Impact of anti-VEGF therapy on distinctive retina layers in patients with macular edema secondary to branch retinal vein occlusion.

BACKGROUND: To explore the impact of anti-vascular epithelial growth factor (ant-VEGF) on the thickness of each retinal layer in patients with macular edema (ME) secondary to the branch retinal vein occlusion (BRVO). METHODS: This retrospective study included patients with ME secondary to monocular BRVO who received anti-VEGF therapy in Ningxia Eye Hospital between January-December 2020. RESULTS: Forty-three patients (25 males) were included, with 31 showed > 25% reduction in central retinal thickness (CRT) after anti-VEGF therapy (response group), and the others showed a ≤25% reduction in CRT (no-response group). The response group showed significantly smaller mean changes in the ganglion cell layer (GCL) (after 2 months) and inner plexiform layer (IPL) (after 1, 2, and 3 months) and significantly greater mean changes in the inner nuclear layer (INL) (after 2 and 3 months), outer plexiform layer (OPL) (after 3 months), outer nuclear layer (ONL) (after 2 and 3 months), and CRT (after 1 and 2 months) (all P < 0.05) as compared to the no-response group. The mean change in the thickness of each retinal layer IPL (P = 0.006) between the two groups was significantly different after controlling for a time and with a significant time trend (P < 0.001). Additionally, patients in the response group were more likely to have an improvement in IPL (43.68 ± 6.01 at 1 month and 41.52 ± 5.45 at 2 months vs. 39.9 ± 6.86 at baseline) after anti-VEGF therapy, while those in no response group might show improvement in GCL (45.75 ± 8.24 at 1 month, 40.00 ± 8.92 at 2 months, and 38.83 ± 9.93 at 3 months vs. 49.67 ± 6.83 at baseline). CONCLUSIONS: Anti-VEGF therapy might help restore the retinal structure and function in patients with ME secondary to BRVO, and those who have a response after anti-VEGF therapy are more likely to improve IPL, while those having no response might show improvement in GCL.

Genetic detection of two novel LRP5 pathogenic variants in patients with familial exudative vitreoretinopathy.

BACKGROUND: Familial exudative vitreoretinopathy (FEVR) is a genetic eye disorder that leads to abnormal development of retinal blood vessels, resulting in vision impairment. This study aims to identify pathogenic variants by targeted exome sequencing in 9 independent pedigrees with FEVR and characterize the novel pathogenic variants by molecular dynamics simulation. METHODS: Clinical data were collected from 9 families with FEVR. The causative genes were screened by targeted next-generation sequencing (TGS) and verified by Sanger sequencing. In silico analyses (SIFT, Polyphen2, Revel, MutationTaster, and GERP + +) were carried out to evaluate the pathogenicity of the variants. Molecular dynamics was simulated to predict protein conformation and flexibility transformation alterations on pathogenesis. Furthermore, molecular docking techniques were employed to explore the interactions and binding properties between LRP5 and DKK1 proteins relevant to the disease. RESULTS: A 44% overall detection rate was achieved with four variants including c.4289delC: p.Pro1431Argfs*8, c.2073G > T: p.Trp691Cys, c.1801G > A: p.Gly601Arg in LRP5 and c.633 T > A: p.Tyr211* in TSPAN12 in 4 unrelated probands. Based on in silico analysis and ACMG standard, two of them, c.4289delC: p.Pro1431Argfs*8 and c.2073G > T: p.Trp691Cys of LRP5 were identified as novel pathogenic variants. Based on computational predictions using molecular dynamics simulations and molecular docking, there are indications that these two variants might lead to alterations in the secondary structure and spatial conformation of the protein, potentially impacting its rigidity and flexibility. Furthermore, these pathogenic variants are speculated to potentially influence hydrogen bonding interactions and could result in an increased binding affinity with the DKK1 protein. CONCLUSIONS: Two novel genetic variants of the LRP5 gene were identified, expanding the range of mutations associated with FEVR. Through molecular dynamics simulations and molecular docking, the potential impact of these variants on protein structure and their interactions with the DKK1 protein has been explored. These findings provide further support for the involvement of these variants in the pathogenesis of the disease.

ESR1 mediated circ_0004018 suppresses angiogenesis in hepatocellular carcinoma via recruiting FUS and stabilizing TIMP2 expression.

Angiogenesis has been certified to account for tumor pathobiology. Circular RNAs (circRNAs) have been demonstrated to be involved in angiogenesis-related diseases, including hepatocellular carcinoma (HCC). Nevertheless, the regulatory roles of most circRNAs remain obscure. This study aims to uncover the function of hsa_circ_0004018 on angiogenesis in HCC. Firstly, quantitative real-time RT-PCR (RT-qPCR) analyzed that circ_0004018 was definitely down-regulated in HCC. Western blot analysis was conducted to detect the protein level of fused protein in sarcoma (FUS) and TIMP metallopeptidase inhibitor 2 (TIMP2). Functional assays were carried out to assess the impacts of circ_0004018 on HCC. From the experimental results, we found that overexpression of circ_0004018 significantly inhibited angiogenesis in HCC. The regulatory mechanism of circ_0004018 in HCC was determined by chromatin immunoprecipitation (ChIP), luciferase reporter assays and RNA immunoprecipitation (RIP) assay. Therefore, we proved that estrogen receptor 1 (ESR1) mediated circ_0004018 regulated TIMP2 by recruiting FUS. A series of rescue assays verified that circ_0004018 participated in angiogenesis in HCC via modulating TIMP2. In summary, this paper disclosed that ESR1 activated circ_0004018 inhibited angiogenesis in HCC via binding to FUS and stabilizing TIMP2 expression.

Multi-Task Transformer with Adaptive Cross-Entropy Loss for Multi-Dialect Speech Recognition.

At present, most multi-dialect speech recognition models are based on a hard-parameter-sharing multi-task structure, which makes it difficult to reveal how one task contributes to others. In addition, in order to balance multi-task learning, the weights of the multi-task objective function need to be manually adjusted. This makes multi-task learning very difficult and costly because it requires constantly trying various combinations of weights to determine the optimal task weights. In this paper, we propose a multi-dialect acoustic model that combines soft-parameter-sharing multi-task learning with Transformer, and introduce several auxiliary cross-attentions to enable the auxiliary task (dialect ID recognition) to provide dialect information for the multi-dialect speech recognition task. Furthermore, we use the adaptive cross-entropy loss function as the multi-task objective function, which automatically balances the learning of the multi-task model according to the loss proportion of each task during the training process. Therefore, the optimal weight combination can be found without any manual intervention. Finally, for the two tasks of multi-dialect (including low-resource dialect) speech recognition and dialect ID recognition, the experimental results show that, compared with single-dialect Transformer, single-task multi-dialect Transformer, and multi-task Transformer with hard parameter sharing, our method significantly reduces the average syllable error rate of Tibetan multi-dialect speech recognition and the character error rate of Chinese multi-dialect speech recognition.

Learning Hierarchical Representations with Spike-and-Slab Inception Network.

Recently, deep convolutional neural networks (CNN) with inception modules have attracted much attention due to their excellent performances on diverse domains. Nevertheless, the basic CNN can only capture a univariate feature, which is essentially linear. It leads to a weak ability in feature expression, further resulting in insufficient feature mining. In view of this issue, researchers incessantly deepened the network, bringing parameter redundancy and model over-fitting. Hence, whether we can employ this efficient deep neural network architecture to improve CNN and enhance the capacity of image recognition task still remains unknown. In this paper, we introduce spike-and-slab units to the modified inception module, enabling our model to capture dual latent variables and the average and covariance information. This operation further enhances the robustness of our model to variations of image intensity without increasing the model parameters. The results of several tasks demonstrated that dual variable operations can be well-integrated into inception modules, and excellent results have been achieved.

Dissecting myocardial mechanics in patients with severe aortic stenosis: 2-dimensional vs 3-dimensional-speckle tracking echocardiography.

BACKGROUND: Aortic stenosis (AS) causes left ventricular (LV) pressure overload, leading to adverse LV remodeling and dysfunction. Identifying early subclinical markers of LV dysfunction in patients with significant AS is critical as this could provide support for earlier intervention, which may result in improved long-term outcomes. We therefore examined the impact of severe AS and its consequent increase in LV afterload on myocardial deformation and rotational mechanics by 2-dimensional (2D) and 3-dimensional (3D) speckle-tracking echocardiography. METHODS: We prospectively measured various strain parameters in 168 patients (42% female, mean age 72 ± 12 years) with severe AS and LV ejection fraction (EF) ≥50%, and compared them to normal values found in literature. 2D and 3D images were analyzed for global longitudinal strain (GLS), global circumferential strain (GCS), global radial strain (GRS), basal rotation, apical rotation, and peak systolic twist. We further assessed the degree of concordance between 2D and 3D strain, and examined their association with measures of LV preload and afterload. RESULTS: Patients with severe AS exhibited significantly lower GLS and GRS but higher GCS, apical rotation, and twist by 2D and 3D echocardiography compared with published normal values (P = 0.003 for 3D twist, P < 0.001 for all others). Agreement between 2D- and 3D-GLS by concordance correlation coefficient was 0.49 (95% confidence interval: 0.39-0.57). GLS was correlated with valvulo-arterial impedance, a measure of LV afterload (r = 0.34, p < 0.001 and r = 0.23, p = 0.003, respectively). CONCLUSION: Patients with severe AS demonstrated lower-than-normal GLS and GRS but appear to compensate with higher-than-normal GCS, apical rotation, and twist in order to maintain a preserved LVEF. GLS showed a modest correlation with valvulo-arterial impedance.

Using arterial blood as a substitute for venous blood in routine biochemistry parameter examinations in rabbits.

BACKGROUND: It has been widely accepted that there is a significant difference in peripheral blood oxygen between arteries and veins. Therefore, arterial blood has been collected for blood gas analysis, and venous blood, because it is convenient to collect, has been used for most laboratory examinations. However, venous blood is always difficult to collect in rabbits; in contrast, arterial blood is easier to obtain, and research on whether arterial blood can be used instead of venous blood for routine biochemical parameter examination is rare. Therefore, the present study was designed to explore whether arterial blood can be used as a substitute for venous blood for routine biochemistry parameter examination in rabbits. RESULTS: Three venous blood samples with gross hemolysis were excluded. Venous and arterial blood samples were obtained from forty-two rabbits. Arterial blood samples correlate well with venous blood in alkaline phosphatase (ALP), aspartate aminotransferase (AST), alanine transaminase (ALT), gamma-glutamyl transpeptidase (GGT), total protein (TP), globulin (GLB), serum total cholesterol (TC), serum triglyceride (TG), high-density lipoprotein cholesterol (HDL), low-density lipoprotein cholesterol (LDL), urea (Ur) and creatinine (Cr) levels by Deming regression analysis with slopes ranging from 0.893 to 1.176 and intercepts ranging from - 4.886 to 5.835. Bland-Altman analysis showed that the two sample parameters had 93%-98% of the points within the 95% consistency limits. There were significant differences between venous blood and arterial blood in ALP, TP, TC, TG, HDL, LDL and Cr, while AST, ALT, GGT, GLB and Ur showed no significant differences. CONCLUSIONS: Arterial blood can be a substitute for venous blood in routine biochemistry parameter examinations in rabbits, especially in situations where venous blood is difficult to collect.

Multi-Level Cross Residual Network for Lung Nodule Classification.

Computer-aided algorithm plays an important role in disease diagnosis through medical images. As one of the major cancers, lung cancer is commonly detected by computer tomography. To increase the survival rate of lung cancer patients, an early-stage diagnosis is necessary. In this paper, we propose a new structure, multi-level cross residual convolutional neural network (ML-xResNet), to classify the different types of lung nodule malignancies. ML-xResNet is constructed by three-level parallel ResNets with different convolution kernel sizes to extract multi-scale features of the inputs. Moreover, the residuals are connected not only with the current level but also with other levels in a crossover manner. To illustrate the performance of ML-xResNet, we apply the model to process ternary classification (benign, indeterminate, and malignant lung nodules) and binary classification (benign and malignant lung nodules) of lung nodules, respectively. Based on the experiment results, the proposed ML-xResNet achieves the best results of 85.88% accuracy for ternary classification and 92.19% accuracy for binary classification, without any additional handcrafted preprocessing algorithm.

Influence of the Left-Behind Experience on Self-Efficacy of College Students.

The left-behind experience refers to the experience of children who were taken care of by caregivers in their hometowns for ≥6 months before reaching age 18, as one or both parents left the rural hometown for better paying jobs in bigger cities. The current study was conducted with college students to test the effect of the left-behind experience on self-efficacy. The General Self-Efficacy Scale and a questionnaire designed by the authors were adopted for a survey of 200 Chinese college students in their final year of university. Of 200 students, 72 (36%) students had ≥6 months of left-behind experience before reaching age 18. Results did not show statistically significant difference in self-efficacy between college students with or without left-behind experience (p > 0.05). In addition, researchers found no significant difference in self-efficacy between college students with left-behind experience from different backgrounds. Further research and focus on this area is necessary. [Journal of Psychosocial Nursing and Mental Health Services, xx(x), xx-xx.].

Quantitative Assessment of Left Ventricular Diastolic Stiffness Using Cardiac Shear Wave Elastography: A Pilot Study.

OBJECTIVES: The purpose of this study was to systematically investigate the feasible echocardiographic views for human transthoracic cardiac shear wave elastography (SWE) and the impact of myocardial anisotropy on myocardial stiffness measurements. METHODS: A novel cardiac SWE technique using pulse inversion harmonic imaging and time-aligned sequential tracking was developed for this study. The technique can measure the quantitative local myocardial stiffness noninvasively. Ten healthy volunteers were recruited and scanned by the proposed technique 3 times on 3 different days. RESULTS: Seven combinations of echocardiographic views and left ventricular (LV) segments were found to be feasible for LV diastolic stiffness measurements: basal interventricular septum under parasternal short- and long-axis views; mid interventricular septum under parasternal short- and long-axis views; anterior LV free wall under parasternal short- and long-axis views; and posterior LV free wall under a parasternal short-axis view. Statistical analyses showed good repeatability of LV diastolic stiffness measurements among 3 different days from 70% of the participants for the basal interventricular septum and posterior LV free wall short-axis views. On the same LV segment, the mean diastolic shear wave speed measurements from the short-axis view were statistically different from the long-axis measurements: 1.82 versus 1.29 m/s for the basal interventricular septum; 1.81 versus 1.45 m/s for mid interventricular septum; and 1.96 versus 1.77 m/s for the anterior LV free wall, indicating that myocardial anisotropy plays a substantial role in LV diastolic stiffness measurements. CONCLUSIONS: These results establish the preliminary normal range of LV diastolic stiffness under different scan views and provide important guidance for future clinical studies using cardiac SWE.

Validation of PDE9A Gene Identified in GWAS Showing Strong Association with Milk Production Traits in Chinese Holstein.

Phosphodiesterase9A (PDE9A) is a cyclic guanosine monophosphate (cGMP)-specific enzyme widely expressed among the tissues, which is important in activating cGMP-dependent signaling pathways. In our previous genome-wide association study, a single nucleotide polymorphism (SNP) (BTA-55340-no-rs(b)) located in the intron 14 of PDE9A, was found to be significantly associated with protein yield. In addition, we found that PDE9A was highly expressed in mammary gland by analyzing its mRNA expression in different tissues. The objectives of this study were to identify genetic polymorphisms of PDE9A and to determine the effects of these variants on milk production traits in dairy cattle. DNA sequencing identified 11 single nucleotide polymorphisms (SNPs) and six SNPs in 5' regulatory region were genotyped to test for the subsequent association analyses. After Bonferroni correction for multiple testing, all these identified SNPs were statistically significant for one or more milk production traits (p < 0.0001~0.0077). Interestingly, haplotype-based association analysis revealed similar effects on milk production traits (p < 0.01). In follow-up RNA expression analyses, two SNPs (c.-1376 G>A, c.-724 A>G) were involved in the regulation of gene expression. Consequently, our findings provide confirmatory evidences for associations of PDE9A variants with milk production traits and these identified SNPs may serve as genetic markers to accelerate Chinese Holstein breeding program.

Automatic classification of spinal osteosarcoma and giant cell tumor of bone using optimized DenseNet.

Objective: This study aims to explore an optimized deep-learning model for automatically classifying spinal osteosarcoma and giant cell tumors. In particular, it aims to provide a reliable method for distinguishing between these challenging diagnoses in medical imaging. Methods: This research employs an optimized DenseNet model with a self-attention mechanism to enhance feature extraction capabilities and reduce misclassification in differentiating spinal osteosarcoma and giant cell tumors. The model utilizes multi-scale feature map extraction for improved classification accuracy. The paper delves into the practical use of Gradient-weighted Class Activation Mapping (Grad-CAM) for enhancing medical image classification, specifically focusing on its application in diagnosing spinal osteosarcoma and giant cell tumors. The results demonstrate that the implementation of Grad-CAM visualization techniques has improved the performance of the deep learning model, resulting in an overall accuracy of 85.61%. Visualizations of images for these medical conditions using Grad-CAM, with corresponding class activation maps that indicate the tumor regions where the model focuses during predictions. Results: The model achieves an overall accuracy of 80% or higher, with sensitivity exceeding 80% and specificity surpassing 80%. The average area under the curve AUC for spinal osteosarcoma and giant cell tumors is 0.814 and 0.882, respectively. The model significantly supports orthopedics physicians in developing treatment and care plans. Conclusion: The DenseNet-based automatic classification model accurately distinguishes spinal osteosarcoma from giant cell tumors. This study contributes to medical image analysis, providing a valuable tool for clinicians in accurate diagnostic classification. Future efforts will focus on expanding the dataset and refining the algorithm to enhance the model's applicability in diverse clinical settings.

Improvements of myocardial strain and work in diabetes patients with normal ejection fraction after empagliflozin treatment.

AIMS/INTRODUCTION: To assess the effect of empagliflozin treatment on left ventricular (LV), right ventricular (RV) and left atrial (LA) functions in diabetes patients with normal ejection fraction. MATERIALS AND METHODS: The study included a total of 128 diabetes patients with multiple cardiovascular risk factors who were subjected to a 6-month follow up from the initiation of empagliflozin treatment. Before and after treatment with empagliflozin, LV, RV and LA strain, and noninvasive myocardial work parameters were evaluated by speckle tracking echocardiography. RESULTS: In 128 diabetes patients (mean age 56 ± 8 years, 85 men) with multiple cardiovascular risk factors, myocardial strain and work parameters were impaired, despite the absence of significant clinical symptoms of heart failure. After 6-month treatment with empagliflozin, the absolute value of LV strain in all directions increased, represented by LV global longitudinal strain (-18.0 ± 1.7% to -19.2 ± 1.7% [mean ± SD]). The same trend in LV global work efficiency (93 [91-94] % to 94 [93-95] % [median (IQR)]), RV free-wall longitudinal strain (-24.0 ± 2.7% to -25.0 ± 2.8%), LA reservoir (31 ± 5% to 34 ± 5%) and conduit strain (-14 ± 4% to -16 ± 4%) was also observed. LV mass index (106.9 ± 16.8-103.6 ± 16.4 g/m2) and LV global wasted work (143 [111-185] mmHg% to 108 [88-141] mmHg%) decreased after treatment (P < 0.05 for all). LV volume and LA volume index remained unchanged after treatment. In the multivariable analysis, the change in LA reservoir strain (ß = 0.050, P = 0.035) and baseline global longitudinal strain (ß = -0.488, P < 0.001) were independent predictors of improvement in LV global longitudinal strain. CONCLUSIONS: This study suggests that 6-month treatment with empagliflozin improved LV, RV and LA functions in diabetes patients with normal ejection fraction.

A Radiomics-Based Nomogram Using Ultrasound Carotid Plaque Evaluation For Predicting Cerebro-Cardiovascular Events In Asymptomatic Patients.

RATIONALE AND OBJECTIVES: This study aims to assess whether a radiomics-based nomogram correlates with a higher risk of future cerebro-cardiovascular events in patients with asymptomatic carotid plaques. Additionally, it investigates the nomogram's contribution to the revised Framingham Stroke Risk Profile (rFSRP) for predicting cerebro-cardiovascular risk. MATERIALS AND METHODS: Predictive models aimed at identifying an increased risk of future cerebro-cardiovascular events were developed and internally validated at one center, then externally validated at two other centers. Survival curves, constructed using the Kaplan-Meier method, were compared through the log-rank test. RESULTS: This study included a total of 2009 patients (3946 images). The final nomogram was generated using multivariate Cox regression variables, including dyslipidemia, lumen diameter, plaque echogenicity, and ultrasonography (US)-based radiomics risk. The Harrell's concordance index (C-index) for predicting events-free survival (EFS) was 0.708 in the training cohort, 0.574 in the external validation cohort 1, 0.632 in the internal validation cohort, and 0.639 in the external validation cohort 2. The final nomogram showed a significant increase in C-index compared to the clinical, conventional US, and US-based radiomics models (all P < 0.05). Furthermore, the final nomogram-assisted method significantly improved the sensitivity and accuracy of radiologists' visual qualitative score of plaque (both P < 0.001). Among 1058 patients with corresponding 1588 plaque US images classified as low-risk by the rFSRP, 75 (7.1%) patients with corresponding 93 (5.9%) carotid plaque images were appropriately reclassified to the high-risk category by the final nomogram. CONCLUSION: The radiomics-based nomogram demonstrated accurate prediction of cerebro-cardiovascular events in patients with asymptomatic carotid plaques. It also improved the sensitivity and accuracy of radiologists' visual qualitative score of carotid plaque and enhanced the risk stratification ability of rFSRP. SUMMARY: The radiomics-based nomogram allowed accurate prediction of cerebro-cardiovascular events, especially ipsilateral ischemic stroke in patients with asymptomatic carotid atherosclerotic plaques. KEY RESULTS: The radiomics-based nomogram allowed accurate prediction of cerebro-cardiovascular events, especially ipsilateral ischemic stroke in patients with asymptomatic carotid atherosclerotic plaques. The radiomics-based nomogram improved the sensitivity and accuracy of radiologists' visual qualitative score of carotid plaque. The radiomics-based nomogram improved the discrimination of high-risk populations from low-risk populations in asymptomatic patients with carotid atherosclerotic plaques and the risk stratification capability of the rFSRP.

Use of carotid plaque neovascularization at contrast-enhanced US to predict coronary events in patients with coronary artery disease.

PURPOSE: To determine whether carotid plaque neovascularization as assessed with contrast-enhanced ultrasonography (US) can help predict future coronary events in patients with stable coronary artery disease (CAD). MATERIALS AND METHODS: The study was approved by the hospital ethics committee, and informed consent was obtained from all patients. Three hundred twelve consecutive patients (228 men; mean age, 63 years ± 9; age range, 42-88 years) with both CAD and at least one carotid plaque thicker than 2.0 mm underwent both standard and contrast-enhanced carotid US. Patients with stable CAD were followed up for 8-47 months (mean, 33 months ± 9) or until a coronary event occurred. Statistical analysis was performed with the Student t test, χ(2) analysis, Kaplan-Meier method, and Cox proportional hazards regression models. RESULTS: Contrast material enhancement of plaque was seen in 42 of 51 patients (82%) with acute coronary syndrome (ACS) and 114 of 261 patients (43.7%) with stable CAD (P < .001). Coronary events occurred during the follow-up period in 24 of 111 patients (21.6%) with contrast material enhancement of plaque and only seven of 137 patients (5.1%) without enhancement (P< .001). In 248 patients with stable CAD and follow-up, Kaplan-Meier analysis demonstrated a significantly higher probability of developing coronary events in patients with contrast material enhancement of plaque than in those without contrast material enhancement (P < .001). The presence of contrast material enhancement of plaque was a significant and independent predictor of future coronary events in patients with stable CAD (odds ratio: 3.90; 95% confidence interval: 1.60, 9.46; P = .003). CONCLUSION: Contrast material enhancement of plaque is more common in patients with ACS than in those with stable CAD and is a significant and independent predictor of future coronary events in patients with stable CAD, suggesting that noninvasive contrast-enhanced carotid US may be used as a method for risk stratification of patients with stable CAD.

Word self-update contrastive adversarial networks for text-to-image synthesis.

Synthesizing realistic fine-grained images from text descriptions is a significant computer vision task. Although many GANs-based methods have been proposed to solve this task, generating high-quality images consistent with text information remains a difficult problem. These existing GANs-based methods ignore important words due to the use of fixed initial word features in generator, and neglect to learn semantic consistency between images and texts for discriminators. In this article, we propose a novel attentional generation and contrastive adversarial framework for fine-grained text-to-image synthesis, termed as Word Self-Update Contrastive Adversarial Networks (WSC-GAN). Specifically, we introduce a dual attention module for modeling color details and semantic information. With a new designed word self-update module, the generator can leverage visually important words to compute attention maps in the feature synthesis module. Furthermore, we contrive multi-branch contrastive discriminators to maintain better consistency between the generated image and text description. Two novel contrastive losses are proposed for our discriminators to impose image-sentence and image-word consistency constraints. Extensive experiments on CUB and MS-COCO datasets demonstrate that our method achieves better performance compared with state-of-the-art methods.

Lightweight image de-snowing: A better trade-off between network capacity and performance.

The single image de-snowing task is an essential topic in computer vision, as images captured on snowy days degrade the performance of current vision-based intelligent systems. Existing methods build complex network structures with numerous parameters to pursue continuous performance improvement. Nonetheless, they generally ignore the negative impact of large memory consumption in real applications. This paper aims to address the above problem by making a trade-off between network capacity and performance. We propose two novel networks suitable for different application scenarios. For devices with small memory and requiring fast inference speed, we propose an extremely lightweight recursive network (XLRNet). XLRNet is constructed by a single recursive strategy and two novel lightweight modules. For devices with large memory and pursuing better de-snowing performance, we propose a coupled lightweight dual recursive network (CLDRNet). CLDRNet cascades two XLRNets by a novel dual recursive strategy and a novel dual coupled LSTM module (DC-LSTM). Extensive experiments demonstrate the effectiveness and superiority of our two models on three synthetic datasets and real-world datasets.

Exploring the prospect of intrinsic wave propagation in evaluating myocardial stiffness among patients with type 2 diabetes.

Background: Diabetes predisposes affected individuals to impaired myocardial perfusion and ischemia, leading to cardiac dysfunction. Increased myocardial stiffness is an independent and significant risk factor in diastolic dysfunction. This study sought to estimate myocardial stiffness in Type 2 diabetes (T2DM) patients using the intrinsic wave velocity propagation (IVP) along the longitudinal wall motion during late diastole and evaluate the value of IVP in assessing cardiac function and structure. Methods: 87 and 53 participants with and without T2DM (control group) were enrolled. Of the 87 T2DM patients (DM group), 43 were complicated with hypertension (DM + H group), and 44 were not (DM-H group). Ultrasound parameters were measured and analyzed, including color M-mode flow propagation velocity, global longitudinal systolic strain (GLS), and IVP. Results: IVP was higher in the DM group than in the control group (1.62 ± 0.25 m/s and 1.40 ± 0.19 m/s, P < 0.001). After stratification for hypertension, IVP in both DM + H (1.71 ± 0.25 m/s) and DM-H (1.53 ± 0.20 m/s) groups were found to be significantly higher than that in the control group (1.40 ± 0.19 m/s); also, the difference of IVP between DM + H and DM-H group reached statistical significance. Moreover, IVP was significantly correlated with flow propagation velocity during early diastole (Pve) (r = -0.580, P < 0.001), flow propagation velocity during late diastole (Pva) (r = 0.271, P < 0.001), GLS (r = 0.330, P < 0.001), interventricular septal thickness at end-diastole (IVSd) (r = 0.321, P < 0.001), blood glucose (r = 0.246, P < 0.003), systolic blood pressure (r = 0.370, P < 0.001) and diastolic blood pressure (r = 0.389, P < 0.001). Conclusions: The results indicated the application potential of IVP in assessing the early detection of cardiac function changes noninvasively and sensitively. The correlation with myocardial stiffness warrants further studies to substantiate its potential clinical utility.

Value of Intraplaque Neovascularization on Contrast-Enhanced Ultrasonography in Predicting Ischemic Stroke Recurrence in Patients With Carotid Atherosclerotic Plaque.

OBJECTIVE: Patients with a history of ischemic stroke are at risk for a second ischemic stroke. This study aimed to investigate the relationship between carotid plaque enhancement on perfluorobutane microbubble contrast-enhanced ultrasonography (CEUS) and future recurrent stroke, and to determine whether plaque enhancement can contribute to risk assessment for recurrent stroke compared with the Essen Stroke Risk Score (ESRS). MATERIALS AND METHODS: This prospective study screened 151 patients with recent ischemic stroke and carotid atherosclerotic plaques at our hospital between August 2020 and December 2020. A total of 149 eligible patients underwent carotid CEUS, and 130 patients who were followed up for 15-27 months or until stroke recurrence were analyzed. Plaque enhancement on CEUS was investigated as a possible risk factor for stroke recurrence and as a possible adjunct to ESRS. RESULTS: During follow-up, 25 patients (19.2%) experienced recurrent stroke. Patients with plaque enhancement on CEUS had an increased risk of stroke recurrence events (22/73, 30.1%) compared to those without plaque enhancement (3/57, 5.3%), with an adjusted hazard ratio (HR) of 38.264 (95% confidence interval [CI]:14.975-97.767; P < 0.001) according to a multivariable Cox proportional hazards model analysis, indicating that the presence of carotid plaque enhancement was a significant independent predictor of recurrent stroke. When plaque enhancement was added to the ESRS, the HR for stroke recurrence in the high-risk group compared to that in the low-risk group (2.188; 95% CI, 0.025-3.388) was greater than that of the ESRS alone (1.706; 95% CI, 0.810-9.014). A net of 32.0% of the recurrence group was reclassified upward appropriately by the addition of plaque enhancement to the ESRS. CONCLUSION: Carotid plaque enhancement was a significant and independent predictor of stroke recurrence in patients with ischemic stroke. Furthermore, the addition of plaque enhancement improved the risk stratification capability of the ESRS.

Quantitative Assessment of Right Ventricular Function in Patients With Systemic Lupus Erythematosus Using the Novel Non-invasive Pressure-Strain Loop.

OBJECTIVE: Current echocardiography evaluation of right ventricular (RV) function, which heralds the prognosis in patients with systemic lupus erythematosus (SLE), is of limited utility. The non-invasive pressure-strain loop (PSL), an emerging technique, has been found to feasible, sensitive and accurate in the diagnosis of cardiovascular diseases. The aim of this study was to quantitatively evaluate, using the non-invasive PSL, the right ventricular myocardial work (RVMW) in SLE patients. METHODS: Seventy-five SLE patients were recruited and grouped by pulmonary artery systolic pressure (PASP) into normal (group A, N = 26), mild (group B, N = 22) and moderate to severe (group C, N = 27) groups. Twenty-five healthy volunteers undergoing physical examination were recruited as the control group. Right ventricular global myocardial work index (RVGWI), global constructive work (RVGCW), global wasted work (RVGWW), global work efficiency (RVGWE), global longitudinal strain (RVGLS) and other conventional parameters were measured. DISCUSSION: There were no differences between group A and the control group with respect to RVLS, RVGLS and all RVMW parameters (all p values > 0.05). RVGWI and RVGCW significantly differed among the other groups (all p values < 0.05). RVGWE was significantly lower and RVGWW was significantly higher in group C than in the control group and groups A and B (all p values < 0.05). Compared with the control group, RVGWW was significantly increased and RVGLS was significantly decreased in group B (all p values < 0.05). All but one RVMW parameter moderately to strongly correlated with SLE disease activity index (SLEDAI) and World Health Organization Functional Class (WHO-FC). RVGWW (area under the receiver operating characteristic curve [AUC] = 0.893) and RVGWE (AUC = 0.877) were sensitive parameters in detecting earlier cardiac dysfunction in SLE patients. CONCLUSION: RVGWW and RVGWE serve as sensitive and promising parameters in the integrative analysis of early right ventricular dysfunction in SLE patients. To conclude, non-invasive PSL, the novel method, facilitates the quantitative assessment of RVMW in SLE patients.