Accelerating GluCEST imaging using deep learning for B0 correction.
Ontology highlight
ABSTRACT: Purpose
Glutamate weighted Chemical Exchange Saturation Transfer (GluCEST) MRI is a noninvasive technique for mapping parenchymal glutamate in the brain. Because of the sensitivity to field (B0 ) inhomogeneity, the total acquisition time is prolonged due to the repeated image acquisitions at several saturation offset frequencies, which can cause practical issues such as increased sensitivity to patient motions. Because GluCEST signal is derived from the small z-spectrum difference, it often has a low signal-to-noise-ratio (SNR). We proposed a novel deep learning (DL)-based algorithm armed with wide activation neural network blocks to address both issues.Methods
B0 correction based on reduced saturation offset acquisitions was performed for the positive and negative sides of the z-spectrum separately. For each side, a separate deep residual network was trained to learn the nonlinear mapping from few CEST-weighted images acquired at different ppm values to the one at 3 ppm (where GluCEST peaks) in the same side of the z-spectrum.Results
All DL-based methods outperformed the "traditional" method visually and quantitatively. The wide activation blocks-based method showed the highest performance in terms of Structural Similarity Index (SSIM) and peak signal-to-noise ratio (PSNR), which were 0.84 and 25dB respectively. SNR increases in regions of interest were over 8dB.Conclusion
We demonstrated that the new DL-based method can reduce the entire GluCEST imaging time by ˜50% and yield higher SNR than current state-of-the-art.
SUBMITTER: Li Y
PROVIDER: S-EPMC8082274 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:PurposeTo propose a novel combined parallel imaging and deep learning-based reconstruction framework for robust reconstruction of highly accelerated 2D cardiac cine MRI data.MethodsWe propose DL-ESPIRiT, an unrolled neural network architecture that utilizes an extended coil sensitivity model to address SENSE-related field-of-view (FOV) limitations in previously proposed deep learning-based reconstruction frameworks. Additionally, we propose a novel neural network design based on (2+1)D spatiotemporal convolutions to produce more accurate dynamic MRI reconstructions than conventional 3D convolutions. The network is trained on fully sampled 2D cardiac cine datasets collected from 11 healthy volunteers with IRB approval. DL-ESPIRiT is compared against a state-of-the-art parallel imaging and compressed sensing method known as l1 -ESPIRiT. The reconstruction accuracy of both methods is evaluated on retrospectively undersampled datasets (R = 12) with respect to standard image quality metrics as well as automatic deep learning-based segmentations of left ventricular volumes. Feasibility of DL-ESPIRiT is demonstrated on two prospectively undersampled datasets acquired in a single heartbeat per slice.ResultsThe (2+1)D DL-ESPIRiT method produces higher fidelity image reconstructions when compared to l1 -ESPIRiT reconstructions with respect to standard image quality metrics (P < .001). As a result of improved image quality, segmentations made from (2+1)D DL-ESPIRiT images are also more accurate than segmentations from l1 -ESPIRiT images.ConclusionsDL-ESPIRiT synergistically combines a robust parallel imaging model and deep learning-based priors to produce high-fidelity reconstructions of retrospectively undersampled 2D cardiac cine data acquired with reduced FOV. Although a proof-of-concept is shown, further experiments are necessary to determine the efficacy of DL-ESPIRiT in prospectively undersampled data.
| S-EPMC7722220 | biostudies-literature
Project description:Superresolution imaging of solids is essential to explore local symmetry breaking and derived material properties. Electron ptychography is one of the most promising schemes to realize superresolution imaging beyond aberration correction. However, to reach both deep sub-angstrom resolution imaging and accurate measurement of atomic structures, it is still required for the electron beam to be nearly parallel to the zone axis of crystals. Here, we report an efficient and robust method to correct the specimen misorientation in electron ptychography, giving deep sub-angstrom resolution for specimens with large misorientations. The method largely reduces the experimental difficulties of electron ptychography and paves the way for widespread applications of ptychographic deep sub-angstrom resolution imaging.
| S-EPMC9106290 | biostudies-literature
Project description:A unidirectional imager would only permit image formation along one direction, from an input field-of-view (FOV) A to an output FOV B, and in the reverse path, B → A, the image formation would be blocked. We report the first demonstration of unidirectional imagers, presenting polarization-insensitive and broadband unidirectional imaging based on successive diffractive layers that are linear and isotropic. After their deep learning-based training, the resulting diffractive layers are fabricated to form a unidirectional imager. Although trained using monochromatic illumination, the diffractive unidirectional imager maintains its functionality over a large spectral band and works under broadband illumination. We experimentally validated this unidirectional imager using terahertz radiation, well matching our numerical results. We also created a wavelength-selective unidirectional imager, where two unidirectional imaging operations, in reverse directions, are multiplexed through different illumination wavelengths. Diffractive unidirectional imaging using structured materials will have numerous applications in, e.g., security, defense, telecommunications, and privacy protection.
| S-EPMC11318665 | biostudies-literature
Project description:Psychosis commonly develops in adolescence or early adulthood. Youths at clinical high risk (CHR) for psychosis exhibit similar, subtle symptoms to those with schizophrenia (SZ). Malfunctioning neurotransmitter systems, such as glutamate, are implicated in the disease progression of psychosis. Yet, in vivo imaging techniques for measuring glutamate across the cortex are limited. Here, we use a novel 7 Tesla MRI glutamate imaging technique (GluCEST) to estimate changes in glutamate levels across cortical and subcortical regions in young healthy individuals and ones on the psychosis spectrum. Individuals on the psychosis spectrum (PS; n=19) and healthy young individuals (HC; n=17) underwent MRI imaging at 3 and 7 T. At 7 T, a single slice GluCEST technique was used to estimate in vivo glutamate. GluCEST contrast was compared within and across the subcortex, frontal, parietal and occipital lobes. Subcortical (χ2 (1)=4.65, P=0.031) and lobular (χ2 (1)=5.17, P=0.023) GluCEST contrast levels were lower in PS compared with HC. Abnormal GluCEST contrast levels were evident in both CHR (n=14) and SZ (n=5) subjects, and correlated differentially, across regions, with clinical symptoms. Our findings describe a pattern of abnormal brain neurochemistry early in the course of psychosis. Specifically, CHR and young SZ exhibit diffuse abnormalities in GluCEST contrast attributable to a major contribution from glutamate. We suggest that neurochemical profiles of GluCEST contrast across cortex and subcortex may be considered markers of early psychosis. GluCEST methodology thus shows promise to further elucidate the progression of the psychosis disease state.
| S-EPMC5822706 | biostudies-literature
Project description:PurposeTo describe and implement a strategy for dynamic slice-by-slice and multiband B0 shimming using spherical harmonic shims in the human brain at 7T.TheoryFor thin axial slices, spherical harmonic shims can be divided into pairs of shims (z-degenerate and non-z-degenerate) that are spatially degenerate, such that only ½ of the shims (non-z-degenerate) are required for single slice optimizations. However, when combined, the pairs of shims can be used to simultaneously generate the same in-plane symmetries but with different amplitudes as a function of their z location. This enables multiband shimming equivalent to that achievable by single slice-by-slice optimization.MethodsAll data were acquired at 7T using a spherical harmonic shim insert enabling shimming up through 4th order with two additional 5th order shims (1st-4th+). Dynamic shim updating was achieved using a 10A shim power supply with 2 ms ramps and constrained optimizations to minimize eddy currents.ResultsIn groups of eight subjects, we demonstrated that: 1) dynamic updating using 1st-4th+ order shims reduced the SD of the B0 field over the whole brain from 32.4 ± 2.6 and 24.9 ± 2 Hz with 1st-2nd and 1st-4th+ static global shimming to 15.1 ± 1.7 Hz; 2) near equivalent performance was achieved when dynamically updating only the non-z-degenerate shims (14.3 ± 1.5 Hz), or when a using multiband shim factor of 2, MBs = 2, and all shims (14.4 ± 2.0 Hz).ConclusionHigh order spherical harmonics provide substantial improvements over static global shimming and enable dynamic multiband shimming with near equivalent performance to that of dynamic slice-by-slice shimming. This reduces distortion in echo planar imaging.
| S-EPMC8098050 | biostudies-literature
Project description:Quantitative [15O]H2O positron emission tomography (PET) is the accepted reference method for regional cerebral blood flow (rCBF) quantification. To perform reliable quantitative [15O]H2O-PET studies in PET/MRI scanners, MRI-based attenuation-correction (MRAC) is required. Our aim was to compare two MRAC methods (RESOLUTE and DeepUTE) based on ultrashort echo-time with computed tomography-based reference standard AC (CTAC) in dynamic and static [15O]H2O-PET. We compared rCBF from quantitative perfusion maps and activity concentration distribution from static images between AC methods in 25 resting [15O]H2O-PET scans from 14 healthy men at whole-brain, regions of interest and voxel-wise levels. Average whole-brain CBF was 39.9 ± 6.0, 39.0 ± 5.8 and 40.0 ± 5.6 ml/100 g/min for CTAC, RESOLUTE and DeepUTE corrected studies respectively. RESOLUTE underestimated whole-brain CBF by 2.1 ± 1.50% and rCBF in all regions of interest (range -2.4%- -1%) compared to CTAC. DeepUTE showed significant rCBF overestimation only in the occipital lobe (0.6 ± 1.1%). Both MRAC methods showed excellent correlation on rCBF and activity concentration with CTAC, with slopes of linear regression lines between 0.97 and 1.01 and R2 over 0.99. In conclusion, RESOLUTE and DeepUTE provide AC information comparable to CTAC in dynamic [15O]H2O-PET but RESOLUTE is associated with a small but systematic underestimation.
| S-EPMC8669198 | biostudies-literature
Project description:PurposeAttenuation correction and scatter compensation (AC/SC) are two main steps toward quantitative PET imaging, which remain challenging in PET-only and PET/MRI systems. These can be effectively tackled via deep learning (DL) methods. However, trustworthy, and generalizable DL models commonly require well-curated, heterogeneous, and large datasets from multiple clinical centers. At the same time, owing to legal/ethical issues and privacy concerns, forming a large collective, centralized dataset poses significant challenges. In this work, we aimed to develop a DL-based model in a multicenter setting without direct sharing of data using federated learning (FL) for AC/SC of PET images.MethodsNon-attenuation/scatter corrected and CT-based attenuation/scatter corrected (CT-ASC) 18F-FDG PET images of 300 patients were enrolled in this study. The dataset consisted of 6 different centers, each with 50 patients, with scanner, image acquisition, and reconstruction protocols varying across the centers. CT-based ASC PET images served as the standard reference. All images were reviewed to include high-quality and artifact-free PET images. Both corrected and uncorrected PET images were converted to standardized uptake values (SUVs). We used a modified nested U-Net utilizing residual U-block in a U-shape architecture. We evaluated two FL models, namely sequential (FL-SQ) and parallel (FL-PL) and compared their performance with the baseline centralized (CZ) learning model wherein the data were pooled to one server, as well as center-based (CB) models where for each center the model was built and evaluated separately. Data from each center were divided to contribute to training (30 patients), validation (10 patients), and test sets (10 patients). Final evaluations and reports were performed on 60 patients (10 patients from each center).ResultsIn terms of percent SUV absolute relative error (ARE%), both FL-SQ (CI:12.21-14.81%) and FL-PL (CI:11.82-13.84%) models demonstrated excellent agreement with the centralized framework (CI:10.32-12.00%), while FL-based algorithms improved model performance by over 11% compared to CB training strategy (CI: 22.34-26.10%). Furthermore, the Mann-Whitney test between different strategies revealed no significant differences between CZ and FL-based algorithms (p-value > 0.05) in center-categorized mode. At the same time, a significant difference was observed between the different training approaches on the overall dataset (p-value < 0.05). In addition, voxel-wise comparison, with respect to reference CT-ASC, exhibited similar performance for images predicted by CZ (R2 = 0.94), FL-SQ (R2 = 0.93), and FL-PL (R2 = 0.92), while CB model achieved a far lower coefficient of determination (R2 = 0.74). Despite the strong correlations between CZ and FL-based methods compared to reference CT-ASC, a slight underestimation of predicted voxel values was observed.ConclusionDeep learning-based models provide promising results toward quantitative PET image reconstruction. Specifically, we developed two FL models and compared their performance with center-based and centralized models. The proposed FL-based models achieved higher performance compared to center-based models, comparable with centralized models. Our work provided strong empirical evidence that the FL framework can fully benefit from the generalizability and robustness of DL models used for AC/SC in PET, while obviating the need for the direct sharing of datasets between clinical imaging centers.
| S-EPMC9742659 | biostudies-literature
Project description:Patient motion during dynamic PET imaging can induce errors in myocardial blood flow (MBF) estimation. Motion correction for dynamic cardiac PET is challenging because the rapid tracer kinetics of 82Rb leads to substantial tracer distribution change across different dynamic frames over time, which can cause difficulties for image registration-based motion correction, particularly for early dynamic frames. In this paper, we developed an automatic deep learning-based motion correction (DeepMC) method for dynamic cardiac PET. In this study we focused on the detection and correction of inter-frame rigid translational motion caused by voluntary body movement and pattern change of respiratory motion. A bidirectional-3D LSTM network was developed to fully utilize both local and nonlocal temporal information in the 4D dynamic image data for motion detection. The network was trained and evaluated over motion-free patient scans with simulated motion so that the motion ground-truths are available, where one million samples based on 65 patient scans were used in training, and 600 samples based on 20 patient scans were used in evaluation. The proposed method was also evaluated using additional 10 patient datasets with real motion. We demonstrated that the proposed DeepMC obtained superior performance compared to conventional registration-based methods and other convolutional neural networks (CNN), in terms of motion estimation and MBF quantification accuracy. Once trained, DeepMC is much faster than the registration-based methods and can be easily integrated into the clinical workflow. In the future work, additional investigation is needed to evaluate this approach in a clinical context with realistic patient motion.
| S-EPMC8670362 | biostudies-literature
Project description:The materials discovery process can be significantly expedited and simplified if we can learn effectively from available knowledge and data. In the present contribution, we show that efficient and accurate prediction of a diverse set of properties of material systems is possible by employing machine (or statistical) learning methods trained on quantum mechanical computations in combination with the notions of chemical similarity. Using a family of one-dimensional chain systems, we present a general formalism that allows us to discover decision rules that establish a mapping between easily accessible attributes of a system and its properties. It is shown that fingerprints based on either chemo-structural (compositional and configurational information) or the electronic charge density distribution can be used to make ultra-fast, yet accurate, property predictions. Harnessing such learning paradigms extends recent efforts to systematically explore and mine vast chemical spaces, and can significantly accelerate the discovery of new application-specific materials.
| S-EPMC3786293 | biostudies-literature
Project description:PurposeTo measure spatiotemporal B0 field changes in real time using FID navigators (FIDnavs) and to demonstrate the efficacy of retrospectively correcting high-resolution T2* -weighted images using a novel FIDnav framework.MethodsA forward model of the complex FIDnav signals was generated by simulating the effect of changes in the underlying B0 inhomogeneity coefficients, with spatial encoding provided by a multi-channel reference image. Experiments were performed at 3T to assess the accuracy of B0 field estimates from FIDnavs acquired from a 64-channel head coil under different shim settings and in 5 volunteers performing deep-breathing and nose-touching tasks designed to modulate the B0 field. Second-order, in-plane spherical harmonic (SH) inhomogeneity coefficients estimated from FIDnavs were incorporated into an iterative reconstruction to retrospectively correct 2D gradient-echo images acquired in both axial and sagittal planes.ResultsSpatiotemporal B0 field changes measured from rapidly acquired FIDnavs were in good agreement with the results of second-order SH fitting to the measured field maps. FIDnav field estimates accounted for a significant proportion of the ΔB0 variance induced by deep breathing (64 ± 21%) and nose touching (67 ± 34%) across all volunteers. Ghosting, blurring, and intensity modulation artifacts in T2* -weighted images, induced by spatiotemporal field changes, were visibly reduced following retrospective correction with FIDnav inhomogeneity coefficients.ConclusionsSpatially resolved B0 inhomogeneity changes up to second order can be characterized in real time using the proposed approach. Retrospective FIDnav correction substantially improves T2* -weighted image quality in the presence of strong B0 field modulations, with potential for real-time shimming.
| S-EPMC6824939 | biostudies-literature