Radial Undersampled MRI Reconstruction Using Deep Learning With Mutual Constraints Between Real and Imaginary Components of K-Space.

The deep learning method is an efficient solution for improving the quality of undersampled magnetic resonance (MR) image reconstruction while reducing lengthy data acquisition. Most deep learning methods neglect the mutual constraints between the real and imaginary components of complex-valued k-space data. In this paper, a new complex-valued convolutional neural network, namely, Dense-U-Dense Net (DUD-Net), is proposed to interpolate the undersampled k-space data and reconstruct MR images. The proposed network comprises dense layers, U-Net, and other dense layers in sequence. The dense layers are used to simulate the mutual constraints between real and imaginary components, and U-Net performs feature sparsity and interpolation estimation for k-space data. Two MRI datasets were used to evaluate the proposed method: brain magnitude-only MR images and knee complex-valued k-space data. Several operations were conducted for data preprocessing. First, the complex-valued MR images were synthesized by phase modulation on magnitude-only images. Second, a radial trajectory based on the golden angle was used for k-space undersampling, whereby a reversible normalization method was proposed to balance the distribution of positive and negative values in k-space data. The optimal performance of DUD-Net was demonstrated based on a quantitative evaluation of inter-method and intra-method comparisons. When compared with other methods, significant improvements were achieved, PSNRs were increased by 10.78 and 5.74dB, whereas RMSEs were decreased by 71.53% and 30.31% for magnitude and phase image, respectively. It is concluded that DUD-Net significantly improves the performance of MR image reconstruction.

Feasibility study on the clinical application of CT-based synthetic brain T1-weighted MRI: comparison with conventional T1-weighted MRI.

OBJECTIVES: This study aimed to examine the equivalence of computed tomography (CT)-based synthetic T1-weighted imaging (T1WI) to conventional T1WI for the quantitative assessment of brain morphology. MATERIALS AND METHODS: This prospective study examined 35 adult patients undergoing brain magnetic resonance imaging (MRI) and CT scans. An image synthesis method based on a deep learning model was used to generate synthetic T1WI (sT1WI) from CT data. Two senior radiologists used sT1WI and conventional T1WI on separate occasions to independently measure clinically relevant brain morphological parameters. The reliability and consistency between conventional and synthetic T1WI were assessed using statistical consistency checks, comprising intra-reader, inter-reader, and inter-method agreement. RESULTS: The intra-reader, inter-reader, and inter-method reliability and variability mostly exhibited the desired performance, except for several poor agreements due to measurement differences between the radiologists. All the measurements of sT1WI were equivalent to that of T1WI at 5% equivalent intervals. CONCLUSION: This study demonstrated the equivalence of CT-based sT1WI to conventional T1WI for quantitatively assessing brain morphology, thereby providing more information on imaging diagnosis with a single CT scan. CLINICAL RELEVANCE STATEMENT: Real-time synthesis of MR images from CT scans reduces the time required to acquire MR signals, improving the efficiency of the treatment planning system and providing benefits in the clinical diagnosis of patients with contraindications such as presence of metal implants or claustrophobia. KEY POINTS: • Deep learning-based image synthesis methods generate synthetic T1-weighted imaging from CT scans. • The equivalence of synthetic T1-weighted imaging and conventional MRI for quantitative brain assessment was investigated. • Synthetic T1-weighted imaging can provide more information per scan and be used in preoperative diagnosis and radiotherapy.

The automatic evaluation of steno-occlusive changes in time-of-flight magnetic resonance angiography of moyamoya patients using a 3D coordinate attention residual network.

Background: Moyamoya disease (MMD) is a rare cerebrovascular occlusive disease with progressive stenosis of the terminal portion of internal cerebral artery (ICA) and its main branches, which can cause complications, such as high risks of disability and increased mortality. Accurate and timely diagnosis may be difficult for physicians who are unfamiliar to MMD. Therefore, this study aims to achieve a preoperative deep-learning-based evaluation of MMD by detecting steno-occlusive changes in the middle cerebral artery or distal ICA areas. Methods: A fine-tuned deep learning model was developed using a three-dimensional (3D) coordinate attention residual network (3D CA-ResNet). This study enrolled 50 preoperative patients with MMD and 50 controls, and the corresponding time of flight magnetic resonance angiography (TOF-MRA) imaging data were acquired. The 3D CA-ResNet was trained based on sub-volumes and tested using patch-based and subject-based methods. The performance of the 3D CA-ResNet, as evaluated by the area under the curve (AUC) of receiving-operator characteristic, was compared with that of three other conventional 3D networks. Results: With the resulting network, the patch-based test achieved an AUC value of 0.94 for the 3D CA-ResNet in 480 patches from 10 test patients and 10 test controls, which is significantly higher than the results of the others. The 3D CA-ResNet correctly classified the MMD patients and normal healthy controls, and the vascular lesion distribution in subjects with the disease was investigated by generating a stenosis probability map and 3D vascular structure segmentation. Conclusions: The results demonstrated the reliability of the proposed 3D CA-ResNet in detecting stenotic areas on TOF-MRA imaging, and it outperformed three other models in identifying vascular steno-occlusive changes in patients with MMD.

Fusing hand-crafted and deep-learning features in a convolutional neural network model to identify prostate cancer in pathology images.

Prostate cancer can be diagnosed by prostate biopsy using transectal ultrasound guidance. The high number of pathology images from biopsy tissues is a burden on pathologists, and analysis is subjective and susceptible to inter-rater variability. The use of machine learning techniques could make prostate histopathology diagnostics more precise, consistent, and efficient overall. This paper presents a new classification fusion network model that was created by fusing eight advanced image features: seven hand-crafted features and one deep-learning feature. These features are the scale-invariant feature transform (SIFT), speeded up robust feature (SURF), oriented features from accelerated segment test (FAST) and rotated binary robust independent elementary features (BRIEF) (ORB) of local features, shape and texture features of the cell nuclei, the histogram of oriented gradients (HOG) feature of the cavities, a color feature, and a convolution deep-learning feature. Matching, integrated, and fusion networks are the three essential components of the proposed deep-learning network. The integrated network consists of both a backbone and an additional network. When classifying 1100 prostate pathology images using this fusion network with different backbones (ResNet-18/50, VGG-11/16, and DenseNet-121/201), we discovered that the proposed model with the ResNet-18 backbone achieved the best performance in terms of the accuracy (95.54%), specificity (93.64%), and sensitivity (97.27%) as well as the area under the receiver operating characteristic curve (98.34%). However, each of the assessment criteria for these separate features had a value lower than 90%, which demonstrates that the suggested model combines differently derived characteristics in an effective manner. Moreover, a Grad-CAM++ heatmap was used to observe the differences between the proposed model and ResNet-18 in terms of the regions of interest. This map showed that the proposed model was better at focusing on cancerous cells than ResNet-18. Hence, the proposed classification fusion network, which combines hand-crafted features and a deep-learning feature, is useful for computer-aided diagnoses based on pathology images of prostate cancer. Because of the similarities in the feature engineering and deep learning for different types of pathology images, the proposed method could be used for other pathology images, such as those of breast, thyroid cancer.

Synthesis of magnetic resonance images from computed tomography data using convolutional neural network with contextual loss function.

Background: Magnetic resonance imaging (MRI) images synthesized from computed tomography (CT) data can provide more detailed information on pathological structures than that of CT data alone; thus, the synthesis of MRI has received increased attention especially in medical scenarios where only CT images are available. A novel convolutional neural network (CNN) combined with a contextual loss function was proposed for synthesis of T1- and T2-weighted images (T1WI and T2WI) from CT data. Methods: A total of 5,053 and 5,081 slices of T1WI and T2WI, respectively were selected for the dataset of CT and MRI image pairs. Affine registration, image denoising, and contrast enhancement were done on the aforementioned multi-modality medical image dataset comprising T1WI, T2WI, and CT images of the brain. A deep CNN was then proposed by modifying the ResNet structure to constitute the encoder and decoder of U-Net, called double ResNet-U-Net (DRUNet). Three different loss functions were utilized to optimize the parameters of the proposed models: mean squared error (MSE) loss, binary crossentropy (BCE) loss, and contextual loss. Statistical analysis of the independent-sample t-test was conducted by comparing DRUNets with different loss functions and different network layers. Results: DRUNet-101 with contextual loss yielded higher values of peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and Tenengrad function (i.e., 34.25±2.06, 0.97±0.03, and 17.03±2.75 for T1WI and 33.50±1.08, 0.98±0.05, and 19.76±3.54 for T2WI respectively). The results were statistically significant at P<0.001 with a narrow confidence interval of difference, indicating the superiority of DRUNet-101 with contextual loss. In addition, both image zooming and difference maps presented for the final synthetic MR images visually reflected the robustness of DRUNet-101 with contextual loss. The visualization of convolution filters and feature maps showed that the proposed model can generate synthetic MR images with high-frequency information. Conclusions: The results demonstrated that DRUNet-101 with contextual loss function provided better high-frequency information in synthetic MR images compared with the other two functions. The proposed DRUNet model has a distinct advantage over previous models in terms of PSNR, SSIM, and Tenengrad score. Overall, DRUNet-101 with contextual loss is recommended for synthesizing MR images from CT scans.

Computer-Aided Diagnosis of Spinal Tuberculosis From CT Images Based on Deep Learning With Multimodal Feature Fusion.

Background: Spinal tuberculosis (TB) has the highest incidence in remote plateau areas, particularly in Tibet, China, due to inadequate local healthcare services, which not only facilitates the transmission of TB bacteria but also increases the burden on grassroots hospitals. Computer-aided diagnosis (CAD) is urgently required to improve the efficiency of clinical diagnosis of TB using computed tomography (CT) images. However, classical machine learning with handcrafted features generally has low accuracy, and deep learning with self-extracting features relies heavily on the size of medical datasets. Therefore, CAD, which effectively fuses multimodal features, is an alternative solution for spinal TB detection. Methods: A new deep learning method is proposed that fuses four elaborate image features, specifically three handcrafted features and one convolutional neural network (CNN) feature. Spinal TB CT images were collected from 197 patients with spinal TB, from 2013 to 2020, in the People's Hospital of Tibet Autonomous Region, China; 3,000 effective lumbar spine CT images were randomly screened to our dataset, from which two sets of 1,500 images each were classified as tuberculosis (positive) and health (negative). In addition, virtual data augmentation is proposed to enlarge the handcrafted features of the TB dataset. Essentially, the proposed multimodal feature fusion CNN consists of four main sections: matching network, backbone (ResNet-18/50, VGG-11/16, DenseNet-121/161), fallen network, and gated information fusion network. Detailed performance analyses were conducted based on the multimodal features, proposed augmentation, model stability, and model-focused heatmap. Results: Experimental results showed that the proposed model with VGG-11 and virtual data augmentation exhibited optimal performance in terms of accuracy, specificity, sensitivity, and area under curve. In addition, an inverse relationship existed between the model size and test accuracy. The model-focused heatmap also shifted from the irrelevant region to the bone destruction caused by TB. Conclusion: The proposed augmentation effectively simulated the real data distribution in the feature space. More importantly, all the evaluation metrics and analyses demonstrated that the proposed deep learning model exhibits efficient feature fusion for multimodal features. Our study provides a profound insight into the preliminary auxiliary diagnosis of spinal TB from CT images applicable to the Tibetan area.

Complete genome sequence of a novel virga-like virus infecting Hevea brasiliensis.

Here, we report the complete genome sequence and organization of a novel virus detected in rubber trees (Hevea brasiliensis). Because the infected plants were asymptomatic, this virus was tentatively named "rubber tree latent virus 1" (RTLV1). The full genome of RTLV1 is 9,422 nt in length and contains three open reading frames with a 157-nt 5' untranslated region (UTR) and a 316-nt 3' UTR. The replicase shares the highest amino acid (aa) sequence identity (32.62%), with only 31% query coverage, with the replicase of Hubei virga-like virus 11. Phylogenetic analysis based on the aa sequence of ORF1 showed that RTLV1 clustered with unclassified members of the family Virgaviridae in a clade that was not closely related to any genus in this family.

Post COVID-19: Health crisis management for the cruise industry.

The cruise industry is gravely affected by the COVID-19 pandemic due to rising public health concerns. This study combines and examines health crisis management and marketing theories to address public health concerns and improve the usage of cruise services. Combining social exchange theory, customers' perceived value theory, and trust theory, a theoretical model is proposed. Survey data (n = 376) are then collected through an online survey that is conducted on the Chinese tourism market. The finding shows that quality management, health management, social and communication strategies, and financial strategies contribute to customers' perceived value of cruise service. In addition, perceived value directly and indirectly influences customers' intention to use cruise service through trust in cruise company's pandemic management capability. This study expands the current literature on cruise crisis recovery and provides recommendations for policy and strategy formulation for the cruise industry to cope with the pandemic by focusing on public health concerns and psychology.

Optimal Diffusion Gradient Encoding Scheme for Diffusion Tensor Imaging Based on Golden Ratio.

BACKGROUND: The accuracy of the estimated diffusion tensor elements can be improved by using a well-chosen magnetic resonance imaging (MRI) diffusion gradient encoding scheme (DGES). Conversely, diffusion tensor imaging (DTI) is typically challenged by the subject's motion during data acquisition and results in corrupted image data. PURPOSE: To identify a reliable DGES based on the golden ratio (GR) that can generate an arbitrary number of uniformly distributed directions to precisely estimate the DTI parameters of partially acquired datasets owing to subject motion. STUDY TYPE: Prospective. POPULATION: Simulations study; three healthy volunteers. FIELD STRENGTH/SEQUENCE: 3 T/DTI data were obtained using a single-shot echo planar imaging sequence. STATISTICAL TESTS: A paired sample t-test and the Wilcoxon test were used, P < 0.05 was considered statistically significant. ASSESSMENT: Two corrupted scenarios A and B were considered and evaluated. For the simulation study, the GR DGES and generated subsets were compared with the Jones and spiral DGESs by electric potential (EP) and condition number (CN). For the human study, the specific subsets A and B selected from scenarios A and B were used for MRI to evaluate fractional anisotropic (FA) map. RESULTS: For the simulation study, the EPs of the GR (14034.25 ± 12957.24) DGES were significantly lower than the Jones (15112.81 ± 13926.08) and spiral (14297.49 ± 13232.94) DGESs. CN variations of GR (1.633 ± 0.024) DGES were significantly lower than Jones (1.688 ± 0.119) and spiral (4.387 ± 2.915) DGESs. For the human study, GR (0.008 ± 0.020) DGES performed similarly with Jones (0.008 ± 0.022) DGES and was superior to spiral (0.022 ± 0.054) DGES in the FA map error. DATA CONCLUSION: The GR DGES ensured that directions of the complete sets and subsets were uniform. The GR DGES had lower error propagation sensitivity, which can help image infants or patients who cannot stay still during scanning. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 1.

Prevalence of Yellow Leaf Disease (YLD) and its Associated Areca Palm Velarivirus 1 (APV1) in Betel Palm (Areca catechu) Plantations in Hainan, China.

Yellow leaf disease (YLD) is an economically important disease affecting betel palm in several countries, the cause of which remains unclear despite associations with putative agents, including phytoplasmas. In this study, we screened the potential casual agents associated with YLD in Hainan, China using next-generation sequencing and revealed the association of areca palm velarivirus 1 (APV1) with the YLD-affected palm. The complete genome of the APV1-WNY isolate was determined to be 17,546 nucleotides in length, approximately 1.5 kb longer than the previously reported APV1_HN genome. Transmission electron microscopy showed that APV1 particles are flexuous and filamentous, a typical morphology of species in the Closteroviridae family. Comparison of symptomatic and symptomless tree populations showed a strong association between APV1 and YLD. APV1 was detected in Pseudococcus sp. mealybugs sampled from YLD-affected trees in many locations, suggesting that mealybugs are a potential transmission vector for APV1. Although further studies are needed to confirm a causal relationship, these results provide timely information for the prevention and management of YLD associated with APV1.

Complete genome sequence of a novel capillovirus infecting Hevea brasiliensis in China.

Tapping panel dryness (TPD) is a complex disorder that causes partial or complete cessation of latex drainage upon tapping of rubber trees (Hevea brasiliensis). In this work, we determined the complete genome sequences of a novel virus identified in a rubber tree with TPD syndrome in China. The genome of the virus consists of 6811 nt and possesses two overlapping open reading frames (ORF1 and ORF2), encoding a polyprotein and a movement protein, respectively. The polyprotein shares 37% amino acid sequence identity with cherry virus A (CVA, ARQ83874.1) over 99% coverage. The genome architecture is similar to that of members of the genus Capillovirus (family Betaflexiviridae). Phylogenetic analysis of the replicase proteins showed that the virus clustered together with members of the genus Capillovirus. The new virus is tentatively called "rubber tree virus 1" (RTV1). RTV1 is the first virus reported to infect rubber trees. This work lays a foundation for research into finding the potential causal agent of TPD in Hevea brasiliensis.

Exponential subtraction in 3D ultrashort echo time imaging to visualize short T2 tissues in tibia.

BACKGROUND: Short T2 tissues can be directly visualized by dual-echo ultrashort echo time imaging with weighted subtraction. As a type of post-processing method, exponential subtraction of ultrashort echo time images with an optimal exponential factor is expected to provide improved positive short T2 contrast. PURPOSE: To test the feasibility and effectiveness of exponential subtraction in three-dimensional ultrashort echo time imaging and to determine the optimal exponential factor. MATERIAL AND METHODS: A dual-echo three-dimensional ultrashort echo time sequence was implemented on a 3-T MRI system. Exponential subtraction was performed on dual three-dimensional ultrashort echo time images of the tibia of seven healthy volunteers with exponential factors in the range of 1.00-3.00 in increments of 0.01. The regions of interest, including cortical bone, marrow, and muscle, were depicted on subtracted images of different exponential factors. Contrast-to-noise ratio values were calculated from these regions of interest and then used to assess the optimal exponential factor. To determine intra-observer agreement regarding region of interest selection, paired intra-observer measurements of regions of interest in all direct subtraction images were conducted with a one-week interval and the paired measurements were assessed using Bland-Altman analysis and paired-samples t-test. RESULTS: Cortical bone can be better visualized by using exponential subtraction in three-dimensional ultrashort echo time imaging; the suggested optimal exponential factor is 1.99-2.03 in the tibia. Paired measurements showed excellent intra-observer agreement. CONCLUSION: It is feasible to visualize cortical bone of the tibia using exponential subtraction in three-dimensional ultrashort echo time imaging. Compared with weighted subtraction images, exponential subtraction images with an optimal exponential factor provide enhanced visualization of short T2 tissues.

Enhancing adsorption capacity while maintaining specific recognition performance of mesoporous silica: a novel imprinting strategy with amphiphilic ionic liquid as surfactant.

In order to facilitate the broad applications of molecular recognition materials in biomedical areas, it is critical to enhance their adsorption capacity while maintaining their excellent recognition performance. In this work, we designed and synthesized well-defined peptide-imprinted mesoporous silica (PIMS) for specific recognition of an immunostimulating hexapeptide from human casein (IHHC) by using amphiphilic ionic liquid as the surfactant to anchor IHHC via a combination of one-step sol-gel method and docking oriented imprinting approach. Thereinto, theoretical calculation was employed to reveal the multiple binding interactions and dual-template configuration between amphiphilic ionic liquid and IHHC. The fabricated PIMS was characterized and an in-depth analysis of specific recognition mechanism was conducted. Results revealed that both adsorption and recognition capabilities of PIMS far exceeded that of the NIMS's. More significantly, the PIMS exhibited a superior binding capacity (60.5 mg g-1), which could increase 18.9% than the previous work. The corresponding imprinting factor and selectivity coefficient could reach up to 4.51 and 3.30, respectively. The PIMS also possessed lickety-split kinetic binding for IHHC, where the equilibrium time was only 10 min. All of these merits were due to the high surface area and the synergistic effect of multiple interactions (including hydrogen bonding, π-π stacking, ion-ion electrostatic interactions and van der Waals interactions, etc) between PIMS and IHHC in imprinted sites. The present work suggests the potential application of PIMS for large-scale and high-effective separation of IHHC, which may lead to their broad applications in drug/gene deliver, biosensors, catalyst and so on.