Search

With the rise of deep learning technology, the field of medical image segmentation has undergone rapid development. In recent years, convolutional neural networks (CNNs) have brought many achievements and become the consensus in medical image segmentation tasks. Although many neural networks based on U-shaped structures and methods, such as skip connections have achieved excellent results in medical image segmentation tasks, the properties of convolutional operations limit their ability to effectively learn local and global features. To address this problem, the Transformer from the field of natural language processing (NLP) was introduced to the image segmentation field. Various Transformer-based networks have shown significant performance advantages over mainstream neural networks in different visual tasks, demonstrating the huge potential of Transformers in the field of image segmentation. However, Transformers were originally designed for NLP and ignore the multidimensional nature of images. In the process of operation, they may destroy the 2D structure of the image and cannot effectively capture low-level features. Therefore, we propose a new multi-scale cross-attention method called M-VAN Unet, which is designed based on the Visual Attention Network (VAN) and can effectively learn local and global features. We propose two attention mechanisms, namely MSC-Attention and LKA-Cross-Attention, for capturing low-level features and promoting global information interaction. MSC-Attention is designed for multi-scale channel attention, while LKA-Cross-Attention is a cross-attention mechanism based on the large kernel attention (LKA). Extensive experiments show that our method outperforms current mainstream methods in evaluation metrics such as Dice coefficient and Hausdorff 95 coefficient.

Introduction: Esophageal variceal bleeding (EVB) is a potentially fatal complication of cirrhosis. The purpose of the present study was to evaluate the safety and efficacy of a novel technique of balloon-occluded esophageal varices (EVs) obliteration (BEVO) for EVs classified as F2 (medium size, F2-EVs) and F3 (large size, F3-EVs). Materials and Methods: Between December 2020 and December 2021, a total of 73 consecutive patients with EVs were treated using BEVO. An injection of sclerosant was administered via direct puncture of the varices during balloon occlusion. Immediate postprocedural Doppler endoscopic ultrasonography (EUS) was conducted to evaluate the blood flow in the EVs. Several factors, including the technical success, controlling of acute bleeding, intraoperative bleeding at the injection site, variceal eradication, variceal recurrence, and BEVO-related complications, were assessed. Results: BEVO was successfully performed in all patients. Immediate hemostasis was achieved in 100% (25/25) of patients with active EVB. The incidence of injection site intraoperative bleeding presenting as oozing and spurting bleeding was 76.71% (56/73) and 8.22% (6/73), respectively. Based on Doppler EUS and endoscopic examination, EVs were completely eradicated in 71 out of 73 patients (97.26%) after three sessions. A total of 3 (4.11%) patients were diagnosed with EV recurrence during follow-up after complete EV eradication. Retrosternal chest discomfort (13.70%; 10/73) and abdominal bloating (2.74%; 2/73) were spontaneously relieved after a few days. No serious BEVO-related complications were observed. Discussion: BEVO is a convenient and effective treatment for obliterating F2 and F3 EVs. Clinical Trial Registration No. ChiCTR2000039974. [ABSTRACT FROM AUTHOR]

The optimal management of esophageal variceal bleeding (EVB) and portal vein thrombosis (PVT) in liver cirrhosis has not been well-established. The aim of the present study was to compare the efficacy and safety of transjugular intrahepatic portosystemic shunt (TIPS) and endoscopic treatment (ET) plus anticoagulation in cirrhotic patients with EVB and PVT. A total of 66 cirrhotic patients with PVT and EVB (31 in the TIPS group and 35 in the ET plus anticoagulation group) were evaluated retrospectively between January 2016 and January 2022. During the follow-up period, 85.5% of patients in the TIPS group achieved complete recanalization of the portal vein, as compared with 19.6% in the ET plus anticoagulation group (p <.001). The cumulative 5-year rate of variceal rebleeding in the TIPS group was significantly lower than that in the ET plus anticoagulation group (31.0 vs. 50.1%; p =.017). The TIPS group exhibited a significantly higher incidence of overt hepatic encephalopathy (HE) than the ET plus anticoagulation group (25.8 vs. 5.7%; p =.037). No difference in the 5-year survival rate (74.1 vs. 85.7%; p =.692) and probability of other complications was observed between the two groups. TIPS was superior to ET plus anticoagulation in preventing variceal rebleeding and achieving recanalization of PVT but increased the incidence of overt HE without improving the survival rate. [ABSTRACT FROM AUTHOR]

[Display omitted] • A preparation method of CoFe alloy carbide catalyst was proposed. • Co-Fe interdiffusion happened in reduction and affect the alloy carbide formation. • ε'-(Co x Fe 1-x) 2.2 C facilitated the synergy between CO insertion and dissociation. • 0.25 wt% Na promoted ε'-(Co x Fe 1-x) 2.2 C formation and enhanced STY of higher alcohols. Higher alcohols synthesis (HAS) from syngas is quite attractive but still challenging due to the unsatisfied product selectivity and stability. Here, we prepared a series of CoFe bimetallic carbide catalysts and investigated the evolution of active species. The catalysts reduced above 300 °C could ensure an adequate interdiffusion between Co and Fe species to form CoFe alloy and then achieved a high fraction of ε'-(Co x Fe 1-x) 2.2 C species during HAS reaction, which exhibited excellent catalytic performance and stability over 200 h. The ε'-(Co x Fe 1-x) 2.2 C, acting as uniformly atomic neighboring Co x C-Fe x C dual sites, facilitated the synergy between CO nondissociative adsorption and CO dissociation, thus boosting the selective production of higher alcohols. It is also demonstrated that adding proper content of Na (0.25 wt%) can enhance the formation of ε'-(Co x Fe 1-x) 2.2 C as well as modulate the surface concentration of intermediates. These results may provide new idea for the design of high-performance catalysts for HAS. [ABSTRACT FROM AUTHOR]

In this paper, the viewing-angle-switching (VAS) films based on polymer dispersed liquid crystals (PDLCs) were fabricated and the optical properties of the VAS films were studied. The VAS films based on PDLC films can exhibit electrically switchable viewing-angle properties, and the films showed a narrow viewing angle under no electric field and presented a wide viewing angle under electric field (24 V). Furthermore, the smart anti-peeping liquid crystal display (LCD) containing the VAS film based on PDLC film also displayed the electrically switchable viewing-angle performance and the corresponding mechanism was discussed. The results demonstrated both the anti-peeping effect and electrically switchable view-angle effect of the as-fabricated VAS film, promising potential applications in intelligent optical films and novel information displays. [ABSTRACT FROM AUTHOR]

WordNets organize words into synonymous word sets, and the connections between words present the semantic relationships between them, which have become an indispensable source for natural language processing (NLP) tasks. With the development and evolution of languages, WordNets need to be constantly updated manually. To address the problem of inadequate word semantic knowledge of "new words", this study explores a novel method to automatically update the WordNet knowledge base by incorporating word-embedding techniques with sememe knowledge from HowNet. The model first characterizes the relationships among words and sememes with a graph structure and jointly learns the embedding vectors of words and sememes; finally, it synthesizes word similarities to predict concepts (synonym sets) of new words. To examine the performance of the proposed model, a new dataset connected to sememe knowledge and WordNet is constructed. Experimental results show that the proposed model outperforms the existing baseline models. [ABSTRACT FROM AUTHOR]

With the broad application of iris recognition systems, iris images acquired under cross-sensor and cross-spectrum conditions are increasingly common. Differences in the information stored in these images place high requirements for accuracy and robustness on cross-domain iris recognition algorithms, and current algorithms have not performed well in cross-domain iris recognition. Hence, we propose a dense squeeze and excitation network (DenseSENet) to extract common features in cross-domain iris images and use real-valued features for cross-domain matching. DenseSENet is a deep convolutional neural network based on an attention mechanism. It uses the batch normalization-rectified linear unit-Conv1 × 1 structure to increase the nonlinearity of the network and repeatedly uses features through dense connections to effectively prevent overfitting. The proposed dense squeeze and excitation structure extracts more accurate and stable features by learning the relations between channels and assigning corresponding attention values. The structure of global average pooling-fully connected-softmax is utilized to obtain compact and normalized real-valued feature templates to effectively reduce the size of the iris template. The proposed cross-domain iris recognition framework based on DenseSENet is end-to-end, which transforms the cross-domain matching problem to a many-to-one classification problem. Experiments are conducted on publicly available cross-sensor and cross-spectral iris databases. The results show that the proposed real-valued feature matching based on DenseSENet has better matching accuracy and robustness than the state-of-art cross-domain matching algorithm on two publicly available cross-sensor and cross-spectral iris databases. [ABSTRACT FROM AUTHOR]

With the rapid development of deep learning, iris recognition methods based on deep learning are constantly being proposed. These methods generally consist of iris segmentation and normalization to more accurately locate iris regions and reduce the impact of iris feature changes caused by pupil expansion. We propose a non-segmentation (NS) iris recognition framework based on a deep learning classification model, which directly takes a raw image as input for feature extraction and recognition without performing iris segmentation and normalization. This method outperforms other methods. We proposed a non-segmentation network (NSNet). NSNet is a convolutional neural network (CNN) based on an attention mechanism that enhances the robustness and accuracy of the network by reusing features and assigning channel attention values. In addition, while ensuring the advanced performance of the network, it uses only 31 convolutional layers to complete iris feature extraction and recognition tasks in the NS iris recognition framework. Since the deep learning classification model cannot recognize that the category of an image is an untrained image category, we proposed a dual-threshold iris framework. In the proposed framework, all untrained image categories are classified as impostor classes, and the first threshold set in the proposed framework can effectively prevent impostors and eliminate inferior images. The proposed framework is suitable for the pursuit of more accurate, more robust, and safer private iris recognition scenarios. We conduct experiments with uniform parameter settings on four publicly available databases and shows that even in the challenging situation where a raw image is used as the input image, the proposed method is still the most advanced algorithm. Moreover, a series of ablation experiments were conducted to further confirm that the proposed framework has a higher accuracy, robustness, and generalizability through verification and discussion. [ABSTRACT FROM AUTHOR]

Iris recognition (IR) is widely adopted for human identification in the biometric fields. However, results using handcrafted features designed with the domain knowledge are obtained by conventional IR algorithms which are complex. Convolutional neural networks (convnets) have been proved to be robust with superior learning capability in many computer vision fields. The deep convnets perform better than shallower convnets. However, training the deep convnets for IR is difficult. IrisConvDeeper, an adaptive architecture based on convnets combining Softmax classifier that is used to alleviate the problems of training deep convnets for IR, is proposed. In addition, a better performance is obtained. Furthermore, a shallower architecture termed as IrisConvShallower is designed for comparison. IrisConvDeeper consists of structures called dense block and has 12 or 14 convolutional layers. Meanwhile, IrisConvShallower consists of six or seven standard convolutional layers. The performance of the proposed architectures is tested on two public iris databases, which are collected under different conditions: CASIA-Iris-V3 and IITD iris image databases. The experimental results demonstrate that the proposed IrisConvDeeper outperforms most of state-of-the-art approaches in terms of correct recognition rate, except some approaches that adopted specific methods and identified fewer subjects compared with our methods. [ABSTRACT FROM AUTHOR]

• The increased ratio of Co 2 C to Co with GO addition linearly increased the alcohols selectivity. • GO addition could enhance Co dispersion and act as an electron donor to promote CO adsorption and disassociation. • Too many GO addition would lead to stacks and then cover part of active sites and weaken the electron donor effect. Higher alcohols synthesis (HAS) directly from syngas has attracted lots of attentions due to its advantages on environmental-friendliness, versatility and economy. Recently, researches proposed that the Co-Co 2 C dual sites could act as efficient active species for HAS. But the role of carbides as well as its formation process needs further investigation. Here, we synthesized a series of graphene oxide (GO)-ordered mesoporous silica (OMS) composite supported cobalt-based catalysts, which presented excellent activity and selectivity in HAS. It is found that the addition of GO could not only improve the dispersion of cobalt species, but also act as an electron donor to promote CO adsorption and disassociation on Co species, both leading to a higher catalytic activity. The electron donor effect of GO could facilitate the carbide formation process as well. The increased ratio of Co 2 C/Co with GO addition linearly improved the alcohol selectivity in HAS. However, excess GO addition would cover a part of active sites and weaken the electron donor effect, resulting in a lower alcohol yield. This work gives insights into the effects of carbon materials on active sites distribution and performance of cobalt-based catalysts, which could be used to rationalize the design of catalysts for HAS. [ABSTRACT FROM AUTHOR]