We report on the UV properties of a sample of ETGs galaxies inhabiting the RS. The analysis of their structures, as derived by fitting a Sersic law to their UV luminosity profiles, suggests the presence of an underlying disk. This is the hallmark of dissipation processes that still must have a role in the evolution of this class of galaxies. SPH simulations with chemo-photometric implementations able to match the global properties of our targets are used to derive their evolutionary paths through UV-optical CDM, providing some fundamental information such as the crossing time through the GV, which depends on their luminosity. The transition from the BC to the RS takes several Gyrs, being about 3-5 Gyr for the the brightest galaxies and more long for fainter ones, if it occurs. The photometric study of nearby galaxy structures in UV is seriously hampered by either the limited FoV of the cameras (e.g in HST) or by the low spatial resolution of the images (e.g in the GALEX). Current missions equipped with telescopes and cameras sensitive to UV wavelengths, such as Swift-UVOT and Astrosat-UVIT, provide a relatively large FoV and better resolution than the GALEX. More powerful UV instruments (size, resolution and FoV) are obviously bound to yield fundamental advances in the accuracy and depth of the surface photometry and in the characterisation of the galaxy environment.

GBVS360, BMS360, ProSal: Extending existing saliency prediction models from 2D to omnidirectional images

In this paper it is studied how existing visual saliency models designed for 2D images can be extended and applied to Omnidirectional images in the equirectangular format. Three different frameworks, BMS360, GBVS360 and Projected Saliency were designed to address this task. In the particular case of BMS360 and GBVS360, the 2D models Boolean Map Saliency (BMS) and Graph-Based Visual Saliency (GBVS) were rigorously extended to integrate the specific properties of equirectangular images to model visual attention in omnidirectional images watched using Head Mounted Displays (HMDs). With the proposed extensions, the saliency prediction performance in comparison to their original design could significantly be improved. Another key result of this paper comes from the design of a framework called projected saliency which allows applying existing saliency models on equirectangular images without requiring in-depth adaptation of the prediction algorithms. The projected saliency framework enabled to study the interaction of features in the feature activation process of computational saliency models. It shows that the activation should not be done per-viewport, but should account for neighboring regions. As a consequence, we argue that the feature activation process should be performed at a global level instead of a local one. Here, the use of a local approach with a large FOV was found to be a good compromise between global activation and appropriate feature computation to directly port existing saliency models designed for rectilinear images to the domain of equirectangular images. In this paper an adaptive equatorial prior is also described. This equatorial prior allows to account for pitch shift of the horizon line. Finally, results show that the newly designed BMS360 and GVBS360 outperform even recent state-of-the-art models at head motion prediction and are competitive at head/eye saliency-map prediction. Moreover, all models and their code are made publicly available on GitHub allowing to further iterate on this work.

Feasibility of high spatiotemporal resolution for an abbreviated 3D radial breast MRI protocol.

To develop a volumetric imaging technique with 0.8-mm isotropic resolution and 10-s/volume rate to detect and analyze breast lesions in a bilateral, dynamic, contrast-enhanced MRI exam. A local low-rank temporal reconstruction approach that also uses parallel imaging and spatial compressed sensing was designed to create rapid volumetric frame rates during a contrast-enhanced breast exam (vastly undersampled isotropic projection [VIPR] spatial compressed sensing with temporal local low-rank [STELLR]). The dynamic-enhanced data are subtracted in k-space from static mask data to increase sparsity for the local low-rank approach to maximize temporal resolution. A T1 -weighted 3D radial trajectory (VIPR iterative decomposition with echo asymmetry and least squares estimation [IDEAL]) was modified to meet the data acquisition requirements of the STELLR approach. Additionally, the unsubtracted enhanced data are reconstructed using compressed sensing and IDEAL to provide high-resolution fat/water separation. The feasibility of the approach and the dual reconstruction methodology is demonstrated using a 16-channel breast coil and a 3T MR scanner in 6 patients. The STELLR temporal performance of subtracted data matched the expected temporal perfusion enhancement pattern in small and large vascular structures. Differential enhancement within heterogeneous lesions is demonstrated with corroboration from a basic reconstruction using a strict 10-second temporal footprint. Rapid acquisition, reliable fat suppression, and high spatiotemporal resolution are presented, despite significant data undersampling. The STELLR reconstruction approach of 3D radial sampling with mask subtraction provides a high-performance imaging technique for characterizing enhancing structures within the breast. It is capable of maintaining temporal fidelity, while visualizing breast lesions with high detail over a large FOV to include both breasts.