Search

To solve the problem of poor face recognition performance in unrestricted environments. A face recognition algorithm based on improved residual IR-ResNet-SE is designed. Firstly, the IR structure is added to the 34-layer residual network to reduce the variability of different features; Secondly, we add the channel attention module to increase the weight of important channel features; Finally, the Arcface loss function is used to improve the classification ability of the model. The LFW, AgeDB, and AR datasets reflect unrestricted factors such as pose, age, expression, occlusion, and illumination. The algorithm proposed in this paper is experimented on these three datasets. The experimental results show that the IR-ResNet-SE algorithm proposed in this paper can achieve 99.74% accuracy in the dataset LFW. And it has excellent robustness in face recognition under unrestricted conditions. &nbsp

Under ideal illumination conditions, the existing face recognition algorithms can obtain better recognition results.However, under non-uniform illumination conditions, the performance of the face recognition algorithm will be greatly reduced. Moreover, the commonfacial image classification modelshavelowprecisionandspeed. To address these problems, this paper proposes a non-uniform illumination face recognition method based on the improved MobileNetV1, and constructs a lightweight feature extraction network with MobileNetV1 as the core. To reduce the influence of non-uniform illumination on the face image, we use the MSRCR algorithm to preprocess the face image. A lightweight I-MobileNet model is proposed. We introduce the attention module into the MobileNetV1 model, which enhances the feature extraction capability of the network and improves the recognition performance. At the same time, the model parameters are optimized with Arcface loss function to increase the distance between classes and reduce the distance within classes. Compared with the original MobileNetV1 model, our method shows a significant improvement in recognition accuracy. In comparison with other network models, experiments conducted on the Extended Yale B dataset with the I-MobileNet model demonstrate the effectiveness of the method. &nbsp

Shao-Long Chen,1 Hong Xiao,2 Guo-Jun Li,3 Yao-Jie Shen4 1Shulan International Medical College, Zhejiang Shuren University, Hangzhou, 310015, People’s Republic of China; 2Department of Infectious Diseases, Shanghai Public Health Clinical Center, Fudan University, Shanghai, 201508, People’s Republic of China; 3Department of Hepatology, The Second Hospital of Yinzhou of Ningbo, Ningbo, 315100, People’s Republic of China; 4Department of Infectious Diseases, Huashan Hospital, Fudan University, Shanghai, 200040, People’s Republic of ChinaCorrespondence: Shao-Long Chen, Shulan International Medical College, Zhejiang Shuren University, Hangzhou, 310015, People’s Republic of China, Email chenshaolong@zjsru.edu.cnPurpose: Chronic hepatitis B virus (CHB) infection is a worldwide health problem. Polyethylene glycol (PEG)ylated interferon (PEG-IFN) is an available therapy for CHB that has antiviral and immunomodulatory effects. However, PEG-IFN therapy is limited by the fact that only a subset of patients show a sustained response, its severe side effects, and high cost. The aim of this study was to explore novel biomarkers for the early prediction of PEG-IFN treatment response and to uncover its underlying mechanism.Patients and Methods: We enrolled 10 paired patients with Hepatitis B e antigen (HBeAg)-positive CHB who received PEG-IFN-α 2a monotherapy. Patient serum samples were collected at 0, 4, 12, 24, and 48 weeks and serum samples were collected from eight healthy people as healthy controls. For confirmation, we enrolled 27 patients with HBeAg-positive CHB receiving PEG-IFN therapy and serum samples at 0 and 12 weeks were obtained. Serum samples were analyzed using Luminex technology.Results: Among 27 assessed cytokines, 10 cytokines were identified to have high expression levels. Among them, six cytokines had significant differences in their levels between the patients with HBeAg-positive CHB and the healthy controls (P < 0.05). Potentially, treatment response could be predicted using the early time points of 4, 12, and 24 weeks. Moreover, after 12 weeks of PEG-IFN treatment, increased levels of pro-inflammatory cytokines and decreased levels of anti-inflammatory cytokines were observed. The fold change of IP-10 between 12 weeks and 0 weeks correlated with the decrease in ALT levels from 0 to 12 weeks (r = 0.2675, P = 0.0024).Conclusion: In patients with CHB, we observed a certain pattern in the levels of cytokines during treatment with PEG-IFN, and the cytokine IP-10 might be a potential biomarker for treatment response.Keywords: chronic hepatitis B, PEGylated interferon, cytokine, chemokine, Luminex

We present the Topology Transformation Equivariant Representation learning, a general paradigm of self-supervised learning for node representations of graph data to enable the wide applicability of Graph Convolutional Neural Networks (GCNNs). We formalize the proposed model from an information-theoretic perspective, by maximizing the mutual information between topology transformations and node representations before and after the transformations. We derive that maximizing such mutual information can be relaxed to minimizing the cross entropy between the applied topology transformation and its estimation from node representations. In particular, we seek to sample a subset of node pairs from the original graph and flip the edge connectivity between each pair to transform the graph topology. Then, we self-train a representation encoder to learn node representations by reconstructing the topology transformations from the feature representations of the original and transformed graphs. In experiments, we apply the proposed model to the downstream node classification, graph classification and link prediction tasks, and results show that the proposed method outperforms the state-of-the-art unsupervised approaches., Comment: Accepted to IEEE Transactions on Knowledge and Data Engineering (TKDE)

The development of cloud storage is indispensable for today’s Internet information technology. The superiority of cloud storage is that the ownership of the storage medium and the ownership of the stored data are divided and held by two parties with respective interests - cloud server providers and data users. However, the transmission, storage and utilization of data constantly caused the violation of users’ rights and interests. Data integrity as a countermeasure has been paid growing attention to by researchers. Traditional data integrity protection schemes are based on the cloud server-client architecture, which requires high computational overhead and communication. To address this scenario, this paper proposes an improved data integrity verification scheme that enables local users to perform rapid and efficient data integrity verification of data stored on cloud servers, reduces the volume of communication data during the verification process, and effectively prevents the loss or tampering of data stored in the cloud server. &nbsp

Facial expression recognition is widely used, but there are some problems such as complex scenes, lack of data sets and low recognition rate. In this paper, we construct a new network model and name it RNFC. The RNFC network adopts 6 improved residual blocks to extract features. Features are passed into the fully connected layer by flattening the data, and Dropout techniques are introduced between the fully connected layers to prevent overfitting of the model. Based on the pytorch framework, we use a cross-entropy loss function to improve the training speed of the network. And perform denoising and enhancement pre-processing on the FER2013 dataset. The RNFC network is trained and tested on the pretreated FER2013. It has a higher recognition rate than classical networks such as VGGnet19 and ResNet18. &nbsp

Photocatalytic CO

Nano zero-valent iron biochar (nZVI–BC), an environmentally-friendly functional material prepared from waste biomass, has attracted extensive attention. This material has potential to solve the problem of biomass conversion. However, the lack of a method of converting biomass to the nZVI–BC involved in biomass modification and pyrolysis hinders its further production and application. In this study, we introduced the green solvent polyethylene glycol 400 (PEG400) to a biomass (rice straw, RS) modification system with FeCl3·6H2O, and activated RS was prepared to nZVI–BC by one-step pyrolysis. The addition of PEG400 promoted the hydrolysis of iron ion and improved the RS surface structure, promoting the attachment of Fe2O3 to the RS surface. The mild activation conditions with temperatures of 60, 80, 100 °C and a time of 0.5 h prevented the excessive loss of the lignin component and were conducive to the formation of carbon skeletons. Amorphous carbon and Fe2O3 were subjected to redox reactions to form nZVI–BC with the assistance of reducing gas produced from pyrolysis. In addition, the prepared nZVI–BC was tested for dye (Congo red) removal, showing rapid absorption (70.6% at 5 min) and high catalysis in advanced oxidation (75.67% at 5 min, 90% at 60 min). This work proposed a novel mechanistic strategy for preparing nZVI–BC and set a foundation for its scaled production and application.

The using of O atoms to modify metallophthalocyanine sheets effectively adjusts the orbital arrangement of the electronic configuration of d-orbitals in transition metals, thereby modulating the MAE of the structure.

An iodine-containing reagent promoted three-component method for the selective synthesis of phenothiazines and bis-phenothiazines from simple and cheap cyclohexanones, elemental sulfur, and inorganic ammonium salts has been developed.

Radical-mediated photoredox hydroarylation reaction of unactivated alkenes using thiosulfonate compounds as key radical precursors to synthesize pyridyquinazolinone and pyrroloquinazolinone derivatives is described.

Surgical treatment is the most important and effective therapy for resectable esophageal cancer. Minimally invasive esophagectomy (MIE) can reduce surgical trauma. A neck incision can be performed for extraction of surgical specimen. This study was performed to investigate the safety and feasibility of neck incision to extract surgical specimen in thoracolaparoscopic esophagectomy for esophageal cancer.Thirty-four patients who experienced thoracolaparoscopic esophagectomy for esophageal cancer and a neck incision for extraction of surgical specimen were enrolled. The clinical, surgical and follow-up data were analyzed.The procedure was successful in all patients (100%), with a neck incision to extract the surgical specimen. The median surgical time was 309 min, and the median blood loss was 186 ml, with the mean length of hospital stay of 11.5 days. Pulmonary complications occurred in 8 patients (23.5%). Anastomotic leakage occurred in 5 patients (14.7%), with one patient being treated conservatively to recover and four (11.8%) who received interventional drainage. One patient with interventional drainage died of severe infection, resulting in a 30-day surgical mortality of 2.9% (n = 1). Gastrointestinal complications happened in 5 patients (14.7%), including ileus in three patients and anastomotic stenosis in two patients. Follow-up was performed at a median time of 20 months (interquartile range, 14-32 months), with no death during this period. No recurrence was found in the first 12 months after radical resection.The cervical incision to extract surgical specimen is safe and feasible with improved cosmetic effect in thoracolaparoscopic esophagectomy for esophageal cancer.

TBHP-mediated thiyl couplings: A mild and metal-free visible-light-induced system allows sequential thiolation/cyclization of N-aryl acrylamides and mercaptans and provides an access to thionated oxindoles with potential biological activity.

Sustainable photoredox catalysis with bromine radical-mediated hydrogen atom transfer enables stoichiometric pyridylation to synthesize a wide range of diarylmethanes and triarylmethanes in good to excellent yields.

Recent progress in the assembly of chiral α-tertiary amines in the last decade via transition-metal catalysis is summarized.

To explore the effectiveness and safety of laparoscopic transhiatal complete mesenteric resection (CME) surgery compared with those of the traditional laparoscopic transhiatal approach in the treatment of Siewert II/III adenocarcinoma of the esophagogastric junction (AEG).Ninety-nine patients with Siewert type II/III AEG were enrolled and divided into two groups: the laparoscopic CME transhiatal approach (CEM-TH, n = 61) group and traditional laparoscopic transhiatal (TH, n = 38) group. Intraoperative and postoperative clinical data of both groups were analyzed.The laparoscopic trasihiatal surgery was technically successful in all patients. The surgical time, intraoperative bleeding, and hospital stay were all significantly (P 0.05) reduced in the CME-TH group compared with those in the TH group. The levels of white blood cells on postoperative day (POD) 1 and 5, postoperative CRP on POD 3 and 5, and postoperative PCT were significantly (P 0.05) lower while lymph nodes were harvested significantly (P 0.05) more in the CME-TH group than in the TH group. Complications were not significantly (P 0.05) different between two groups. No death occurred within 90 days.The CME theory could be safely and effectively applied laparoscopically to treat patients with Siewert II/III AEG. Mesogastrium and lower mesoesophagus can be completely resected together with the tumor, lymph nodes, adipose tissue, and blood vessels as an "intact package," leading to better short-term outcomes.

Peizhang Wu,1,2 Jun Guo,2 Hongwei Yang,2 Debin Yuan,2 Chaoxiang Wang,2 Zhong Wang1 1Department of Neurosurgery, The First Affiliated Hospital of Soochow University, Suzhou, 215006, People’s Republic of China; 2Department of Neurosurgery, Yancheng First People’s Hospital, Yancheng, 224000, People’s Republic of ChinaCorrespondence: Zhong Wang, Department of Neurosurgery, The First Affiliated Hospital of Soochow University, No. 188 Shizi Street, Suzhou, 215006, People’s Republic of China, Email wangzhong6439@163.comBackground: Hypoxia is a frequent feature of solid tumors which significantly affects the efficacy of treatments such as chemotherapy. In addition, exosomes from hypoxic cancer cells could contribute to the chemoresistance of tumor cells through carrying miRNAs. It has been shown that miR-106-5p level was upregulated in glioma. However, whether exosomes derived from hypoxic glioma cells could affect temozolomide (TMZ) resistance in glioma through carrying miR-106a-5p remains unexplored.Methods: Exosomes were isolated from glioma cells under normoxia or hypoxia condition. EdU staining and flow cytometry assays were used to assess the cell proliferation and cell apoptosis. The relation between miR-106a-5p and PTEN was investigated by dual luciferase assay.Results: MiR-106a-5p was enriched in exosomes derived from hypoxic glioma cells compared to exosomes from cells under normoxia condition. Additionally, hypoxic glioma cells were able to transfer exosomes to glioma cells, resulting in a significant increase of miR-106a-5p level in cells. TMZ remarkably suppressed glioma cell proliferation and triggered cell apoptosis. However, hypoxic glioma cell-derived exosomes markedly promoted the proliferation and suppressed the apoptosis in TMZ-treated glioma cells, and miR-106a-5p inhibitor was able to abolish these phenomena. Meanwhile, PTEN was verified to be a direct target of miR-106a-5p. Furthermore, TMZ elevated PTEN and Bax level and reduced p-Akt level in glioma cells, whereas these changes were reversed by hypoxia glioma cell-derived exosomes. Furthermore, hypoxia glioma cell-derived exosomes reduced the sensitivity of glioma cells to TMZ in vivo via downregulating PTEN.Conclusion: Collectively, exosomal miR-106a-5p derived from hypoxia glioma cells could reduce the sensitivity of glioma cells to TMZ through downregulating PTEN. Thus, our study might provide new strategies for improving the clinical efficacy of TMZ on glioma.Keywords: cancer, hypoxia, miR-106a-5p, chemosensitivity

The path toward field-effect transistor (FET) application from laboratory to clinic has delivered a compelling push in the biomedical domain, yet ultrasensitive and timely pathogen identification without PCR remains a long-lasting challenge. Herein, we create a generic check station termed "CRISPR-FET", first incorporating the CRISPR/Cas13a system within the FET modality, for accelerated and unamplified detection of viral RNA. Unlike conventional FETs bearing target-specific receptors, this sensor holds three unique advancements: (i) an ingenious sensing mechanism is used, which converts the signal of a large-sized analyte into an on-chip cleavage response of an immobilized CRISPR reporter, enabling signal generation events to occur all within the Debye length; (ii) the multipurpose inspection of the CoV ORF1ab, CoV N gene, and HCV RNA unveils the potential for "one-for-all" scalable FET-based molecular diagnostics; and (iii) it is shown that Cas13a-crRNAs targeting different sites of the viral genome can be deployed in tandem to amplify the FET response, empowering the detection limit down to 1.56 aM, which is a world-record level of sensitivity in the FET for direct viral gene sensing. Notably, a brilliant clinical applicability was made in distinguishing HCV-infected patients from normal controls. Overall, this study sheds new insights into FET-based nucleic acid sensing technology and invokes a vision for its possible future roles in diagnosis of various viral diseases.