Search

Purpose With rising safety concerns over the use of gadolinium-based contrast agents (GBCAs) in contrast-enhanced MRI, there is a need for dose reduction while maintaining diagnostic capability. This work proposes comprehensive technical solutions for a deep learning (DL) model that predicts contrast-enhanced images of the brain with approximately 10% of the standard dose, across different sites and scanners. Methods The proposed DL model consists of a set of methods that improve the model robustness and generalizability. The steps include multi-planar reconstruction, 2.5D model, enhancement-weighted L1, perceptual, and adversarial losses. The proposed model predicts contrast-enhanced images from corresponding pre-contrast and low-dose images. With IRB approval and informed consent, 640 heterogeneous patient scans (56 train, 13 validation, and 571 test) from 3 institutions consisting of 3D T1-weighted brain images were used. Quantitative metrics were computed and 50 randomly sampled test cases were evaluated by 2 board-certified radiologists. Quantitative tumor segmentation was performed on cases with abnormal enhancements. Ablation study was performed for systematic evaluation of proposed technical solutions. Results The average peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) between full-dose and model prediction were 35.07 ± 3.84 dB and 0.92 ± 0.02 , respectively. Radiologists found the same enhancing pattern in 45/50 (90%) cases; discrepancies were minor differences in contrast intensity and artifacts, with no effect on diagnosis. The average segmentation Dice score between full-dose and synthesized images was 0.88 ± 0.06 (median = 0.91). Conclusions We have proposed a DL model with technical solutions for low-dose contrast-enhanced brain MRI with potential generalizability under diverse clinical settings.

BACKGROUND AND PURPOSE: In acute stroke patients with large vessel occlusions, it would be helpful to be able to predict the difference in the size and location of the final infarct based on the outcome of reperfusion therapy. Our aim was to demonstrate the value of deep learning–based tissue at risk and ischemic core estimation. We trained deep learning models using a baseline MR image in 3 multicenter trials. MATERIALS AND METHODS: Patients with acute ischemic stroke from 3 multicenter trials were identified and grouped into minimal (≤20%), partial (20%-80%), and major (≥80%) reperfusion status based on 4- to 24-hour follow-up MR imaging if available or into unknown status if not. Attention-gated convolutional neural networks were trained with admission imaging as input and the final infarct as ground truth. We explored 3 approaches: 1) separate: train 2 independent models with patients with minimal and major reperfusion; 2) pretraining: develop a single model using patients with partial and unknown reperfusion, then fine-tune it to create 2 separate models for minimal and major reperfusion; and 3) thresholding: use the current clinical method relying on apparent diffusion coefficient and time-to-maximum of the residue function maps. Models were evaluated using area under the curve, the Dice score coefficient, and lesion volume difference. RESULTS: Two hundred thirty-seven patients were included (minimal, major, partial, and unknown reperfusion: n = 52, 80, 57, and 48, respectively). The pretraining approach achieved the highest median Dice score coefficient (tissue at risk = 0.60, interquartile range, 0.43–0.70; core = 0.57, interquartile range, 0.30–0.69). This was higher than the separate approach (tissue at risk = 0.55; interquartile range, 0.41–0.69; P = .01; core = 0.49; interquartile range, 0.35–0.66; P = .04) or thresholding (tissue at risk = 0.56; interquartile range, 0.42–0.65; P = .008; core = 0.46; interquartile range, 0.16–0.54; P < .001). CONCLUSIONS: Deep learning models with fine-tuning lead to better performance for predicting tissue at risk and ischemic core, outperforming conventional thresholding methods.

Purpose: Perfusion imaging assesses target mismatch but requires contrast and processing software. Clinical/diffusion mismatch can miss cases that have target mismatch and could benefit from thrombectomy. We explored whether a neural network can predict hypoperfusion and identify target mismatch from diffusion-weighted imaging (DWI) and clinical information alone. Methods: Acute ischemic stroke cases with baseline MR perfusion and DWI were included from two multi-center trials and one registry for model development and a separate randomized trial for external validation. MR perfusion images were processed by RAPID, which segments Tmax lesion (Tmax≥6s) and the ischemic core lesion (apparent diffusion coefficient [ADC]≤ 620). A 3D U-Net was trained using baseline DWI, ADC, NIH stroke scale, and side of stroke as input, and the union of Tmax and ischemic core segmentation as the ground truth. 5-fold cross-validation was performed for model development cohort. Model performance was evaluated by Dice score coefficient (DSC) and volume difference. Sensitivity and specificity of model target mismatch and clinical/diffusion mismatch criteria from the DAWN were compared, using the DEFUSE 3 target mismatch as reference. Results: 413 patients were included for model development and 46 for external validation. In model development and external validation cohort, the model achieved median DSC of 0.61 (IQR 0.45, 0.71) and 0.62 (IQR 0.53, 0.72); and volume difference of 3 ml (IQR -37, 41) and 7 ml (IQR -24, 32), respectively. Compared to the clinical/diffusion mismatch approach, the model identified target mismatch with a sensitivity of 89.5% vs 49.3%, a specificity of 77.5% vs 89.2% in the model development cohort, and a sensitivity of 95.6% vs 41.3% in external validation cohort. Conclusion: A 3D U-Net can predict hypoperfusion lesions from baseline DWI and clinical information, with more sensitive classification of target mismatch than clinical/diffusion mismatch.

To improve the quality of MRI-based cerebral blood flow (CBF) measurements, a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images to predict gold-standard 15O-water PET CBF images obtained on a simultaneous PET/MRI scanner. The dCNN was trained and tested on 64 scans in 16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation. Fidelity to the PET CBF images and the effects of bias due to training on different cohorts were examined. The dCNN significantly improved CBF image quality compared with ASL alone (mean ± standard deviation): structural similarity index (0.854 ± 0.036 vs. 0.743 ± 0.045 [single-delay] and 0.732 ± 0.041 [multi-delay], P < 0.0001); normalized root mean squared error (0.209 ± 0.039 vs. 0.326 ± 0.050 [single-delay] and 0.344 ± 0.055 [multi-delay], P < 0.0001). The dCNN also yielded mean CBF with reduced estimation error in both HC and PT ( P < 0.001), and demonstrated better correlation with PET. The dCNN trained with the mixed HC and PT cohort performed the best. The results also suggested that models should be trained on cases representative of the target population.

More widespread use of positron emission tomography (PET) imaging is limited by its high cost and radiation dose. Reductions in PET scan time or radiotracer dosage typically degrade diagnostic image quality (DIQ). Deep-learning-based reconstruction may improve DIQ, but such methods have not been clinically evaluated in a realistic multicenter, multivendor environment. In this study, we evaluated the performance and generalizability of a deep-learning-based image-quality enhancement algorithm applied to fourfold reduced-count whole-body PET in a realistic clinical oncologic imaging environment with multiple blinded readers, institutions, and scanner types. We demonstrate that the low-count-enhanced scans were noninferior to the standard scans in DIQ (p 0.05) and overall diagnostic confidence (p 0.001) independent of the underlying PET scanner used. Lesion detection for the low-count-enhanced scans had a high patient-level sensitivity of 0.94 (0.83-0.99) and specificity of 0.98 (0.95-0.99). Interscan kappa agreement of 0.85 was comparable to intrareader (0.88) and pairwise inter-reader agreements (maximum of 0.72). SUV quantification was comparable in the reference regions and lesions (lowest p-value=0.59) and had high correlation (lowest CCC = 0.94). Thus, we demonstrated that deep learning can be used to restore diagnostic image quality and maintain SUV accuracy for fourfold reduced-count PET scans, with interscan variations in lesion depiction, lower than intra- and interreader variations. This method generalized to an external validation set of clinical patients from multiple institutions and scanner types. Overall, this method may enable either dose or exam-duration reduction, increasing safety and lowering the cost of PET imaging.

More widespread use of positron emission tomography (PET) imaging is limited by its high cost and radiation dose. Reductions in PET scan time or radiotracer dosage typically degrade diagnostic image quality (DIQ). Deep-learning-based reconstruction may improve DIQ, but such methods have not been clinically evaluated in a realistic multicenter, multivendor environment. In this study, we evaluated the performance and generalizability of a deep-learning-based image-quality enhancement algorithm applied to fourfold reduced-count whole-body PET in a realistic clinical oncologic imaging environment with multiple blinded readers, institutions, and scanner types. We demonstrate that the low-count-enhanced scans were noninferior to the standard scans in DIQ (p p p-value=0.59) and had high correlation (lowest CCC = 0.94). Thus, we demonstrated that deep learning can be used to restore diagnostic image quality and maintain SUV accuracy for fourfold reduced-count PET scans, with interscan variations in lesion depiction, lower than intra- and interreader variations. This method generalized to an external validation set of clinical patients from multiple institutions and scanner types. Overall, this method may enable either dose or exam-duration reduction, increasing safety and lowering the cost of PET imaging.

Undersampled magnetic resonance image (MRI) reconstruction is typically an ill-posed linear inverse task. The time and resource intensive computations require tradeoffs between accuracy and speed . In addition, state-of-the-art compressed sensing (CS) analytics are not cognizant of the image diagnostic quality . To address these challenges, we propose a novel CS framework that uses generative adversarial networks (GAN) to model the (low-dimensional) manifold of high-quality MR images. Leveraging a mixture of least-squares (LS) GANs and pixel-wise $\ell _{1}/\ell _{2}$ cost, a deep residual network with skip connections is trained as the generator that learns to remove the aliasing artifacts by projecting onto the image manifold. The LSGAN learns the texture details, while the $\ell _{1}/\ell _{2}$ cost suppresses high-frequency noise. A discriminator network, which is a multilayer convolutional neural network (CNN), plays the role of a perceptual cost that is then jointly trained based on high-quality MR images to score the quality of retrieved images. In the operational phase, an initial aliased estimate (e.g., simply obtained by zero-filling) is propagated into the trained generator to output the desired reconstruction. This demands a very low computational overhead. Extensive evaluations are performed on a large contrast-enhanced MR dataset of pediatric patients. Images rated by expert radiologists corroborate that GANCS retrieves higher quality images with improved fine texture details compared with conventional Wavelet-based and dictionary-learning-based CS schemes as well as with deep-learning-based schemes using pixel-wise training. In addition, it offers reconstruction times of under a few milliseconds, which are two orders of magnitude faster than the current state-of-the-art CS-MRI schemes.

Objective: We previously developed two separate deep learning (DL) models to segment the ischemic core and critically hypoperfused tissue on baseline imaging of acute ischemic stroke patients. We aimed to validate the models in an external, multi-center randomized clinical trial (DEFUSE3) and compare with the current clinical standard. Methods: The DL models were previously trained in a separate dataset in which follow-up MRI, obtained at 3-7 days, was used as the reference for critically hypoperfused tissue in patients who did not reperfuse and as the reference for the ischemic core in patients who did reperfuse. For validation, we included DEFUSE3 patients with adequate quality baseline MR perfusion and a 24-hour follow-up DWI scan. The 24-hour DWI lesion served as the reference for ischemic core in patients in the thrombectomy arm and for critically hypoperfused tissue for patients in the medical arm. RAPID was used to generate perfusion maps (Tmax, cerebral blood flow, cerebral blood volume, and mean transient time). The accuracy of segmenting the ischemic core and critically hypoperfused tissue on baseline imaging was compared between the DL approach and the traditional thresholding approach implemented in RAPID. Results: In the 46 patients included for analysis, 24 were in the medical arm and 22 in the thrombectomy arm. Compared to a traditional thresholding method, the DL model segmented the ischemic core more accurately (AUC of 0.92 vs 0.72, p=0.0001 and volume difference of -8ml vs -21ml, p=0.001). Similarly, the DL model segmented critically hypoperfused tissue more accurately (AUC of 0.93 vs 0.80, p Conclusions: The DL-based critical hypoperfusion and ischemic core prediction provides more accurate prediction on final infarct than a commonly used thresholding method in this external validation.

Objective: Ischemic core prediction from CT perfusion (CTP) remains inaccurate compared with gold standard diffusion-weighted imaging (DWI). We evaluated if a deep learning model to predict the DWI lesion from MR perfusion (MRP) could facilitate ischemic core prediction on CTP. Method: Using the multi-center CRISP cohort of acute ischemic stroke patient with CTP before thrombectomy, we included patients with major reperfusion (TICI score≥2b), adequate image quality, and follow-up MRI at 3-7 days. Perfusion parameters including Tmax, mean transient time, cerebral blood flow (CBF), and cerebral blood volume were reconstructed by RAPID software. Core lab experts outlined the stroke lesion on the follow-up MRI. A previously trained MRI model in a separate group of patients was used as a starting point, which used MRP parameters as input and RAPID ischemic core on DWI as ground truth. We fine-tuned this model, using CTP parameters as input, and follow-up MRI as ground truth. Another model was also trained from scratch with only CTP data. 5-fold cross validation was used. Performance of the models was compared with ischemic core (rCBF≤30%) from RAPID software to identify the presence of a large infarct (volume>70 or >100ml). Results: 94 patients in the CRISP trial met the inclusion criteria (mean age 67±15 years, 52% male, median baseline NIHSS 18, median 90-day mRS 2). Without fine-tuning, the MRI model had an agreement of 73% in infarct >70ml, and 69% in >100ml; the MRI model fine-tuned on CT improved the agreement to 77% and 73%; The CT model trained from scratch had agreements of 73% and 71%; All of the deep learning models outperformed the rCBF segmentation from RAPID, which had agreements of 51% and 64%. See Table and figure. Conclusions: It is feasible to apply MRP-based deep learning model to CT. Fine-tuning with CTP data further improves the predictions. All deep learning models predict the stroke lesion after major recanalization better than thresholding approaches based on rCBF.

Fluoroscopic imaging relies on ionizing radiation to provide physicians with high quality video feedback during a surgical operation. Radiation exposure is harmful for both the physician and patient, but reducing dosage results in a much noisier video. We hence propose an algorithm that delivers the same quality video with \(4{\times }\) reduction in radiation dose. Our method is a deep learning approximation to VBM4D, a state-of-the-art video denoiser. Neither VBM4D nor previous deep learning methods are clinically feasible, however, as their high inference runtimes prohibit live display on an operating room monitor. On the other hand, we present a video denoising method which executes orders of magnitude faster while achieving state-of-the-art performance. This provides compelling potential for real-time clinical application in fluoroscopic imaging.

Multi-echo saturation recovery sequence can provide redundant information to synthesize multi-contrast magnetic resonance imaging. Traditional synthesis methods, such as GE’s MAGiC platform, employ a model-fitting approach to generate parameter-weighted contrasts. However, models’ over-simplification, as well as imperfections in the acquisition, can lead to undesirable reconstruction artifacts, especially in T2-FLAIR contrast. To improve the image quality, in this study, a multi-task deep learning model is developed to synthesize multi-contrast neuroimaging jointly using both signal relaxation relationships and spatial information. Compared with previous deep learning-based synthesis, the correlation between different destination contrast is utilized to enhance reconstruction quality. To improve model generalizability and evaluate clinical significance, the proposed model was trained and tested on a large multi-center dataset, including healthy subjects and patients with pathology. Results from both quantitative comparison and clinical reader study demonstrate that the multi-task formulation leads to more efficient and accurate contrast synthesis than previous methods.

Objective: We investigated if deep learning models are able to define the penumbra and ischemic core by comparing models from two training strategies (with and without pre-training) and clinical thresholding criteria (MRI parameter time-to-peak of the residue function [Tmax] and apparent diffusion coefficient [ADC]). Methods: We selected patients from two multicenter stroke trials, with baseline perfusion-weighted imaging (PWI) and diffusion-weighted imaging (DWI) and 3-7 day T2-FLAIR. Based on reperfusion rate calculated from baseline and 24 hr PWI, patients were grouped into unknown (no 24 hr PWI scan), minimal (≤20%), partial (20%-80%), and major (≥80%) reperfusion. Attention-gated U-net structure was selected for training, with eight image channels from baseline PWI/DWI as inputs and the infarct lesion manually segmented on T2-FLAIR as ground truth. Two training strategies were used: (1) training two models separately in minimal and major reperfusion patients; (2) pre-training a model using patients with partial and unknown reperfusion, then fine-tuning two models using minimal and major reperfusion patients, respectively. Prediction was evaluated by Dice score coefficient (DSC), and lesion volume error at an optimal threshold. In minimal and major reperfusion patients, the deep learning models and Tmax and ADC thresholding were compared using paired sample Wilcoxon test. Results: 182 patients were included (85 males, age 65±16 yrs, baseline NIHSS 15 IQR 10-19), with a breakdown of minimal/major/partial/unknown reperfusion status of 32/65/43/42 patients, respectively. The pre-training approach performed the best among all approaches (Table 1, Figure 1). Conclusion: Deep learning models to predict penumbra and ischemic core are best trained using general pre-training on a wide range of stroke cases followed by fine-tuning on the extreme cases. This method outperforms conventional DWI-PWI mismatch inspired thresholding approaches.

Positron Emission Tomography (PET) has been widely applied in clinics for diagnosis of cancer, cardiovascular disease, neurological disorder, and other challenging diseases. Radiotracers are injected into patients prior to PET exams, introducing inevitable radiation risks. While recent deep learning methods have shown to enable low-dose PET without compromising image quality, their performance are often limited when the amplified noise in low-dose scans becomes indistinguishable from high-intensity small abnormality. In this paper, we propose a noise-aware dual Res-UNet framework to enable low dose PET scans and achieve the image quality comparable to that from standard-dose PET scans. Specifically, noise-aware dual Res-UNets are designed to identify the location of high intensity noise in the low dose PET images first, followed by an image reconstruction network incorporating the estimated noise attention map to reconstruct the high quality standard-dose PET image. In order to better reduce the Poisson distribution noise, a general and adaptive robust loss is applied. Experimental results show that our method can outperform other state-of-the-art methods quantitatively and qualitatively and can be applied on real clinical application.

To improve the quality of MRI-based cerebral blood flow (CBF) measurements, a deep convolutional neural network (dCNN) was trained to combine single- and multi-delay arterial spin labeling (ASL) and structural images to predict gold-standard (15)O-water PET CBF images obtained on a simultaneous PET/MRI scanner. The dCNN was trained and tested on 64 scans in 16 healthy controls (HC) and 16 cerebrovascular disease patients (PT) with 4-fold cross-validation. Fidelity to the PET CBF images and the effects of bias due to training on different cohorts were examined. The dCNN significantly improved CBF image quality compared with ASL alone (mean ± standard deviation): structural similarity index (0.854 ± 0.036 vs. 0.743 ± 0.045 [single-delay] and 0.732 ± 0.041 [multi-delay], P

BACKGROUND There are concerns over gadolinium deposition from gadolinium-based contrast agents (GBCA) administration. PURPOSE To reduce gadolinium dose in contrast-enhanced brain MRI using a deep learning method. STUDY TYPE Retrospective, crossover. POPULATION Sixty patients receiving clinically indicated contrast-enhanced brain MRI. SEQUENCE 3D T1 -weighted inversion-recovery prepped fast-spoiled-gradient-echo (IR-FSPGR) imaging was acquired at both 1.5T and 3T. In 60 brain MRI exams, the IR-FSPGR sequence was obtained under three conditions: precontrast, postcontrast images with 10% low-dose (0.01mmol/kg) and 100% full-dose (0.1 mmol/kg) of gadobenate dimeglumine. We trained a deep learning model using the first 10 cases (with mixed indications) to approximate full-dose images from the precontrast and low-dose images. Synthesized full-dose images were created using the trained model in two test sets: 20 patients with mixed indications and 30 patients with glioma. ASSESSMENT For both test sets, low-dose, true full-dose, and the synthesized full-dose postcontrast image sets were compared quantitatively using peak-signal-to-noise-ratios (PSNR) and structural-similarity-index (SSIM). For the test set comprised of 20 patients with mixed indications, two neuroradiologists scored blindly and independently for the three postcontrast image sets, evaluating image quality, motion-artifact suppression, and contrast enhancement compared with precontrast images. STATISTICAL ANALYSIS Results were assessed using paired t-tests and noninferiority tests. RESULTS The proposed deep learning method yielded significant (n = 50, P 5 dB PSNR gains and >11.0% SSIM). Ratings on image quality (n = 20, P = 0.003) and contrast enhancement (n = 20, P

PURPOSE: To improve the image quality of highly accelerated multi-channel MRI data by learning a joint variational network that reconstructs multiple clinical contrasts jointly. METHODS: Data from our multi-contrast acquisition was embedded into the variational network architecture where shared anatomical information is exchanged by mixing the input contrasts. Complementary k-space sampling across imaging contrasts and Bunch-Phase/Wave-Encoding were used for data acquisition to improve the reconstruction at high accelerations. At 3T, our joint variational network approach across T1w, T2w and T2-FLAIR-weighted brain scans was tested for retrospective under-sampling at R=6 (2D) and R=4×4 (3D) acceleration. Prospective acceleration was also performed for 3D data where the combined acquisition time for whole brain coverage at 1 mm isotropic resolution across three contrasts was less than three minutes. RESULTS: Across all test datasets, our joint multi-contrast network better preserved fine anatomical details with reduced image-blurring when compared to the corresponding single-contrast reconstructions. Improvement in image quality was also obtained through complementary k-space sampling and Bunch-Phase/Wave-Encoding where the synergistic combination yielded the overall best performance as evidenced by exemplarily slices and quantitative error metrics. CONCLUSION: By leveraging shared anatomical structures across the jointly reconstructed scans, our joint multi-contrast approach learnt more efficient regularizers which helped to retain natural image appearance and avoid over-smoothing. When synergistically combined with advanced encoding techniques, the performance was further improved, enabling up to R=16-fold acceleration with good image quality. This should help pave the way to very rapid high-resolution brain exams.

Purpose Our goal was to use a generative adversarial network (GAN) with feature matching and task-specific perceptual loss to synthesize standard-dose amyloid Positron emission tomography (PET) images of high quality and including accurate pathological features from ultra-low-dose PET images only. Methods Forty PET datasets from 39 participants were acquired with a simultaneous PET/MRI scanner following injection of 330 ± 30 MBq of the amyloid radiotracer 18F-florbetaben. The raw list-mode PET data were reconstructed as the standard-dose ground truth and were randomly undersampled by a factor of 100 to reconstruct 1% low-dose PET scans. A 2D encoder-decoder network was implemented as the generator to synthesize a standard-dose image and a discriminator was used to evaluate them. The two networks contested with each other to achieve high-visual quality PET from the ultra-low-dose PET. Multi-slice inputs were used to reduce noise by providing the network with 2.5D information. Feature matching was applied to reduce hallucinated structures. Task-specific perceptual loss was designed to maintain the correct pathological features. The image quality was evaluated by peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and root mean square error (RMSE) metrics with and without each of these modules. Two expert radiologists were asked to score image quality on a 5-point scale and identified the amyloid status (positive or negative). Results With only low-dose PET as input, the proposed method significantly outperformed Chen et al.'s method (Chen et al. Radiology. 2018;290:649-656) (which shows the best performance in this task) with the same input (PET-only model) by 1.87 dB in PSNR, 2.04% in SSIM, and 24.75% in RMSE. It also achieved comparable results to Chen et al.'s method which used additional magnetic resonance imaging (MRI) inputs (PET-MR model). Experts' reading results showed that the proposed method could achieve better overall image quality and maintain better pathological features indicating amyloid status than both PET-only and PET-MR models proposed by Chen et al. CONCLUSION: Standard-dose amyloid PET images can be synthesized from ultra-low-dose images using GAN. Applying adversarial learning, feature matching, and task-specific perceptual loss are essential to ensure image quality and the preservation of pathological features.

Generative Adversarial Network (GAN) has demonstrated great potentials in computer vision tasks such as image restoration. However, image restoration for specific scenarios, such as medical image enhancement is still facing challenge: How to ensure the visually plausible results while not containing hallucinated features that might jeopardize downstream tasks such as pathology identification? Here, we propose Task-GAN, a generalized model for medical reconstruction problem. A task-specific network that captures the diagnostic/pathology features, was added to couple the GAN based image reconstruction framework. Validated on multiple medical datasets, we demonstrated that the proposed method leads to improved deep learning based image reconstruction while preserving the detailed structure and diagnostic features.

Fast reconstruction of under-sampled acquisitions has always been a central issue in MRI reconstruction. Recently years has seen multiple studies using deep learning as a de-aliasing framework to restore the aliased image. However, restoration of fine details is still problematic, especially when dealing with noisy image datasets. Sparked by the Fourier transform relationship, this work proposed and tested a new hypothesis: can regularization be directly added in the frequency domain to correct the high-frequency imperfection? To achieve this, discriminative networks are applied in both the image domain and the frequency domain (so-called dual-domain GAN). Evaluation on multiple datasets proved that the dual-domain GAN approach is an effective way to improve the quality of accelerated MR reconstruction.

Supplemental material, JCB888123 Supplementary material for Predicting 15O-Water PET cerebral blood flow maps from multi-contrast MRI using a deep convolutional neural network with evaluation of training cohort bias by Jia Guo, Enhao Gong, Audrey P Fan, Maged Goubran, Mohammad M Khalighi and Greg Zaharchuk in Journal of Cerebral Blood Flow & Metabolism

PURPOSE: To reduce radiotracer requirements for amyloid PET/MRI without sacrificing diagnostic quality by using deep learning methods. MATERIALS AND METHODS: Forty data sets from 39 patients (mean age ± standard deviation [SD], 67 years ± 8), including 16 male patients and 23 female patients (mean age, 66 years ± 6 and 68 years ± 9, respectively), who underwent simultaneous amyloid (fluorine 18 [(18)F]–florbetaben) PET/MRI examinations were acquired from March 2016 through October 2017 and retrospectively analyzed. One hundredth of the raw list-mode PET data were randomly chosen to simulate a low-dose (1%) acquisition. Convolutional neural networks were implemented with low-dose PET and multiple MR images (PET-plus-MR model) or with low-dose PET alone (PET-only) as inputs to predict full-dose PET images. Quality of the synthesized images was evaluated while Bland-Altman plots assessed the agreement of regional standard uptake value ratios (SUVRs) between image types. Two readers scored image quality on a five-point scale (5 = excellent) and determined amyloid status (positive or negative). Statistical analyses were carried out to assess the difference of image quality metrics and reader agreement and to determine confidence intervals (CIs) for reading results. RESULTS: The synthesized images (especially from the PET-plus-MR model) showed marked improvement on all quality metrics compared with the low-dose image. All PET-plus-MR images scored 3 or higher, with proportions of images rated greater than 3 similar to those for the full-dose images (−10% difference [eight of 80 readings], 95% CI: −15%, −5%). Accuracy for amyloid status was high (71 of 80 readings [89%]) and similar to intrareader reproducibility of full-dose images (73 of 80 [91%]). The PET-plus-MR model also had the smallest mean and variance for SUVR difference to full-dose images. CONCLUSION: Simultaneously acquired MRI and ultra–low-dose PET data can be used to synthesize full-dose–like amyloid PET images. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Catana in this issue.

When treatment decisions are being made for patients with acute ischemic stroke, timely and accurate outcome prediction plays an important role. The optimal rehabilitation strategy also relies on long-term outcome predictions. The decision-making process involves numerous biomarkers including imaging features and demographic information. The objective of this study was to integrate common stroke biomarkers using machine learning methods and predict patient recovery outcome at 90 days.A total of 512 patients were enrolled in this retrospective study. Extreme gradient boosting (XGB) and gradient boosting machine (GBM) models were used to predict modified Rankin scale (mRS) scores at 90 days using biomarkers available at admission and 24 hours. Feature selections were performed using a greedy algorithm. Fivefold cross validation was applied to estimate model performance.For binary prediction of an mRS score of greater than 2 using biomarkers available at admission, XGB and GBM had an AUC of 0.746 and 0.748, respectively. Adding the National Institutes of Health Stroke Score at 24 hours and performing feature selection improved the AUC of XGB to 0.884 and the AUC of GBM to 0.877. With the addition of the recanalization outcome, XGB's AUC improved to 0.807 for nonrecanalized patients and dropped to 0.670 for recanalized patients. GBM's AUC improved to 0.781 for nonrecanalized patients and dropped to 0.655 for recanalized patients.Decision tree-based GBMs can predict the recovery outcome of stroke patients at admission with a high AUC. Breaking down the patient groups on the basis of recanalization and nonrecanalization can potentially help with the treatment decision process.

Deep learning is a form of machine learning using a convolutional neural network architecture that shows tremendous promise for imaging applications. It is increasingly being adapted from its original demonstration in computer vision applications to medical imaging. Because of the high volume and wealth of multimodal imaging information acquired in typical studies, neuroradiology is poised to be an early adopter of deep learning. Compelling deep learning research applications have been demonstrated, and their use is likely to grow rapidly. This review article describes the reasons, outlines the basic methods used to train and test deep learning models, and presents a brief overview of current and potential clinical applications with an emphasis on how they are likely to change future neuroradiology practice. Facility with these methods among neuroimaging researchers and clinicians will be important to channel and harness the vast potential of this new method.

Introduction: Predicting the final infarct for acute stroke from acute imaging is valuable for triage and prognosis (1). We developed and tested a Deep Learning model, based on a 3D Convolutional Neural Network (CNN) architecture, to predict the lesion based on a 30-90 day scan from acute images. Method: The model has 3 advantages compared with traditional CNN (Figure): 1. 3D CNN more efficiently utilizes spatial information across multiple slices. 2. Patches instead of entire images are used as inputs. With a patch-based approach, we avoid the impact of minimally-relevant information from distant voxels. Training on patches also prevents over-fitting by augmenting each image into thousands of samples. 3. Multi-scale structure is used by processing patches with different resolutions. Segmentation based on a single scale image cannot fully capture varying local information and may miss contextual information. We use two scales of patches to learn both local and global context. The model was trained and tested using the MICCAI ISLES 2017 challenge dataset (2), which consists of 43 cases with acute ADC and PWI maps paired with annotated final infarct segmentations at day 30-90. Dice score coefficient (DSC) and AUC for segmentation were used as quality metrics. Results: Using 70% of dataset for training and rest for testing, we achieved a DSC of 0.43±0.18 and 0.90 for AUC. As a comparison, the winning entry for the ISLES 2016 challenge achieved a DSC of 0.31, while several previous research works (threshold based, cluster based, Generalized Linear Model, etc.) achieve up to 0.84 AUC (1). Conclusion: We demonstrate that a deep learning approach can predict the final stroke lesion based on acute diffusion and perfusion neuroimaging data. Given its inherent speed, high performance, and capacity for further training, deep learning is a promising method for stroke lesion outcome prediction. References 1. Rekik et al., Neuroimage Clin 2012; 2. http://www.isles-challenge.org/

Deep neural networks have demonstrated promising potential for the field of medical image reconstruction. In this work, an MRI reconstruction algorithm, which is referred to as quantitative susceptibility mapping (QSM), has been developed using a deep neural network in order to perform dipole deconvolution, which restores magnetic susceptibility source from an MRI field map. Previous approaches of QSM require multiple orientation data (e.g. Calculation of Susceptibility through Multiple Orientation Sampling or COSMOS) or regularization terms (e.g. Truncated K-space Division or TKD; Morphology Enabled Dipole Inversion or MEDI) to solve the ill-conditioned deconvolution problem. Unfortunately, they either require long multiple orientation scans or suffer from artifacts. To overcome these shortcomings, a deep neural network, QSMnet, is constructed to generate a high quality susceptibility map from single orientation data. The network has a modified U-net structure and is trained using gold-standard COSMOS QSM maps. 25 datasets from 5 subjects (5 orientation each) were applied for patch-wise training after doubling the data using augmentation. Two additional datasets of 5 orientation data were used for validation and test (one dataset each). The QSMnet maps of the test dataset were compared with those from TKD and MEDI for image quality and consistency in multiple head orientations. Quantitative and qualitative image quality comparisons demonstrate that the QSMnet results have superior image quality to those of TKD or MEDI and have comparable image quality to those of COSMOS. Additionally, QSMnet maps reveal substantially better consistency across the multiple orientations than those from TKD or MEDI. As a preliminary application, the network was tested for two patients. The QSMnet maps showed similar lesion contrasts with those from MEDI, demonstrating potential for future applications., Comment: This work is accepted in neuroimage on 8 June, 2018 and soon will be published. The pubmed link is https://www.ncbi.nlm.nih.gov/pubmed/29894829

Recovering images from highly undersampled measurements has a wide range of applications across imaging sciences. State-of-the-art analytics however are not aware of the image perceptual quality, and demand iterative algorithms that incur significant computational overhead. To sidestep these hurdles, this paper brings forth a novel compressive imaging framework using deep neural networks that approximates a low-dimensional manifold of images using generative adversarial networks. To ensure the images are consistent with the measurements a recurrent GAN (RGAN) architecture is deployed that consists of multiple alternative blocks of generator networks and affine projection, which is then followed by a discriminator network to score the perceptual quality of the generated images. A deep residual network with skip connections is used for the generator, while the discriminator is a multilayer perceptron. Experiments performed with real-world contrast enhanced MRI data corroborate the diagnostic quality of the retrieved images relative to state-of-the-art CS schemes. In addition, it achieves about two-orders of magnitude faster reconstruction.

There are concerns over gadolinium deposition from gadolinium-based contrast agents (GBCA) administration.To reduce gadolinium dose in contrast-enhanced brain MRI using a deep learning method.Retrospective, crossover.Sixty patients receiving clinically indicated contrast-enhanced brain MRI.3D TFor both test sets, low-dose, true full-dose, and the synthesized full-dose postcontrast image sets were compared quantitatively using peak-signal-to-noise-ratios (PSNR) and structural-similarity-index (SSIM). For the test set comprised of 20 patients with mixed indications, two neuroradiologists scored blindly and independently for the three postcontrast image sets, evaluating image quality, motion-artifact suppression, and contrast enhancement compared with precontrast images.Results were assessed using paired t-tests and noninferiority tests.The proposed deep learning method yielded significant (n = 50, P 0.001) improvements over the low-dose images (5 dB PSNR gains and11.0% SSIM). Ratings on image quality (n = 20, P = 0.003) and contrast enhancement (n = 20, P 0.001) were significantly increased. Compared to true full-dose images, the synthesized full-dose images have a slight but not significant reduction in image quality (n = 20, P = 0.083) and contrast enhancement (n = 20, P = 0.068). Slightly better (n = 20, P = 0.039) motion-artifact suppression was noted in the synthesized images. The noninferiority test rejects the inferiority of the synthesized to true full-dose images for image quality (95% CI: -14-9%), artifacts suppression (95% CI: -5-20%), and contrast enhancement (95% CI: -13-6%).With the proposed deep learning method, gadolinium dose can be reduced 10-fold while preserving contrast information and avoiding significant image quality degradation.3 Technical Efficacy: Stage 5 J. MAGN. RESON. IMAGING 2018;48:330-340.

Purpose A typical clinical MR examination includes multiple scans to acquire images with different contrasts for complementary diagnostic information. The multicontrast scheme requires long scanning time. The combination of partially parallel imaging and compressed sensing (CS-PPI) has been used to reconstruct accelerated scans. However, there are several unsolved problems in existing methods. The target of this work is to improve existing CS-PPI methods for multicontrast imaging, especially for two-dimensional imaging. Theory and Methods If the same field of view is scanned in multicontrast imaging, there is significant amount of sharable information. It is proposed in this study to use manifold sharable information among multicontrast images to enhance CS-PPI in a sequential way. Coil sensitivity information and structure based adaptive regularization, which were extracted from previously reconstructed images, were applied to enhance the following reconstructions. The proposed method is called Parallel-imaging and compressed-sensing Reconstruction Of Multicontrast Imaging using SharablE information (PROMISE). Results Using L1-SPIRiT as a CS-PPI example, results on multicontrast brain and carotid scans demonstrated that lower error level and better detail preservation can be achieved by exploiting manifold sharable information. Besides, the privilege of PROMISE still exists while there is interscan motion. Conclusion Using the sharable information among multicontrast images can enhance CS-PPI with tolerance to motions. Magn Reson Med 73:523–535, 2015. © 2014 Wiley Periodicals, Inc.

Assembly of protein complexes is a key mechanism for achieving spatial and temporal coordination in processes involving many enzymes. Growth of rod-shaped bacteria is a well-studied example requiring such coordination; expansion of the cell wall is thought to involve coordination of the activity of synthetic enzymes with the cytoskeleton via a stable complex. Here, we use single-molecule tracking to demonstrate that the bacterial actin homolog MreB and the essential cell wall enzyme PBP2 move on timescales orders of magnitude apart, with drastically different characteristic motions. Our observations suggest that PBP2 interacts with the rest of the synthesis machinery through a dynamic cycle of transient association. Consistent with this model, growth is robust to large fluctuations in PBP2 abundance. In contrast to stable complex formation, dynamic association of PBP2 is less dependent on the function of other components of the synthesis machinery, and buffers spatially distributed growth against fluctuations in pathway component concentrations and the presence of defective components. Dynamic association could generally represent an efficient strategy for spatiotemporal coordination of protein activities, especially when excess concentrations of system components are inhibitory to the overall process or deleterious to the cell.