[Analysis of the clinical effect of percutaneous pedicle screw fixation combined with transpedicular bone grafting in the treatment of thoracolumbar fracture].

OBJECTIVE: To investigate the clinical efficacy of percutaneous screw fixation combined with minimally invasive transpedicular bone grafting and non-bone grafting in the treatment of thoracolumbar fractures. METHODS: From Janury 2021 to June 2022, 40 patients with thoracolumbar fracture were divided into the experimental group and the control group. There were 26 patients in the experimental group, including 21 males and 5 females with an aberage age of (47.3±12.3) years old, who underwent percutaneous pedicle screw fixation combined with transpedicular autogenous bone grafting. In the control group, 14 patients received percutaneous pedicle screw fixation only. including 7 makes and 7 females with an average age of (50.2±11.2) years old. The operative time, intraoperative blood loss, anterior height ratio of injured vertebrae, Cobb angle, visual analogue score (VAS), MacNab scores, loosening or broken of the implants. were compared and analyzed. RESULTS: There was no significant difference in operation time, intraoperative blood loss, VAS and anterior height ratio of injured vertebrae between the two groups. Compared with the preoperative results, VAS and anterior height ratio of injured vertebrae were improved statistically(P<0.05). For Cobb angle of injured vertebra, there was no significant difference between the two groups before surgery (P=0.766). While at 1 week, 3 months and 12 months after surgery, there were statistically differences between the two groups (P values were 0.042, 0.007 and 0.039, respectively). The Cobb angle of injured vertebrae one year after operation was statistically decreased in both groups compared with that before surgery (P<0.001). One year after surgery, the excellent and good rate of Macnab scores was 96.15% in the experimental group and 92.86% in the control group, and there was no statistical differences between the two groups (P=0.648). There was one patient in the control group suffering superficial wound infection on the third day, which was cured by dressing change and anti-infection treatment. There were no postoperative screw loosening and broken in both groups. CONCLUSION: The two surgical methods have the advantages of less trauma, less pain and quicker recovery, which can restore the height of the injured vertebra, reconstruct the spinal sequence and reduce the fracture of the vertebral body. Transpedicular autogenous bone grafting can increase the stability of the fractured vertebra and maintain the height of the vertebra better after surgery, thus reducing the possibility of complications such as kyphosis, screw loosening and broken.

Improved genomic prediction using machine learning with Variational Bayesian sparsity.

BACKGROUND: Genomic prediction has become a powerful modelling tool for assessing line performance in plant and livestock breeding programmes. Among the genomic prediction modelling approaches, linear based models have proven to provide accurate predictions even when the number of genetic markers exceeds the number of data samples. However, breeding programmes are now compiling data from large numbers of lines and test environments for analyses, rendering these approaches computationally prohibitive. Machine learning (ML) now offers a solution to this problem through the construction of fully connected deep learning architectures and high parallelisation of the predictive task. However, the fully connected nature of these architectures immediately generates an over-parameterisation of the network that needs addressing for efficient and accurate predictions. RESULTS: In this research we explore the use of an ML architecture governed by variational Bayesian sparsity in its initial layers that we have called VBS-ML. The use of VBS-ML provides a mechanism for feature selection of important markers linked to the trait, immediately reducing the network over-parameterisation. Selected markers then propagate to the remaining fully connected feed-forward components of the ML network to form the final genomic prediction. We illustrated the approach with four large Australian wheat breeding data sets that range from 2665 lines to 10375 lines genotyped across a large set of markers. For all data sets, the use of the VBS-ML architecture improved genomic prediction accuracy over legacy linear based modelling approaches. CONCLUSIONS: An ML architecture governed under a variational Bayesian paradigm was shown to improve genomic prediction accuracy over legacy modelling approaches. This VBS-ML approach can be used to dramatically decrease the parameter burden on the network and provide a computationally feasible approach for improving genomic prediction conducted with large breeding population numbers and genetic markers.

SharpFormer: Learning Local Feature Preserving Global Representations for Image Deblurring.

The goal of dynamic scene deblurring is to remove the motion blur presented in a given image. To recover the details from the severe blurs, conventional convolutional neural networks (CNNs) based methods typically increase the number of convolution layers, kernel-size, or different scale images to enlarge the receptive field. However, these methods neglect the non-uniform nature of blurs, and cannot extract varied local and global information. Unlike the CNNs-based methods, we propose a Transformer-based model for image deblurring, named SharpFormer, that directly learns long-range dependencies via a novel Transformer module to overcome large blur variations. Transformer is good at learning global information but is poor at capturing local information. To overcome this issue, we design a novel Locality preserving Transformer (LTransformer) block to integrate sufficient local information into global features. In addition, to effectively apply LTransformer to the medium-resolution features, a hybrid block is introduced to capture intermediate mixed features. Furthermore, we use a dynamic convolution (DyConv) block, which aggregates multiple parallel convolution kernels to handle the non-uniform blur of inputs. We leverage a powerful two-stage attentive framework composed of the above blocks to learn the global, hybrid, and local features effectively. Extensive experiments on the GoPro and REDS datasets show that the proposed SharpFormer performs favourably against the state-of-the-art methods in blurred image restoration.

M1 Bone Marrow-Derived Macrophage-Derived Extracellular Vesicles Inhibit Angiogenesis and Myocardial Regeneration Following Myocardial Infarction via the MALAT1/MicroRNA-25-3p/CDC42 Axis.

Myocardial infarction (MI) is a severe cardiovascular disease. Some M1 macrophage-derived extracellular vesicles (EVs) are involved in the inhibition of angiogenesis and acceleration dysfunction during MI. However, the potential mechanism of M1 phenotype bone marrow-derived macrophages- (BMMs-) EVs (M1-BMMs-EVs) in MI is largely unknown. This study sought to investigate whether M1-BMMs-EVs increased CDC42 expression and activated the MEK/ERK pathway by carrying lncRNA MALAT1 and competitively binding to miR-25-3p, thus inhibiting angiogenesis and myocardial regeneration after MI. After EV treatment, the cardiac function, infarct size, fibrosis, angiogenesis, and myocardial regeneration of MI mice and the viability, proliferation and angiogenesis of oxygen-glucose deprivation- (OGD-) treated myocardial microvascular endothelial cells (MMECs) were assessed. MALAT1 expression in MI mice, cells, and EVs was detected. MALAT1 downstream microRNAs (miRs), genes, and pathways were predicted and verified. MALAT1 and miR-25-3p were intervened to evaluate EV effects on OGD-treated cells. In MI mice, EV treatment aggravated MI and inhibited angiogenesis and myocardial regeneration. In OGD-treated cells, EV treatment suppressed cell viability, proliferation, and angiogenesis. MALAT1 was highly expressed in MI mice, OGD-treated MMECs, M1-BMMs, and EVs. Silencing MALAT1 weakened the inhibition of EV treatment on OGD-treated cells. MALAT1 sponged miR-25-3p to upregulate CDC42. miR-25-3p overexpression promoted OGD-treated cell viability, proliferation, and angiogenesis. The MEK/ERK pathway was activated after EV treatment. Collectively, M1-BMMs-EVs inhibited angiogenesis and myocardial regeneration following MI via the MALAT1/miR-25-3p/CDC42 axis and the MEK/ERK pathway activation.

Endothelial peroxynitrite causes disturbance of neuronal oscillations by targeting caspase-1 in the arcuate nucleus.

Severe anorexia limits the clinical application of cisplatin, and even leads to the discontinuation of treatment. However, the mechanisms underlying cisplatin-induced anorexia are unknown. Herein, we demonstrated that cisplatin could affect neuronal gamma oscillations and induce abnormal neuronal theta-gamma phase-amplitude coupling in the arcuate nucleus (Arc) of the hypothalamus, and these findings were associated with significantly decreased food intake and weight loss in mice. Chemogenetic activation of AgRP neurons in the Arc reversed the cisplatin-induced food intake reduction in mice. We further demonstrated that endothelial peroxynitrite (ONOO-) formation in the Arc induced nitrosative stress following cisplatin treatment via a previously uncharacterized pathway involving neuronal caspase-1 activation. Strikingly, treatment with the ONOO- scavenger uric acid (UA) reversed the reduced action potential (AP) frequency of AgRP neurons and increased the AP frequency of POMC neurons induced by SIN1, a donor of ONOO-, in the Arc, as determined by whole-cell patch-clamp electrophysiological recording. Consistent with these findings, UA treatment effectively alleviated cisplatin-induced dysfunction of neuronal oscillations and neuronal theta-gamma phase-amplitude coupling in the Arc of mice. Taken together, these results suggest, for the first time, that targeting the overproduction of endothelial ONOO- can regulate cisplatin-induced neurotoxicity through neuronal caspase-1, and thereby serve as a potential therapeutic approach to alleviate chemotherapy-induced anorexia and weight loss.

Learning to Zoom-In via Learning to Zoom-Out: Real-World Super-Resolution by Generating and Adapting Degradation.

Most learning-based super-resolution (SR) methods aim to recover high-resolution (HR) image from a given low-resolution (LR) image via learning on LR-HR image pairs. The SR methods learned on synthetic data do not perform well in real-world, due to the domain gap between the artificially synthesized and real LR images. Some efforts are thus taken to capture real-world image pairs. However, the captured LR-HR image pairs usually suffer from unavoidable misalignment, which hampers the performance of end- to-end learning. Here, focusing on the real-world SR, we ask a different question: since misalignment is unavoidable, can we propose a method that does not need LR-HR image pairing and alignment at all and utilizes real images as they are? Hence we propose a framework to learn SR from an arbitrary set of unpaired LR and HR images and see how far a step can go in such a realistic and "unsupervised" setting. To do so, we firstly train a degradation generation network to generate realistic LR images and, more importantly, to capture their distribution (i.e., learning to zoom out). Instead of assuming the domain gap has been eliminated, we minimize the discrepancy between the generated data and real data while learning a degradation adaptive SR network (i.e., learning to zoom in). The proposed unpaired method achieves state-of- the-art SR results on real-world images, even in the datasets that favour the paired-learning methods more.

COVID-19 Chest CT Image Segmentation Network by Multi-Scale Fusion and Enhancement Operations.

A novel coronavirus disease 2019 (COVID-19) was detected and has spread rapidly across various countries around the world since the end of the year 2019. Computed Tomography (CT) images have been used as a crucial alternative to the time-consuming RT-PCR test. However, pure manual segmentation of CT images faces a serious challenge with the increase of suspected cases, resulting in urgent requirements for accurate and automatic segmentation of COVID-19 infections. Unfortunately, since the imaging characteristics of the COVID-19 infection are diverse and similar to the backgrounds, existing medical image segmentation methods cannot achieve satisfactory performance. In this article, we try to establish a new deep convolutional neural network tailored for segmenting the chest CT images with COVID-19 infections. We first maintain a large and new chest CT image dataset consisting of 165,667 annotated chest CT images from 861 patients with confirmed COVID-19. Inspired by the observation that the boundary of the infected lung can be enhanced by adjusting the global intensity, in the proposed deep CNN, we introduce a feature variation block which adaptively adjusts the global properties of the features for segmenting COVID-19 infection. The proposed FV block can enhance the capability of feature representation effectively and adaptively for diverse cases. We fuse features at different scales by proposing Progressive Atrous Spatial Pyramid Pooling to handle the sophisticated infection areas with diverse appearance and shapes. The proposed method achieves state-of-the-art performance. Dice similarity coefficients are 0.987 and 0.726 for lung and COVID-19 segmentation, respectively. We conducted experiments on the data collected in China and Germany and show that the proposed deep CNN can produce impressive performance effectively. The proposed network enhances the segmentation ability of the COVID-19 infection, makes the connection with other techniques and contributes to the development of remedying COVID-19 infection.

Learning Deep Gradient Descent Optimization for Image Deconvolution.

As an integral component of blind image deblurring, non-blind deconvolution removes image blur with a given blur kernel, which is essential but difficult due to the ill-posed nature of the inverse problem. The predominant approach is based on optimization subject to regularization functions that are either manually designed or learned from examples. Existing learning-based methods have shown superior restoration quality but are not practical enough due to their restricted and static model design. They solely focus on learning a prior and require to know the noise level for deconvolution. We address the gap between the optimization- and learning-based approaches by learning a universal gradient descent optimizer. We propose a recurrent gradient descent network (RGDN) by systematically incorporating deep neural networks into a fully parameterized gradient descent scheme. A hyperparameter-free update unit shared across steps is used to generate the updates from the current estimates based on a convolutional neural network. By training on diverse examples, the RGDN learns an implicit image prior and a universal update rule through recursive supervision. The learned optimizer can be repeatedly used to improve the quality of diverse degenerated observations. The proposed method possesses strong interpretability and high generalization. Extensive experiments on synthetic benchmarks and challenging real-world images demonstrate that the proposed deep optimization method is effective and robust to produce favorable results as well as practical for real-world image deblurring applications.

[Partitioning ecosystem respiration of a Platycladus orientalis forest in the west mountainous area of Beijing, China using stable carbon isotope].

Based on stable carbon isotope, we quantitatively partitioned ecosystem respiration in a Platycladus orientalis forest in the west mountainous area of Beijing. Results from this study could lay the foundation for carbon exchange research in forest ecosystems of this region. The spectroscopy technique was used to continuously measure CO2 concentrations and δ13C values at different height of the forest. Soil and branch chambers were used for measuring nighttime δ13C values in underground and aboveground respiration, and then the proportions of respiration components were calculated. Combined with soil respiration efflux measurement, ecosystem respiration was then quantitatively partitioned. The results showed that δ13C values of respiratory components fluctuated, which ranged from -31.74‰ to -23.33‰ in aboveground respiration of plants and from -32.11‰ to -27.74‰ in soil respiration. The δ13C values of ecosystem respiration was at the middle of those ranges. Soil respiration averaged 1.70 µmol·m-2·s-1 at night, accounting for 47%-91% of ecosystem respiration. Aboveground respiration averaged 0.72 µmol·m-2·s-1, contributing less to ecosystem respiration. Daytime respiration based on isotope mixing model calculation had greater variability than that based on temperature response model, with a mean value of 2.31 µmol·m-2·s-1 and 2.28 µmol·m-2·s-1, respectively.

Endothelial Cdk5 deficit leads to the development of spontaneous epilepsy through CXCL1/CXCR2-mediated reactive astrogliosis.

Blood-brain barrier (BBB) dysfunction has been suggested to play an important role in epilepsy. However, the mechanism mediating the transition from cerebrovascular damage to epilepsy remains unknown. Here, we report that endothelial cyclin-dependent kinase 5 (CDK5) is a central regulator of neuronal excitability. Endothelial-specific Cdk5 knockout led to spontaneous seizures in mice. Knockout mice showed increased endothelial chemokine (C-X-C motif) ligand 1 (Cxcl1) expression, decreased astrocytic glutamate reuptake through the glutamate transporter 1 (GLT1), and increased glutamate synaptic function. Ceftriaxone restored astrocytic GLT1 function and inhibited seizures in endothelial Cdk5-deficient mice, and these effects were also reversed after silencing Cxcl1 in endothelial cells and its receptor chemokine (C-X-C motif) receptor 2 (Cxcr2) in astrocytes, respectively, in the CA1 by AAV transfection. These results reveal a previously unknown link between cerebrovascular factors and epileptogenesis and provide a rationale for targeting endothelial signaling as a potential treatment for epilepsy.

[Experimental study of platelet-rich plasma in treatment of Achilles tendinopathy in rabbits].

OBJECTIVE: To explore the effect of platelet-rich plasma (PRP) in treatment of Achilles tendinopathy in rabbits, and provide experimental evidence for the clinical application of PRP in treatment of Achilles tendinopathy. METHODS: Forty-eight adult New Zealand white rabbits, weighing 2.5-3.0 kg, male or female, were randomly divided into model group (group A), model control group (group B), model+treatment control group (group C), model+treatment group (group D), with 12 in each group. The rabbits were injected with type Ⅰ collagenase to prepare Achilles tendinopathy models in groups A, C, and D, and with an equal dose of normal saline in group B. The blood from the central artery of rabbit ear was taken to preprare PRP by secondary centrifugation in group D. The results of platelet counts showed that PRP platelets reached 3 to 5 times the whole blood. After the model was prepared, the rabbits in groups C and D were injected with physiological saline and autologous PRP at the molding site respectively, once a week, 0.8 mL each time for 4 weeks. At 1 week after PRP injection, the relative hardness (expressed as HRD%) of Achilles tendon was evaluated by ultrasound elastic quantitative imaging detection technique; the maximum breaking load of Achilles tendon was measured by universal electronic tensile testing machine; the contents of collagen type Ⅰ and Ⅲ were determined by ELISA; and the morphology of Achilles tendon collagen fibers was observed by HE and Masson stainings. RESULTS: All animals survived during the experiment. The results of ultrasound elastic quantitative imaging and mechanical tests showed that the HRD% and the maximum breaking load were significantly lower in group A than in group B ( P<0.05) and in group C than in group D ( P<0.05). The results of ELISA showed that the content of collagen type Ⅰ was significantly lower in group A than in group B ( P<0.05) and in group C than in group D ( P<0.05); the content of collagen type Ⅲ was significantly higher in group A than in group B ( P<0.05) and in group D than in group C ( P<0.05). HE and Masson stainings showed that the Achilles tendon collagen fibers were irregularly curled and the structure was severely damaged in group A; the fibers were parallel and ordered, and the structure was complete in group B; the fibers were irregularly curled and structurally disordered in group C; the fibers were slightly curled and the structure was relatively complete in group D. CONCLUSION: A rabbit model of Achilles tendinopathy can be reconstructed by type Ⅰ collagenase injection. PRP treatment can increase the Achilles tendon hardness and maximum breaking load, up-regulate the expression level of collagen type Ⅰ and Ⅲ, improve the structure of Achilles tendon collagen fiber, and promote the repair in rabbit Achilles tendinopathy model.

Functional coupling of Tmem74 and HCN1 channels regulates anxiety-like behavior in BLA neurons.

Anxiety disorders are the most prevalent psychiatric disorders, but their pathogenic mechanism remains poorly understood. Here, we report that transmembrane protein 74 (TMEM74), which contains two putative transmembrane domains and exhibits high levels of mRNA in the brain, is closely associated with the pathogenesis of anxiety disorders. TMEM74 was decreased in the serum of patients with anxiety and the basolateral amygdaloid nucleus (BLA) in chronic stress mice. Furthermore, genetic deletion of Tmem74 or selective knockdown of Tmem74 in BLA pyramidal neurons resulted in anxiety-like behaviors in mice. Whole-cell recordings in BLA pyramidal neurons revealed lower hyperpolarization-activated cation current (Ih) and greater input resistance and excitability in Tmem74-/- neurons than in wild-type neurons. Accordingly, surface expression of hyperpolarization-activated cyclic nucleotide-gated 1 (HCN1) channels was also lower in the BLA of Tmem74-/- mice. The Ih current blocker ZD7288 mimicked these effects in BLA pyramidal neurons in wild-type mice but not in Tmem74-/- mice. Consistent with the improvement in anxiety-like behaviors, Tmem74 overexpression restored HCN1 channel trafficking and pyramidal neuron excitability in the BLA of Tmem74-/- and chronic stress mice. Mechanistically, we demonstrate that interactions between Tmem74 and HCN1 are physiologically relevant and that transmembrane domain 1 (TM1) is essential for the cellular membrane localization of Tmem74 to enhance Ih. Together, our findings suggest that Tmem74 coupling with HCN1 acts as a critical component in the pathophysiology of anxiety and is a potential target for new treatments of anxiety disorders.

Secretory vimentin is associated with coronary artery disease in patients and induces atherogenesis in ApoE-/- mice.

PURPOSE: The present study aimed to investigate the relationship between serum levels of secretory vimentin and coronary artery disease (CAD). The biological effect of secretory vimentin was ascertained by experiments. METHODS: We analysed serum levels of secretory vimentin in CAD patients (n = 288) and non-CAD controls (n = 195) by ELISA. To evaluate the pro-inflammatory effects of secreted vimentin, the human aortic endothelial cells (HAECs) and human peripheral blood mononuclear cells (PBMCs) were treated with recombinant vimentin or saline. Intraperitoneal injection of vimentin (1 µg/each) or saline was performed every other day for 12 weeks in ApoE-/- mice for assessment of atherogenic effect. RESULTS: Serum levels of secretory vimentin were significantly increased in CAD patients than in health controls (p < 0.05), and correlated with the number of diseased coronary arteries, Syntax and Gensini score (for all comparison, p < 0.01). Logistic regression analysis showed that vimentin level is an independent determinant of CAD. In experiments, recombinant vimentin protein enhanced the expression of adhesion molecules and inflammatory cytokines in both endothelial cells and macrophages. This protein also promoted macrophage-endothelial cells adhesion in vitro and the recruitment of leukocytes to mesenteric venules in C57BL/6 mice. Compared with saline, intraperitoneal injection of recombinant vimentin (1 µg/each) every other day induced atherogenesis in ApoE-/- mice at 12-weeks, with significant increase of inflammatory cytokine and adhesion molecules expression in aortic tissue (p < 0.05). CONCLUSION: Serum vimentin levels are associated with the presence and the severity of CAD. Vimentin protein promotes atherogenesis in ApoE-/- mice.

MPTV: Matching Pursuit-Based Total Variation Minimization for Image Deconvolution.

Total variation (TV) regularization has proven effective for a range of computer vision tasks through its preferential weighting of sharp image edges. Existing TV-based methods, however, often suffer from the over-smoothing issue and solution bias caused by the homogeneous penalization. In this paper, we consider addressing these issues by applying inhomogeneous regularization on different image components. We formulate the inhomogeneous TV minimization problem as a convex quadratic constrained linear programming problem. Relying on this new model, we propose a matching pursuit-based total variation minimization method (MPTV), specifically for image deconvolution. The proposed MPTV method is essentially a cutting-plane method that iteratively activates a subset of nonzero image gradients and then solves a subproblem focusing on those activated gradients only. Compared with existing methods, the MPTV is less sensitive to the choice of the trade-off parameter between data fitting and regularization. Moreover, the inhomogeneity of MPTV alleviates the over-smoothing and ringing artifacts and improves the robustness to errors in blur kernel. Extensive experiments on different tasks demonstrate the superiority of the proposed method over the current state of the art.

Cathepsin B inhibition ameliorates leukocyte-endothelial adhesion in the BTBR mouse model of autism.

AIMS: Autism spectrum disorder (ASD) is a wide range of neurodevelopmental disorders involving deficits in social interaction and communication. Unfortunately, autism remains a scientific and clinical challenge owing to the lack of understanding the cellular and molecular mechanisms underlying it. This study aimed to investigate the pathophysiological mechanism underlying leukocyte-endothelial adhesion in autism-related neurovascular inflammation. METHODS: Male BTBR T+tf/J mice were used as an autism model. The dynamic pattern of leukocyte-endothelial adhesion in mouse cerebral vessels was detected by two-photon laser scanning microscopy (TPLSM). Using FACS, RT-PCR, and Western blotting, we explored the expression of cell adhesion molecules, the mRNA expression of endothelial chemokine, the protein levels of cathepsin B, and inflammatory mediators. RESULTS: We found a significant increase in leukocyte-endothelial adhesion in BTBR mice, accompanied by elevated expression of the adhesion molecule neutrophils CD11b and endothelial ICAM-1. Our data further indicate that elevated neutrophil cathepsin B levels contribute to elevated endothelial chemokine CXCL7 levels in BTBR mice. The pharmacological inhibition of cathepsin B reverses the enhanced leukocyte-endothelial adhesion in the cerebral vessels of autistic mice. CONCLUSION: Our results revealed the prominent role of cathepsin B in modulating leukocyte-endothelial adhesion during autism-related neurovascular inflammation and identified a promising novel approach for autism treatment.

Proteomic analysis of the response of porcine adrenal gland to heat stress.

Heat stress (HS) and its associated pathologies are major challenges facing the pig industry in southern China, and are responsible for large economic losses. However, the molecular mechanisms governing the abnormal secretion of HS-responsive hormones, such as glucocorticoids, are not fully understood. The goal of this study was to investigate differentially expressed proteins (DEPs) in the adrenal glands of pigs, and to elucidate changes in the immune neuroendocrine system in pigs following HS. Through a functional proteomics approach, we identified 1202 peptides, corresponding to 415 proteins. Of these, we found 226 DEPs between heat-stressed and control porcine adrenal gland tissue; 99 of these were up-regulated and 127 were down-regulated in response to HS. These DEPs included proteins involved in substrate transport, cytoskeletal changes, and stress responses. Ingenuity Pathway Analysis was used to identify the subcellular characterization, functional pathway involvement, regulatory networks, and upstream regulators of the identified proteins. Functional network and pathway analyses may provide insights into the complexity and dynamics of HS-host interactions, and may accelerate our understanding of the mechanisms of HS.

Two-Stream Convolutional Networks for Blind Image Quality Assessment.

Traditional image quality assessment (IQA) methods do not perform robustly due to the shallow hand-designed features. It has been demonstrated that deep neural network can learn more effective features than ever. In this paper, we describe a new deep neural network to predict the image quality accurately without relying on the reference image. To learn more effective feature representations for non-reference IQA, we propose a two-stream convolution network that includes two subcomponents for image and gradient image. The motivation for this design is using a two-stream scheme to capture different-level information of inputs and easing the difficulty of extracting features from one steam. The gradient stream focuses on extracting structure features in details, and the image stream pays more attention to the information in intensity. In addition, to consider the locally non-uniform distribution of distortion in images, we add a region-based fully convolutional layer for using the information around the center of the input image patch. The final score of the overall image is calculated by averaging of the patch scores. The proposed network performs in an end-to-end manner in both the training and testing phases. The experimental results on a series of benchmark datasets, e.g., LIVE, CISQ, IVC, TID2013, and Waterloo Exploration Database, show that the proposed algorithm outperforms the state-of-the-art methods, which verifies the effectiveness of our network architecture.

[Visualization studies on research status of the ventilator-associated pneumonia based on SinoMed database].

OBJECTIVE: To survey the distribution pattern and subject domain knowledge of the literatures about ventilator-associated pneumonia (VAP). METHODS: Literatures about VAP published until December 2017 were identified in SinoMed database for statistics and analysis. The information of author, organization and province was extracted by BICOMS software for generating co-occurrence matrix, at the same time, the topic words were cluster analyzed by Gcluto software to generate topical visual surface maps and visualization matrices, and the current research hotspots were analyzed. NetDraw from Ucinet 6.0 software was used to arrange the relationship among topic words according to the centrality, and the social network diagrams of authors, authors' provinces and institutions were draw to analyze the current status of VAP research cooperation. RESULTS: 4 851 VAP-related literatures were retrieved preliminarily, and 43 were excluded from abstracts, news reports, information and missing literatures. Finally, a total of 4 808 articles were enrolled in the visual analysis. From 2001 to 2004, the number of VAP-related literatures published was less than 10. Since 2009, the number of VAP documents had increased steadily, from 2010 to 2017, the peak period of publications reached 91.7% (4 411/4 808). According to the analysis of the amount of publications, the top three of 34 provincial administrative regions that published VAP-related literature in China were Guangdong Province (n = 628), Jiangsu Province (n = 478) and Zhejiang Province (n = 404), the number of hospitals issued by the First Affiliated Hospital of Chongqing Medical University was the largest (n = 20); there was only one journal with more than 100 articles, and there were 154 journals with only one article, accounting for 34.8% of the total number of journals. A total of 9 921 authors participated in the VAP-related literature writing, the number of high-yielding authors was not large, and the institution could not establish an effective social network diagram, suggesting that communication and cooperation should be strengthened in hospitals and outside hospitals. The results of the topic words social network analysis showed that the VAP research field was centered around the core of "mechanical ventilation", "intensive care unit (ICU)", "risk factor analysis", "nursing", "etiological analysis", "preventive measures" and "pathogens". The current research hotspots were at the edge of the network map, such as "drug sensitivity analysis", "Acinetobacter baumannii", "bronohoalveolar lavage (BAL)" and "acute exacerbation of chronic obstructive pulmonary disease (AECOPD)". By clustering 80 high-frequency topic words, at present, VAP research hotspots were mainly focus on five topics: obstructive pulmonary disease, especially in acute exacerbation, was prone to VAP; concerned about newborns and children's VAP; types, drug resistance and selection of antimicrobial agents for VAP pathogens in ICU; clinical efficacy and prognosis of VAP through preventive measures, pulmonary supportive care and comprehensive care interventions; oral care and airway management during mechanical ventilation was also the key aspect of the treatment of VAP. CONCLUSIONS: In recent years, the academics had attached great importance to the study of VAP, the number of publications had reached a historical peak, and the research direction was diverse. However, it was necessary to strengthen cooperation among research institutes, collect and count epidemiological data, improve and expand the research quality and scale of clinical diagnosis, nurse, prevention, pathogen distribution and drug resistance analysis.

An Adaptive Markov Random Field for Structured Compressive Sensing.

Exploiting intrinsic structures in sparse signals underpins the recent progress in compressive sensing (CS). The key for exploiting such structures is to achieve two desirable properties: generality (i.e., the ability to fit a wide range of signals with diverse structures) and adaptability (i.e., being adaptive to a specific signal). Most existing approaches, however, often only achieve one of these two properties. In this study, we propose a novel adaptive Markov random field sparsity prior for CS, which not only is able to capture a broad range of sparsity structures, but also can adapt to each sparse signal through refining the parameters of the sparsity prior with respect to the compressed measurements. To maximize the adaptability, we also propose a new sparse signal estimation where the sparse signals, support, noise and signal parameter estimation are unified into a variational optimization problem, which can be effectively solved with an alternative minimization scheme. Extensive experiments on three real-world datasets demonstrate the effectiveness of the proposed method in recovery accuracy, noise tolerance, and runtime.

[Study on cytotoxicity of three-dimensional printed ß-tricalcium phosphate loaded poly (lactide-co-glycolide) anti-tuberculosis drug sustained release microspheres and its effect on osteogenic differentiation of bone marrow mesenchymal stem cells].

Objective: To study the effect of three-dimensional (3D) printed ß-tricalcium phosphate (ß-TCP) scaffold loaded poly (lactide-co-glycolide) (PLGA) anti-tuberculosis drug sustained release microspheres on osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) and its cytotoxicity. Methods: Isoniazid and rifampicin/PLGA sustained release microspheres were prepared by W/O/W multiple emulsion method. The ß-TCP scaffolds were prepared by 3D printing technique. The microspheres were loaded on the scaffolds by centrifugal oscillation method to prepare composite materials. The BMSCs of Sprague Dawley rat were isolated and cultured by whole bone marrow adherent method, and the third generation cells were used for the following experiments. BMSCs were co-cultured with osteogenic induction medium (group A), PLGA anti-tuberculosis drug sustained release microsphere extract (group B), 3D printed ß-TCP scaffold extract (group C), and 3D printed ß-TCP scaffold loaded PLGA anti-tuberculosis drug sustained release microsphere composite extract (group D), respectively. Cytotoxicity was detected by cell counting kit 8 (CCK-8) method; the calcium deposition was observed by alizarin red staining; and the mRNA expressions of alkaline phosphatase (ALP), osteocalcin (OCN), and bone sialoprotein (BSP) were detected by real-time fluorescence quantitative PCR (RT-qPCR). Results: CCK-8 assay showed that the absorbance ( A) value of groups A, B, C, and D increased gradually with the culture time prolonging. After cultured for 24, 48, and 72 hours, the A value decreased in the order of groups A, C, B, and D. There was no significant difference between groups B and D ( P>0.05), but there were significant differences between other groups ( P<0.05). The cytotoxicity was evaluated as grade 0-2, and the toxicity test was qualified. Alizarin red staining showed that red mineralized nodules were formed in all groups at 21 days after osteogenic induction, but the number of mineralized nodules decreased sequentially in groups C, D, A, and B. RT-qPCR test results showed that the relative expressions of OCN and BSP genes in groups A, B, C, and D increased gradually with the culture time prolonging. The relative expression of ALP gene increased at 7 and 14 days, and decreased at 21 days. After cultured for 7, 14, and 21 days, the relative expressions of ALP, OCN, and BSP genes decreased sequentially in groups C, D, A, and B; the differences were significant between groups at different time points ( P<0.05). Conclusion: 3D printed ß-TCP loaded PLGA anti-tuberculosis drug sustained release microsphere composites have no obvious cytotoxicity to BMSCs, and can promote BMSCs to differentiate into osteoblasts to a certain extent.