Reduced Scan Time Research Articles

Accelerated Synthetic MRI with Deep Learning-Based Reconstruction for Pediatric Neuroimaging.

Synthetic MR imaging is a time-efficient technique. However, its rather long scan time can be challenging for children. This study aimed to evaluate the clinical feasibility of accelerated synthetic MR imaging with deep learning-based reconstruction in pediatric neuroimaging and to investigate the impact of deep learning-based reconstruction on image quality and quantitative values in synthetic MR imaging. This study included 47 children 2.3-14.7 years of age who underwent both standard and accelerated synthetic MR imaging at 3T. The accelerated synthetic MR imaging was reconstructed using a deep learning pipeline. The image quality, lesion detectability, tissue values, and brain volumetry were compared among accelerated deep learning and accelerated and standard synthetic data sets. The use of deep learning-based reconstruction in the accelerated synthetic scans significantly improved image quality for all contrast weightings (P < .001), resulting in image quality comparable with or superior to that of standard scans. There was no significant difference in lesion detectability between the accelerated deep learning and standard scans (P > .05). The tissue values and brain tissue volumes obtained with accelerated deep learning and the other 2 scans showed excellent agreement and a strong linear relationship (all, R 2 > 0.9). The difference in quantitative values of accelerated scans versus accelerated deep learning scans was very small (tissue values, <0.5%; volumetry, -1.46%-0.83%). The use of deep learning-based reconstruction in synthetic MR imaging can reduce scan time by 42% while maintaining image quality and lesion detectability and providing consistent quantitative values. The accelerated deep learning synthetic MR imaging can replace standard synthetic MR imaging in both contrast-weighted and quantitative imaging.

Application of deep learning–based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time

ObjectivesTo compare the image quality and diagnostic performance of conventional motion-corrected periodically rotated overlapping parallel line with enhanced reconstruction (PROPELLER) MRI sequences with post-processed PROPELLER MRI sequences using deep learning-based (DL) reconstructions.MethodsIn this prospective study of 30 patients, conventional (19 min 18 s) and accelerated MRI sequences (7 min 16 s) using the PROPELLER technique were acquired. Accelerated sequences were post-processed using DL. The image quality and diagnostic confidence were qualitatively assessed by 2 readers using a 5-point Likert scale. Analysis of the pathological findings of cartilage, rotator cuff tendons and muscles, glenoid labrum and subacromial bursa was performed. Inter-reader agreement was calculated using Cohen’s kappa statistic. Quantitative evaluation of image quality was measured using the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR).ResultsMean image quality and diagnostic confidence in evaluation of all shoulder structures were higher in DL sequences (p value = 0.01). Inter-reader agreement ranged between kappa values of 0.155 (assessment of the bursa) and 0.947 (assessment of the rotator cuff muscles). In 17 cases, thickening of the subacromial bursa of more than 2 mm was only visible in DL sequences. The pathologies of the other structures could be properly evaluated by conventional and DL sequences. Mean SNR (p value = 0.01) and CNR (p value = 0.02) were significantly higher for DL sequences.ConclusionsThe accelerated PROPELLER sequences with DL post-processing showed superior image quality and higher diagnostic confidence compared to the conventional PROPELLER sequences. Subacromial bursa can be thoroughly assessed in DL sequences, while the other structures of the shoulder joint can be assessed in conventional and DL sequences with a good agreement between sequences.Key Points• MRI of the shoulder requires long scan times and can be hampered by motion artifacts.• Deep learning–based convolutional neural networks are used to reduce image noise and scan time while maintaining optimal image quality. The radial k-space acquisition technique (PROPELLER) can reduce the scan time and has potential to reduce motion artifacts.• DL sequences show a higher diagnostic confidence than conventional sequences and therefore are preferred for assessment of the subacromial bursa, while conventional and DL sequences show comparable performance in the evaluation of the shoulder joint.

Artificial Intelligence for Contrast-Free MRI: Scar Assessment in Myocardial Infarction Using Deep Learning-Based Virtual Native Enhancement.

Myocardial scars are assessed noninvasively using cardiovascular magnetic resonance late gadolinium enhancement (LGE) as an imaging gold standard. A contrast-free approach would provide many advantages, including a faster and cheaper scan without contrast-associated problems. Virtual native enhancement (VNE) is a novel technology that can produce virtual LGE-like images without the need for contrast. VNE combines cine imaging and native T1 maps to produce LGE-like images using artificial intelligence. VNE was developed for patients with previous myocardial infarction from 4271 data sets (912 patients); each data set comprises slice position-matched cine, T1 maps, and LGE images. After quality control, 3002 data sets (775 patients) were used for development and 291 data sets (68 patients) for testing. The VNE generator was trained using generative adversarial networks, using 2 adversarial discriminators to improve the image quality. The left ventricle was contoured semiautomatically. Myocardial scar volume was quantified using the full width at half maximum method. Scar transmurality was measured using the centerline chord method and visualized on bull's-eye plots. Lesion quantification by VNE and LGE was compared using linear regression, Pearson correlation (R), and intraclass correlation coefficients. Proof-of-principle histopathologic comparison of VNE in a porcine model of myocardial infarction also was performed. VNE provided significantly better image quality than LGE on blinded analysis by 5 independent operators on 291 data sets (all P<0.001). VNE correlated strongly with LGE in quantifying scar size (R, 0.89; intraclass correlation coefficient, 0.94) and transmurality (R, 0.84; intraclass correlation coefficient, 0.90) in 66 patients (277 test data sets). Two cardiovascular magnetic resonance experts reviewed all test image slices and reported an overall accuracy of 84% for VNE in detecting scars when compared with LGE, with specificity of 100% and sensitivity of 77%. VNE also showed excellent visuospatial agreement with histopathology in 2 cases of a porcine model of myocardial infarction. VNE demonstrated high agreement with LGE cardiovascular magnetic resonance for myocardial scar assessment in patients with previous myocardial infarction in visuospatial distribution and lesion quantification with superior image quality. VNE is a potentially transformative artificial intelligence-based technology with promise in reducing scan times and costs, increasing clinical throughput, and improving the accessibility of cardiovascular magnetic resonance in the near future.

Radiomics analysis of short tau inversion recovery images in cardiac magnetic resonance for the prediction of late gadolinium enhancement in patients with acute myocarditis

Rationale and objectivesCardiac Magnetic Resonance (CMR) imaging is recommended as the reference diagnostic non-invasive modality for myocarditis but is often limited by patients' compliance. The purpose of this study is to evaluate the validity of Radiomics applied to Short Tau Inversion Recovery (STIR) sequences, in predicting the presence of LGE in patients with suspected acute myocarditis. Materials and methods171 STIR images on short-axis view were segmented with “MaZda” software ver 4.6, by placing a region of interest (ROI) on the left ventricle by two radiologists in consensus. Images were classified according to the presence of LGE in the equivalent short-axis T1-IR slice. A total of 337 ROI features were extracted for each image. Dataset was then split into two parts (train and test set) with 70:30 ratio. ResultsEleven classification models were trained. An Ensemble Machine Learning (EML) model was obtained by averaging the predictions of models with accuracy on test set >70%. The EML documented accuracy of 0.75, sensitivity of 0.8 and a specificity of 0.73 with a NPV of 0.81 and a PPV of 0.7, with AUC of 0.79 (95% CI: 0.66–0.92). ConclusionRadiomics and machine learning analysis could be a promising approach in reducing scan times without reducing diagnostic accuracy in predicting LGE in patients with acute myocarditis.

Acoustic-resolution-based spectroscopic photoacoustic endoscopy towards molecular imaging in deep tissues.

Due to many technical difficulties, the study of molecular photoacoustic endoscopic (PAE) imaging in deep tissues is limited. In this work, we have set up a multimodal acoustic-resolution-based PAE (AR-PAE) system to image the rabbit rectum and preliminarily explored the potential of molecular PAE for deep-seated targets in proof-of-concept. We developed an improved back-projection (IBP) algorithm for focused detection over the centimeter-scale imaging depth. We also developed a deep-learning-based algorithm to remove the electrical noise from the step motor to prevent data averaging for reduced scanning time. We injected a dose of indocyanine green (ICG) near the rabbit rectum and compared 2D and 3D photoacoustic/ultrasound (PA/US) images at different wavelengths. We proposed incorporating a small camera to guide the slow PA/US endoscopic scan. Results show that this system has achieved a lateral resolution of about 0.77/0.65 mm for PA/US images with a signal-to-noise ratio (SNR) of 25/38 dB at an imaging depth of 1.4 cm. We found that the rectum wall and the ICG can be well distinguished spectroscopically. Results also show that the PA images at 532 nm have higher signal intensity and reflection artifacts from pelvic tendons and bones than those at longer wavelengths such as 800 nm. The proposed methods and the intuitive findings in this work may guide and promote the development of high-penetration molecular PAE.

Improving Image Quality and Reducing Scan Time for Synthetic MRI of Breast by Using Deep Learning Reconstruction.

Objectives To investigate a deep learning reconstruction algorithm to reduce the time of synthetic MRI (SynMRI) scanning on the breast and improve the image quality. Materials and Methods A total of 192 healthy female volunteers (mean age: 48.1 years) underwent the breast MR examination at 3.0 T from September 2020 to June 2021. Standard SynMRI and fast SynMRI scans were collected simultaneously on the same volunteer. Deep learning technology with a generative adversarial network (GAN) was used to generate high-quality fast SynMRI images by end-to-end training. Peak signal-to-noise ratio (PSNR), mean squared error (MSE), and structural similarity index measure (SSIM) were used to compare the image quality of generated images from fast SynMRI by deep learning algorithms. Results Fast SynMRI acquisition time is half of the standard SynMRI scan, and the generated images of the GAN model show that PSNR and SSIM are improved and MSE is reduced. Conclusion The application of deep learning algorithms with GAN model in breast MAGiC MRI improves the image quality and reduces the scanning time.

CNNT DB-LGE: free-breathing dark blood late enhancement imaging using the convolutional neural network transformer speeds acquisition by 50%

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Supported in part by the Division of Intramural Research of the National Heart, Lung, and Blood Institute, National Institutes of Health (grants Z1A-HL006214-05 and Z1A-HL006242-02). Background Dark blood late gadolinium enhancement (DB-LGE) imaging shows superior delineation of myocardial infarction (MI), especially at the sub-endocardial boundary. Our previous study [1] developed a free-breathing DB-LGE with the single shot SSFP readout, phase sensitive inversion recovery (PSIR) reconstruction, and respiratory motion corrected averaging. To compensate the potential signal-to-noise ratio loss, our previous DB-LGE doubled the measurements, thereby increasing the acquisition time. Purpose In this study, we developed a deep learning image enhancement model using a novel neural network architecture called the convolutional neural network transformer (CNNT) to improve the image quality of DB-LGE and to reduce the acquisition time by decreasing the number of measurements. Methods A novel image enhancement model was developed using a novel network architecture called the Convolutional Neural Network Transformer (CNNT) proposed by us. This architecture is suitable for the 2D+Time CMR acquisition, by exploiting the temporal correlation between images over multiple averages. The evaluation was first retrospectively conducted on a cohort of 12 patients acquired with the original protocol [1] using the full 16 measurements. For every subject, a complete short-axis stack (typically 12 slices) was acquired to cover the entire left ventricular. The imaging data was reconstructed in three ways. Original: using all acquired 16 measurements. This is our base-line protocol. Original 50%: using only the first 8 measurements. CNNT 50%: using only the first 8 averages, but performing the CNNT deep learning image enhancement before MOCO PSIR reconstruction. Two experienced imaging researchers (PK and MF, &amp;gt;10 years of experience for both) scored all DB-LGE images for the overall quality, diagnostic confidence and delineation of MI/boundaries (5 = excellent, 4 = good, 3 = fair, 2 = poor, and 1 = non-diagnostic). The CNNT DB-LGE was deployed to the MR scanner using the Gadgetron InlineAI [2]. Results Figure 1 gives examples of DB-LGE with three reconstruction methods. The CNNT image has higher SNR and well delineated MI. The Original images with the longest acquisition have good quality and the Original-50% acquired with 8 measurements are good quality but have reduced SNR. The mean scores for overall image quality, diagnostic confidence and MI delineation of two reviewers were 4.88±0.23, 4.88±0.23, 4.83±0.25 for CNNT and 4.96±0.14, 4.96±0.14, 4.67±0.39 for the original approach. No significant differences were found between the original and the CNNT (P&amp;gt;0.15 for all). Figure 2 shows an acute MI patient prospectively acquired with the 50% scan time reduction, with and without the CNNT enhancement. The resulting PSIR images well delineate the MVO due to the acute MI, with improved SNR. Conclusion A novel CNNT model was proposed and evaluated to speed up the free-breathing MOCO DB LGE by 50% without sacrificing image quality.

A uniformity correction method to reduce scan time for 7T sodium imaging of brain tumors.

Sodium imaging shows great potential for the characterization of brain tumors. Intensity correction is required but the additional scan time is costly. Recent developments can halve the time but were optimized in normal brains and may not be applicable in brain tumor imaging. We aim to develop an individualized uniformity correction for sodium imaging optimized for brain tumor patients that reduces scan time but provides high-resolution images for clinical practice. Two-, 4-, and 6-mm iso-cubic voxel resolution birdcage coil images were used to calculate the 2-mm iso-cubic voxel individual sensitivity maps in healthy subjects (n=3). Cut profiles were compared to determine the optimal approach. In addition, a 3-dimensional phantom was developed to test a generalized uniformity correction approach in both healthy subjects (n=3) and tumor patients (n=3). The cut profiles showed that the average correlation coefficient between 2- and 4-mm birdcage image correction results was r=.9937, and r=.9876 for 2- and 6-mm birdcage images. The correlation result between individual map correction and phantom map correction was r=.9817. The 4mm birdcage coil image provided the optimal approach for both as a compromise between the time-savings effect and image quality. This method allows for a 2-mm iso-cubic voxel resolution clinical sodium scan within 12minutes. We also presented prescanned phantom sensitivity map results, which were designed to cover all patient head sizes. This approach provides an alternative solution in more time-sensitive cases.

3D Cube FLAIR plus HyperSense compressed sensing is superior to 2D T2WI FLAIR scanning regarding image quality, spatial resolution, detection rate for cortical microinfarcts.

3-dimention (3D) Cube isotropic volumetric magnetic resonance imaging (MRI) facilitates comprehensive recognition of microinfarcts while it takes long scanning time. HyperSense compressed sensing is an emerging technique for accelerating MRI acquisition to reduce scanning time, while its application along with 3D Cube MRI for microinfarcts is seldom reported. Therefore, this study aimed to investigate the efficiency of 3D Cube FLAIR plus HyperSense compressed sensing technique versus conventional 2-dimention (2D) FLAIR scanning in the detection of cortical microinfarcts (CMIs).Totally 59 patients with cerebrovascular disease were enrolled then scanned by 3D Cube FLAIR plus HyperSense compressed sensing and 2D T2WI FLAIR sequences. The image quality scores, signal-to-noise ratio (SNR) for gray matter (GM), SNR for white matter (WM), their contrast-to-noise ratio (WM-to-GM CNR), detected number of CMIs were evaluated.3D Cube FLAIR plus HyperSense showed a dramatically increased scores of uniformity, artifact, degree of lesion displacement, and overall image quality compared to 2D T2WI FLAIR. Meanwhile, it exhibited similar SNRwm and SNRgm, but a higher WM-to-GM contrast-to-noise ratio compared with 2D T2WI FLAIR. Furthermore, the scanning time of 3D Cube FLAIR plus HyperSense and 2D T2WI FLAIR were both set as 2.5 minutes. Encouragingly, 244 CMIs were detected by 3D Cube FLAIR plus HyperSense, which was higher compared to 2D T2WI FLAIR (106 detected CMIs).3D Cube FLAIR plus HyperSense compressed sensing is superior to 2D T2WI FLAIR scanning regarding image quality, spatial resolution, detection rate for CMIs; meanwhile, it does not increase the scanning time. These findings may contribute to early detection and treatment of stroke.

Optimal timing of SPECT/CT to demonstrate parathyroid adenomas in 99mTc-sestamibi scintigraphy.

Accurate preoperative localisation of the parathyroid adenoma is essential to achieve a minimally invasive parathyroidectomy. The purpose of this study was to validate and improve our single-isotope dual-phase parathyroid imaging protocol utilising 99mTechnetium-Sestamibi ([99mTc]MIBI). There has been no accepted gold standard evidence-based protocol regarding timing of single-photon emission computed tomography/computed tomography (SPECT/CT) acquisition in parathyroid imaging with resultant variation between centres. We sought to determine the optimum timing of SPECT/CT post administration of [99mTc]MIBI in the identification of parathyroid adenomas. We aimed to evaluate the efficacy of early and late SPECT/CT and to establish whether SPECT/CT demonstrates increased sensitivity over planar imaging. A sample of 36 patients with primary hyperparathyroidism underwent planar and SPECT/CT acquisition 15 minutes (early) and two hours (late) post [99mTc]MIBI administration. Two radionuclide radiologists reviewed the images and Fisher's exact Chi-squared statistic was used to evaluate the diagnostic performances of early versus late SPECT/CT acquisition and SPECT/CT versus planar imaging. Twenty-one likely parathyroid adenomas were identified with a statistically superior diagnosis rate in the late SPECT/CT acquisition compared with both early SPECT/CT and planar imaging (p < 0.05). All adenomas diagnosed on early SPECT/CT acquisition were also identified on late SPECT/CT images. Single late phase SPECT/CT is significantly superior to early SPECT/CT in the identification of parathyroid adenomas. Late SPECT/CT improves diagnostic accuracy over planar acquisition. Imaging protocols should be revised to include late SPECT/CT acquisition. Early SPECT/CT acquisition can be eliminated from scan protocols with associated implications regarding reduced scan time and increased patient throughput.

Accelerated 3D T2-weighted images using compressed sensing for pediatric brain imaging.

The purpose of this study was to compare the image quality of the 3D T2-weighted images accelerated using conventional method (CAI-SPACE) with the images accelerated using compressed sensing (CS-SPACE) in pediatric brain imaging. A total of 116 brain MRI (53 with CAI-SPACE and 63 with CS-SPACE) were obtained from children 16years old or younger. Quantitative image quality was evaluated using the apparent signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). The sequences were qualitatively evaluated for overall image quality, general artifact, cerebrospinal fluid (CSF)-related artifact, and grey-white matter differentiation. The two sequences were compared for the total and two age groups (< 24months vs. ≥ 24months). Compressed sensing application in 3D T2-weighted imaging resulted in 8.5% reduction in scanning time. Quantitative image quality analysis showed higher apparent SNR (median [Interquartile range]; 29 [25] vs. 23 [14], P = 0.005) and CNR (0.231 [0.121] vs. 0.165 [0.120], P = 0.027) with CS-SPACE compared to CAI-SPACE. Qualitative image quality analysis showed better image quality with CS-SPACE for general (P = 0.024) and CSF-related artifact (P < 0.001). CSF-related artifacts reduction was prominent in the older age group (≥ 24months). Overall image quality (P = 0.162) and grey-white matter differentiation (P = 0.397) were comparable between CAI-SPACE and CS-SPACE. Compressed sensing application in 3D T2-weighted images modestly reduced acquisition time and lowered CSF-related artifact compared to conventional images of the pediatric brain.

Deep-learning prediction of amyloid deposition from early-phase amyloid positron emission tomography imaging.

While the use of biomarkers for the detection of early and preclinical Alzheimer's Disease has become essential, the need to wait for over an hour after injection to obtain sufficient image quality can be challenging for patients with suspected dementia and their caregivers. This study aimed to develop an image-based deep-learning technique to generate delayed uptake patterns of amyloid positron emission tomography (PET) images using only early-phase images obtained from 0-20min after radiotracer injection. We prepared pairs of early and delayed [11C]PiB dynamic images from 253 patients (cognitively normal n = 32, fronto-temporal dementia n = 39, mild cognitive impairment n = 19, Alzheimer's disease n = 163) as a training dataset. The neural network was trained with the early images as the input, and the output was the corresponding delayed image. A U-net convolutional neural network (CNN) and a conditional generative adversarial network (C-GAN) were used for the deep-learning architecture and the data augmentation methods, respectively. Then, an independent test data set consisting of early-phase amyloid PET images (n = 19) was used to generate corresponding delayed images using the trained network. Two nuclear medicine physicians interpreted the actual delayed images and predicted delayed images for amyloid positivity. In addition, the concordance of the actual delayed and predicted delayed images was assessed statistically. The concordance of amyloid positivity between the actual versus AI-predicted delayed images was 79%(κ = 0.60) and 79% (κ = 0.59) for each physician, respectively. In addition, the physicians' agreement rate was at 89% (κ = 0.79) when the same image was interpreted. And, the actual versus AI-predicted delayed images were not readily distinguishable (correct answer rate, 55% and 47% for each physician, respectively). The statistical comparison of the actual versus the predicted delated images indicated that the peak signal-to-noise ratio (PSNR) was 21.8dB ± 2.2dB, and the structural similarity index (SSIM) was 0.45 ± 0.04. This study demonstrates the feasibility of an image-based deep-learning framework to predict delayed patterns of Amyloid PET uptake using only the early phase images. This AI-based image generation method has the potential to reduce scan time for amyloid PET and increase the patient throughput, without sacrificing diagnostic accuracy for amyloid positivity.

Evaluating of the Quality of Hepatic Diffusion Weighted Imaging Using Multiband Imaging With Variable-Rate Selective Excitation.

To assess the image quality of diffusion-weighted imaging (DWI) using multiband (MB) imaging with variable-rate selective excitation (VERSE) and compare it to conventional DWI. We retrospectively evaluated hepatic DWI images of patients (n = 76) according to either the conventional method (SENSE, acceleration factor = 2) (n = 38) or fast scanning method (MB imaging with VERSE, acceleration factor = 2 × 2) (n = 38). We also conducted a volunteer study (n = 15) for those scanning methods. During quantitative analysis, the signal-to-noise ratio (SNR), apparent diffusion coefficient values, and contrast in the liver, spleen, and spinal cord were compared between the 2 groups. During qualitative analysis, all images were independently and blindly evaluated by 2 board-certified radiologists. The image contrast, noise, artifacts, and sharpness were assessed, and the performance of classification was measured using receiver operating characteristic curve analysis. In the retrospective study, the SNRs of the hepatic parenchyma and spinal cord between the 2 protocols were significantly different (liver, 8.9 [interquartile range {IQR}, 7.6-12.2] vs 13.0 [IQR, 10.0-16.7]; P < 0.001 and spinal cord, 6.0 [IQR, 4.7-9.4] vs 4.3 [IQR, 3.8-6.8]; P < 0.02). No significant differences between the 2 protocols in the other retrospective analyses were noted. In the receiver operating characteristic curve analysis, area under the curve was 0.49 (95% confidence intervals, 0.40-0.58). Multiband VERSE reduced scan time and SNR of hepatic DWI; however, subjective image quality parameters were not significantly impacted.

Tissue volume estimation and age prediction using rapid structural brain scans

The multicontrast EPImix sequence generates six contrasts, including a T1-weighted scan, in ~1 min. EPImix shows comparable diagnostic performance to conventional scans under qualitative clinical evaluation, and similarities in simple quantitative measures including contrast intensity. However, EPImix scans have not yet been compared to standard MRI scans using established quantitative measures. In this study, we compared conventional and EPImix-derived T1-weighted scans of 64 healthy participants using tissue volume estimates and predicted brain-age. All scans were pre-processed using the SPM12 DARTEL pipeline, generating measures of grey matter, white matter and cerebrospinal fluid volume. Brain-age was predicted using brainageR, a Gaussian Processes Regression model previously trained on a large sample of standard T1-weighted scans. Estimates of both global and voxel-wise tissue volume showed significantly similar results between standard and EPImix-derived T1-weighted scans. Brain-age estimates from both sequences were significantly correlated, although EPImix T1-weighted scans showed a systematic offset in predictions of chronological age. Supplementary analyses suggest that this is likely caused by the reduced field of view of EPImix scans, and the use of a brain-age model trained using conventional T1-weighted scans. However, this systematic error can be corrected using additional regression of T1-predicted brain-age onto EPImix-predicted brain-age. Finally, retest EPImix scans acquired for 10 participants demonstrated high test-retest reliability in all evaluated quantitative measurements. Quantitative analysis of EPImix scans has potential to reduce scanning time, increasing participant comfort and reducing cost, as well as to support automation of scanning, utilising active learning for faster and individually-tailored (neuro)imaging.

Joint reconstruction and segmentation of noisy velocity images as an inverse Navier–Stokes problem

We formulate and solve a generalized inverse Navier–Stokes problem for the joint velocity field reconstruction and boundary segmentation of noisy flow velocity images. To regularize the problem, we use a Bayesian framework with Gaussian random fields. This allows us to estimate the uncertainties of the unknowns by approximating their posterior covariance with a quasi-Newton method. We first test the method for synthetic noisy images of two-dimensional (2-D) flows and observe that the method successfully reconstructs and segments the noisy synthetic images with a signal-to-noise ratio (SNR) of three. Then we conduct a magnetic resonance velocimetry (MRV) experiment to acquire images of an axisymmetric flow for low ( ${\simeq }6$ ) and high ( ${&gt;}30$ ) SNRs. We show that the method is capable of reconstructing and segmenting the low SNR images, producing noiseless velocity fields and a smooth segmentation, with negligible errors compared with the high SNR images. This amounts to a reduction of the total scanning time by a factor of 27. At the same time, the method provides additional knowledge about the physics of the flow (e.g. pressure) and addresses the shortcomings of MRV (i.e. low spatial resolution and partial volume effects) that otherwise hinder the accurate estimation of wall shear stresses. Although the implementation of the method is restricted to 2-D steady planar and axisymmetric flows, the formulation applies immediately to three-dimensional (3-D) steady flows and naturally extends to 3-D periodic and unsteady flows.

Synthetically trained convolutional neural networks for improved tensor estimation from free-breathing cardiac DTI.

Cardiac diffusion tensor imaging (cDTI) provides invaluable information about the state of myocardial microstructure. For further clinical dissemination, free-breathing acquisitions are desired, which however require image registration prior to tensor estimation. Due to the varying contrast and the intrinsically low signal-to-noise ratio (SNR), registration is very challenging and thus can introduce additional errors in the tensor estimation. In the work at hand it is hypothesized, that by incorporating spatial information and physiologically plausible priors into the fitting algorithm, the robustness of diffusion tensor estimation can be improved. To this end, we present a parameterized pipeline to generate synthetic data, that captures the statistics including spatial correlations of diffusion tensors and motion of the heart. The synthetic data is used to train a residual convolutional neural network (CNN) to estimate diffusion tensors from unregistered in-vivo cDTI data. Using in-silico data, the synthetically trained CNN is demonstrated to yield increased tensor estimation accuracy and precision when compared to conventional registration followed by least squares fitting. The network outputs fewer outliers especially at the myocardial borders. In-vivo feasibility using data from five healthy subjects demonstrates the utility of the synthetically trained network. The in-vivo results predicted by the synthetically trained CNN are found to be consistent with the registered least-squares estimates while showing fewer outliers and reduced noise. Even in low SNR regimes, the network results in robust tensor estimation, enabling scan time reduction by reduced-average acquisition in-vivo. Finally, to investigate the network's capability of discriminating between healthy and lesioned tissue, the in-vivo data was artificially augmented showing preserved classification of tissue states based on diffusion metrics.

Clinical feasibility of accelerated diffusion weighted imaging of the abdomen with deep learning reconstruction: Comparison with conventional diffusion weighted imaging

PurposeTo assess the clinical feasibility of accelerated deep learning–reconstructed diffusion weighted imaging (DWI) and to compare its image quality and acquisition time with those of conventional DWI. MethodsSeventy-four consecutive patients who underwent 3 T abdominal magnetic resonance imaging (MRI) were retrospectively enrolled. DWI were acquired using both conventional DWI and DWI with deep-learning reconstruction (DL DWI). Image quality (overall image quality, anatomic sharpness and details, artifacts, noise, and lesion conspicuity) was scored by two radiologists and compared between two DWI sequences. The apparent diffusion coefficient (ADC) was measured in six locations of the liver parenchyma and focal lesions and compared between two DWI sequences. ResultsThe mean acquisition time for the DL DWI (216.87 ± 49.23 sec) was significantly shorter (P < 0.001) than for conventional DWI (358.69 ± 105.93 sec). DL DWI achieved higher scores than conventional DWI for all qualitative image quality parameters (P < 0.001). DL DWI had a more homogeneous distribution of ADC values throughout the liver, except for the left superior section, compared with conventional DWI. The standard deviations of the ADC values for all hepatic areas were significantly lower in DL DWI than in conventional DWI (all, P < 0.001). The ADC values for the liver parenchyma and focal hepatic lesions were lower in DL DWI than in conventional DWI. ConclusionsDL DWI is a feasible acquisition technique in clinical routines and provides improved image quality and simultaneously significant reduction in scan time compared with conventional DWI.

Development of compact muon spectrometer using multiple pressurized gas Cherenkov radiators

In both particle physics and muon applications, a high-resolution muon momentum measurement capability plays a significant role not only in providing valuable information on the properties of subatomic particles but also in improving the utilizability of cosmic ray muons. Typically, muon momentum is measured by reconstructing a curved muon path using a strong magnetic field and muon trackers. Alternatively, a time-of-flight and Cherenkov ring imager are less frequently applied, especially when there is a need to avoid a magnetic field. However, measurement resolution is much less than that of magnetic spectrometers, approximately 20% whereas it is nearly 4% or less when using magnets and trackers. Here, we propose a different paradigm to estimate muon momentum that utilizes multiple pressurized gas Cherenkov radiators. Using the fact that the refractive index of gas medium varies depending on its pressure and temperature, we can optimize the muon Cherenkov threshold momentum levels for which a muon signal will be detected. In this work, we demonstrate that muon momentum can be estimated with mean resolution of σp/p < 20% and mean classification rate of 90.08% in the momentum range of 0.1 to 10.0 GeV/c by analyzing optical photon signals from each Cherenkov radiator. We anticipate our new spectrometer will significantly improve quality of imaging and reduce scanning time in cosmic ray muon applications by being incorporated with existing instruments.

Tuberculosis conundrum - current and future scenarios: A proposed comprehensive approach combining laboratory, imaging, and computing advances.

Tuberculosis (TB) remains a global threat, with the rise of multiple and extensively drug resistant TB posing additional challenges. The International health community has set various 5-yearly targets for TB elimination: mathematical modelling suggests that a 2050 target is feasible with a strategy combining better diagnostics, drugs, and vaccines to detect and treat both latent and active infection. The availability of rapid and highly sensitive diagnostic tools (Gene-Xpert, TB-Quick) will vastly facilitate population-level identification of TB (including rifampicin resistance and through it, multi-drug-resistant TB). Basic-research advances have illuminated molecular mechanisms in TB, including the protective role of Vitamin D. Also, Mycobacterium tuberculosis impairs the host immune response through epigenetic mechanisms (histone-binding modulation). Imaging will continue to be key, both for initial diagnosis and follow-up. We discuss advances in multiple imaging modalities to evaluate TB tissue changes, such as molecular imaging techniques (including pathogen-specific positron emission tomography imaging agents), non-invasive temporal monitoring, and computing enhancements to improve data acquisition and reduce scan times. Big data analysis and Artificial Intelligence (AI) algorithms, notably in the AI sub-field called “Deep Learning”, can potentially increase the speed and accuracy of diagnosis. Additionally, Federated learning makes multi-institutional/multi-city AI-based collaborations possible without sharing identifiable patient data. More powerful hardware designs - e.g., Edge and Quantum Computing- will facilitate the role of computing applications in TB. However, “Artificial Intelligence needs real Intelligence to guide it!” To have maximal impact, AI must use a holistic approach that incorporates time tested human wisdom gained over decades from the full gamut of TB, i.e., key imaging and clinical parameters, including prognostic indicators, plus bacterial and epidemiologic data. We propose a similar holistic approach at the level of national/international policy formulation and implementation, to enable effective culmination of TB’s endgame, summarizing it with the acronym “TB - REVISITED”.

Accelerated dual-venc 4D flow MRI with variable high-venc spatial resolution for neurovascular applications.

PurposeDual‐velocity encoded (dual‐venc or DV) 4D flow MRI achieves wide velocity dynamic range and velocity‐to‐noise ratio (VNR), enabling accurate neurovascular flow characterization. To reduce scan time, we present interleaved dual‐venc 4D Flow with independently prescribed, prospectively undersampled spatial resolution of the high‐venc (HV) acquisition: Variable Spatial Resolution Dual Venc (VSRDV).MethodsA prototype VSRDV sequence was developed based on a Cartesian acquisition with eight‐point phase encoding, combining PEAK‐GRAPPA acceleration with zero‐filling in phase and partition directions for HV. The VSRDV approach was optimized by varying z, the zero‐filling fraction of HV relative to low‐venc, between 0%–80% in vitro (realistic neurovascular model with pulsatile flow) and in vivo (n = 10 volunteers). Antialiasing precision, mean and peak velocity quantification accuracy, and test–retest reproducibility were assessed relative to reference images with equal‐resolution HV and low venc (z = 0%).ResultsIn vitro results for all z demonstrated an antialiasing true positive rate at least 95% for RPEAK−GRAPPA = 2 and 5, with no linear relationship to z (p = 0.62 and 0.13, respectively). Bland–Altman analysis for z = 20%, 40%, 60%, or 80% versus z = 0% in vitro and in vivo demonstrated no bias >1% of venc in mean or peak velocity values at any RZF. In vitro mean and peak velocity, and in vivo peak velocity, had limits of agreement within 15%.ConclusionVSRDV allows up to 34.8% scan time reduction compared to PEAK‐GRAPPA accelerated DV 4D Flow MRI, enabling large spatial coverage and dynamic range while maintaining VNR and velocity measurement accuracy.