Dendritic quinary PtRhMoCoFe high-entropy alloy as a robust immunosensing nanoplatform for ultrasensitive detection of biomarker

Recently, high-entropy alloys have superior physicochemical properties as compared to conventional alloys for their glamorous "cocktail effect". Nevertheless, they are scarcely applied to electrochemical immunoassays until now. Herein, uniform PtRhMoCoFe high-entropy alloyed nanodendrites (HEANDs) were synthesized by a wet-chemical co-reduction method, where glucose and oleylamine behaved as the co-reducing agents. Then, a series of characterizations were conducted to illustrate the synergistic effect among multiple metals and fascinating structural characteristics of PtRhMoCoFe HEANDs. The obtained high-entropy alloy was adopted to build a electrochemical label-free biosensor for ultrasensitive bioassay of biomarker cTnI. In the optimized analytical system, the resultant sensor exhibited a dynamic linear range of 0.0001-200 ng mL-1 and a low detection limit of 0.0095 pg mL-1 (S/N = 3). Eventually, this sensing platform was further explored in serum samples with satisfied recovery (102.0 %). This research renders some constructive insights for synthesis of high-entropy alloys and their expanded applications in bioassays and bio-devices.

Filler Mixed Into Adhesives Does Not Necessarily Improve Their Mechanical Properties

OBJECTIVES

To investigate the influence of filler type/loading on the micro-tensile fracture strength (μTFS) of adhesive resins, as measured 'immediately' upon preparation and after 1-week water storage ('water-stored').

METHODS

The morphology and particle-size distribution of three filler particles, referred to as 'Glass-S' (Esschem Europe), 'BioUnion' (GC), and 'CPC_Mont', were correlatively characterized by SEM, TEM, and particle-size analysis. These filler particles were incorporated into an unfilled adhesive resin ('BZF-29unfilled', GC) in different concentrations to measure the 'immediate' μTFS. After 1-week water storage, the 'water-stored' μTFS of the experimental particle-filled adhesive resins with the most optimum filler loading, specific for each filler type, was measured. In addition, the immediate and water-stored μTFS of the adhesive resins of three experimental two-step universal adhesives based on the same resin matrix but varying for filler type/loading, coded as 'BZF-21' (containing silica and bioglass), 'BZF-29' (containing solely silica), and 'BZF-29_hv' (highly viscous with a higher silica loading than BZF-29), and of the adhesive resins of the gold-standard adhesives OptiBond FL ('Opti-FL', Kerr) and Clearfil SE Bond 2 ('C-SE2', Kuraray Noritake) was measured along with that of BZF-29unfilled (GC) serving as control/reference. Statistics involved one-way and two-way ANOVA followed by post-hoc multiple comparisons (α<0.05).

RESULTS

Glass-S, BioUnion, and CPC_Mont represent irregular fillers with an average particle size of 8.5-9.9 μm. Adding filler to BZF-29unfilled decreased μTFS regardless of filler type/loading. One-week water storage reduced μTFS of all adhesive resins except BZF-21, with the largest reduction in μTFS recorded for BZF-29unfilled. Among the three filler types, the μTFS of the 30 wt% Glass-S and 20 wt% BioUnion filled adhesive resin was not significantly different from the μTFS of BZF-29unfilled upon water storage.

CONCLUSIONS

Adding filler particles into adhesive resin did not enhance its micro-tensile fracture strength but appeared to render it less sensitive to water storage as compared to the unfilled adhesive resin investigated.

On the Consistency and Large-Scale Extension of Multiple Kernel Clustering

Existing multiple kernel clustering (MKC) algorithms have two ubiquitous problems. From the theoretical perspective, most MKC algorithms lack sufficient theoretical analysis, especially the consistency of learned parameters, such as the kernel weights. From the practical perspective, the high complexity makes MKC unable to handle large-scale datasets. This paper tries to address the above two issues. We first make a consistency analysis of an influential MKC method named Simple Multiple Kernel k-Means (SimpleMKKM). Specifically, suppose that ∧γn are the kernel weights learned by SimpleMKKM from the training samples. We also define the expected version of SimpleMKKM and denote its solution as γ*. We establish an upper bound of ||∧γn-γ*||∞ in the order of ~O(1/√n), where n is the sample number. Based on this result, we also derive its excess clustering risk calculated by a standard clustering loss function. For the large-scale extension, we replace the eigen decomposition of SimpleMKKM with singular value decomposition (SVD). Consequently, the complexity can be decreased to O(n) such that SimpleMKKM can be implemented on large-scale datasets. We then deduce several theoretical results to verify the approximation ability of the proposed SVD-based method. The results of comprehensive experiments demonstrate the superiority of the proposed method. The code is publicly available at https://github.com/weixuan-liang/SVD-based-SimpleMKKM.

Effect of COVID-19 infection on pregnant women in plateau regions

OBJECTIVE

The present study aims to explore the effect of COVID-19 infection on pregnant women in plateau regions.

STUDY DESIGN

Data from 381 pregnant women infected with COVID-19 who underwent prenatal examination or treatment at Women and Children's Hospital of Tibet Autonomous Region between January 2020 and December 2022 and 314 pregnant women not infected with COVID-19 were retrospectively collected.

METHODS

The study participants were divided into an infected and non-infected group according to whether they were infected with COVID-19. Basic information (ethnicity, age, body mass index and gestational age [GA]), vaccination status, intensive care unit (ICU) admission and delivery outcomes were compared. Binary logistic regression was used to analyse the influencing factors of ICU admission.

RESULTS

The results revealed significant differences in the GA, vaccination rate, blood pressure, partial pressure of oxygen, white blood cell (WBC) count, ICU admission rate, preeclampsia rate, forearm presentation rate, thrombocytopenia rate, syphilis infection rate and placental abruption rate between the two groups (P < 0.05). A univariate analysis showed that COVID-19 infection, hepatitis B virus infection, the WBC count and hypoproteinaemia were risk factors for ICU admission. The results of the multivariate analysis of the ICU admission of pregnant women showed that COVID-19 infection (odds ratio [OR] = 4.271, 95 % confidence interval [CI]: 3.572-5.820, P < 0.05) was a risk factor for ICU admission and the WBC count (OR = 0.935, 95 % CI: 0.874-0.947, P < 0.05) was a protective factor for ICU admission.

CONCLUSION

Pregnant women are vulnerable to the adverse consequences of COVID-19 infection, and public health measures such as vaccination are needed to protect this population subgroup.

Anchor-Sharing and Clusterwise Contrastive Network for Multiview Representation Learning

Multiview clustering (MVC) has gained significant attention as it enables the partitioning of samples into their respective categories through unsupervised learning. However, there are a few issues as follows: 1) many existing deep clustering methods use the same latent features to achieve the conflict objectives, namely, reconstruction and view consistency. The reconstruction objective aims to preserve view-specific features for each individual view, while the view-consistency objective strives to obtain common features across all views; 2) some deep embedded clustering (DEC) approaches adopt view-wise fusion to obtain consensus feature representation. However, these approaches overlook the correlation between samples, making it challenging to derive discriminative consensus representations; and 3) many methods use contrastive learning (CL) to align the view's representations; however, they do not take into account cluster information during the construction of sample pairs, which can lead to the presence of false negative pairs. To address these issues, we propose a novel multiview representation learning network, called anchor-sharing and clusterwise CL (CwCL) network for multiview representation learning. Specifically, we separate view-specific learning and view-common learning into different network branches, which addresses the conflict between reconstruction and consistency. Second, we design an anchor-sharing feature aggregation (ASFA) module, which learns the sharing anchors from different batch data samples, establishes the bipartite relationship between anchors and samples, and further leverages it to improve the samples' representations. This module enhances the discriminative power of the common representation from different samples. Third, we design CwCL module, which incorporates the learned transition probability into CL, allowing us to focus on minimizing the similarity between representations from negative pairs with a low transition probability. It alleviates the conflict in previous sample-level contrastive alignment. Experimental results demonstrate that our method outperforms the state-of-the-art performance.

Hierarchical and dynamic graph attention network for drug-disease association prediction

In the realm of biomedicine, the prediction of associations between drugs and diseases holds significant importance. Yet, conventional wet lab experiments often fall short of meeting the stringent demands for prediction accuracy and efficiency. Many prior studies have predominantly focused on drug and disease similarities to predict drug-disease associations, but overlooking the crucial interactions between drugs and diseases that are essential for enhancing prediction accuracy. Hence, in this paper, a resilient and effective model named Hierarchical and Dynamic Graph Attention Network (HDGAT) has been proposed to predict drug-disease associations. Firstly, it establishes a heterogeneous graph by leveraging the interplay of drug and disease similarities and associations. Subsequently, it harnesses the capabilities of graph convolutional networks and bidirectional long short-term memory networks (Bi-LSTM) to aggregate node-level information within the heterogeneous graph comprehensively. Furthermore, it incorporates a hierarchical attention mechanism between convolutional layers and a dynamic attention mechanism between nodes to learn embeddings for drugs and diseases. The hierarchical attention mechanism assigns varying weights to embeddings learned from different convolutional layers, and the dynamic attention mechanism efficiently prioritizes inter-node information by allocating each node with varying rankings of attention coefficients for neighbour nodes. Moreover, it employs residual connections to alleviate the over-smoothing issue in graph convolution operations. The latent drug-disease associations are quantified through the fusion of these embeddings ultimately. By conducting 5-fold cross-validation, HDGAT's performance surpasses the performance of existing state-of-the-art models across various evaluation metrics, which substantiates the exceptional efficacy of HDGAT in predicting drug-disease associations.

Boundary uncertainty aware network for automated polyp segmentation

Recently, leveraging deep neural networks for automated colorectal polyp segmentation has emerged as a hot topic due to the favored advantages in evading the limitations of visual inspection, e.g., overwork and subjectivity. However, most existing methods do not pay enough attention to the uncertain areas of colonoscopy images and often provide unsatisfactory segmentation performance. In this paper, we propose a novel boundary uncertainty aware network (BUNet) for precise and robust colorectal polyp segmentation. Specifically, considering that polyps vary greatly in size and shape, we first adopt a pyramid vision transformer encoder to learn multi-scale feature representations. Then, a simple yet effective boundary exploration module (BEM) is proposed to explore boundary cues from the low-level features. To make the network focus on the ambiguous area where the prediction score is biased to neither the foreground nor the background, we further introduce a boundary uncertainty aware module (BUM) that explores error-prone regions from the high-level features with the assistance of boundary cues provided by the BEM. Through the top-down hybrid deep supervision, our BUNet implements coarse-to-fine polyp segmentation and finally localizes polyp regions precisely. Extensive experiments on five public datasets show that BUNet is superior to thirteen competing methods in terms of both effectiveness and generalization ability.

Predicting miRNA-disease association via graph attention learning and multiplex adaptive modality fusion

miRNAs are a class of small non-coding RNA molecules that play important roles in gene regulation. They are crucial for maintaining normal cellular functions, and dysregulation or dysfunction of miRNAs which are linked to the onset and advancement of multiple human diseases. Research on miRNAs has unveiled novel avenues in the realm of the diagnosis, treatment, and prevention of human diseases. However, clinical trials pose challenges and drawbacks, such as complexity and time-consuming processes, which create obstacles for many researchers. Graph Attention Network (GAT) has shown excellent performance in handling graph-structured data for tasks such as link prediction. Some studies have successfully applied GAT to miRNA-disease association prediction. However, there are several drawbacks to existing methods. Firstly, most of the previous models rely solely on concatenation operations to merge features of miRNAs and diseases, which results in the deprivation of significant modality-specific information and even the inclusion of redundant information. Secondly, as the number of layers in GAT increases, there is a possibility of excessive smoothing in the feature extraction process, which significantly affects the prediction accuracy. To address these issues and effectively complete miRNA disease prediction tasks, we propose an innovative model called Multiplex Adaptive Modality Fusion Graph Attention Network (MAMFGAT). MAMFGAT utilizes GAT as the main structure for feature aggregation and incorporates a multi-modal adaptive fusion module to extract features from three interconnected networks: the miRNA-disease association network, the miRNA similarity network, and the disease similarity network. It employs adaptive learning and cross-modality contrastive learning to fuse more effective miRNA and disease feature embeddings as well as incorporates multi-modal residual feature fusion to tackle the problem of excessive feature smoothing in GATs. Finally, we employ a Multi-Layer Perceptron (MLP) model that takes the embeddings of miRNA and disease features as input to anticipate the presence of potential miRNA-disease associations. Extensive experimental results provide evidence of the superior performance of MAMFGAT in comparison to other state-of-the-art methods. To validate the significance of various modalities and assess the efficacy of the designed modules, we performed an ablation analysis. Furthermore, MAMFGAT shows outstanding performance in three cancer case studies, indicating that it is a reliable method for studying the association between miRNA and diseases. The implementation of MAMFGAT can be accessed at the following GitHub repository: https://github.com/zixiaojin66/MAMFGAT-master.

Can digitalization reduce industrial pollution? Roles of environmental investment and green innovation

Industrialized nations have witnessed a decline in environmental quality over the years. The potential of digitalization in mitigating environmental pollution is of significant interest. Drawing on firm-level data from listed Chinese companies between 2010 and 2020, including pollutant and financial metrics, this study investigates the influence of digitalization on industrial environmental pollution. We found that digitalization substantially diminishes the intensity of industrial pollution emissions. These findings hold even after employing instrumental variable tests, substituting the dependent variable with carbon dioxide emissions, and conducting a quasi-natural experiment in intelligent manufacturing. Moreover, our exploration of the underlying mechanisms reveals that the decline in pollution emission intensity attributable to digitalization stems from both structural and technological factors; specifically, it enhances environmental investment and fosters green innovation. The benefits of digitalization in curbing emission intensity are pronounced for firms characterized by lower pollution levels, executive leadership with environmental work backgrounds, heightened capital intensity, and elevated media coverage.

Efficacy and Safety of Hydrogen Therapy in Patients with Early-Stage Interstitial Lung Disease: A Single-Center, Randomized, Parallel-Group Controlled Trial

Purpose

Several in vivo experiments have shown that molecular hydrogen is a promising therapeutic agent for interstitial lung diseases (ILD). In this study, hydrogen therapy was investigated to determine whether it is superior to N-Acetylcysteine (NAC) for the treatment of patients with early-stage ILD.

Patients and Methods

A prospective, single-center, randomized, controlled clinical trial was conducted in 87 patients with early-stage ILD. Hydrogen or NAC therapy was randomly assigned (1:1 ratio) to the eligible patients. The primary endpoint was the change in the high-resolution computed tomography (HRCT) and composite physiologic index (CPI) scores from baseline to week 48. Pulmonary function was evaluated as a secondary endpoint, and adverse events were recorded for safety analysis.

Results

The rate of HRCT image improvement from the baseline in the HW group (63.6%) was higher than that in the NAC group (39.5%). A significant decrease in CPI and improvement in DLCO-sb were observed in the hydrogen group compared with those in the control group. Changes in other pulmonary function parameters, including FVC, FEV1, FEV1/FVC%, and TLC, were not significantly different between the two groups. Adverse events were reported in 7 (15.9%) patients in the HW group and 10 (23.3%) patients in the NAC group, but the difference was not significant (P=0.706).

Conclusion

Hydrogen therapy exhibits superior efficacy and acceptable safety compared with NAC therapy in patients with early-stage ILD.

Sparse Low-Rank Multi-View Subspace Clustering With Consensus Anchors and Unified Bipartite Graph

Anchor technology is popularly employed in multi-view subspace clustering (MVSC) to reduce the complexity cost. However, due to the sampling operation being performed on each individual view independently and not considering the distribution of samples in all views, the produced anchors are usually slightly distinguishable, failing to characterize the whole data. Moreover, it is necessary to fuse multiple separated graphs into one, which leads to the final clustering performance heavily subject to the fusion algorithm adopted. What is worse, existing MVSC methods generate dense bipartite graphs, where each sample is associated with all anchor candidates. We argue that this dense-connected mechanism will fail to capture the essential local structures and degrade the discrimination of samples belonging to the respective near anchor clusters. To alleviate these issues, we devise a clustering framework named SL-CAUBG. Specifically, we do not utilize sampling strategy but optimize to generate the consensus anchors within all views so as to explore the information between different views. Based on the consensus anchors, we skip the fusion stage and directly construct the unified bipartite graph across views. Most importantly, l1 norm and Laplacian-rank constraints employed on the unified bipartite graph make it capture both local and global structures simultaneously. l1 norm helps eliminate the scatters between anchors and samples by constructing sparse links and guarantees our graph to be with clear anchor-sample affinity relationship. Laplacian-rank helps extract the global characteristics by measuring the connectivity of unified bipartite graph. To deal with the nondifferentiable objective function caused by l1 norm, we adopt an iterative re-weighted method and the Newton's method. To handle the nonconvex Laplacian-rank, we equivalently transform it as a convex trace constraint. We also devise a four-step alternate method with linear complexity to solve the resultant problem. Substantial experiments show the superiority of our SL-CAUBG.

Molecular docking technology and network pharmacology based on Rhapontici Radix-Cremastrae Pseudobulbus drug pair in treating breast cancer

OBJECTIVE

Network pharmacology is a bioinformatics-based research strategy for identifying the mechanisms of drugs and promoting drug development. This study used network pharmacology to investigate the mechanism of the Loulu-Cremastrae Pseudobulbus drug pair treating breast cancer (BC).

MATERIALS AND METHODS

The ingredients and potential targets of the drug pair were searched with Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCSMP). National Center for Biotechnology Information (NCBI) and gene cards were used to search the targets of BC. Networks of "drugs-components-targets" and protein-protein interaction were constructed through Cytoscape. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were carried out through common targets. Using AutoDock tool, molecular docking was performed to verify the binding between key targets and compounds.

RESULTS

Finally, we selected 6 active compounds from the drug pair. A total of 61 targets were associated with the drug pair, and 15,295 targets were related to BC. 55 common targets were obtained after the intersection. The key targets included Transcription factor Jun (JUN), Heat shock protein HSP 90-alpha (HSP90AA1), and Caspase-3(CASP3). 327 terms were obtained by GO analysis. 78 pathways (p < 0.05) were identified through KEGG analysis. Molecular docking indicated that important compounds combined well with key targets.

CONCLUSIONS

Various active compounds, including beta-sitosterol, 2-methoxy-9,10-dihydrophenanthrene-4,5-diol, and stigmasterol, can regulate multiple signaling pathways related to BC, such as the estrogen and prolactin signaling pathways, playing therapeutic roles in BC.

Peripheral T-Cell Priming and Micrometastatic Disease Control with Metastasis-Directed Therapy: Multidimensional Immunogenomic Profiling of Oligometastatic Prostate Cancer in the EXTEND Trial

PURPOSE/OBJECTIVE(S)

Comprehensive metastasis-directed therapy (MDT) for oligometastatic prostate cancer extended progression-free survival (PFS) and time to new lesion formation in the intermittent hormone therapy (HT) basket of EXTEND. To better understand the mechanism of MDT benefit, we pooled the intermittent and continuous HT baskets of EXTEND and tested the hypothesis that adding MDT to HT would program systemic T-cells to control micrometastatic disease.

MATERIALS/METHODS

A total of 174 men were randomized to HT with or without MDT to up to 5 sites of metastases. HT was given for 6 months (intermittent basket, n = 87) or indefinitely (continuous basket, n = 87). Peripheral blood samples were drawn at enrollment, at the end of MDT, at 3 months follow-up (3 mo F/U), and at progression and then analyzed by flow cytometry, T-cell receptor (TCR)-β CDR3 variable region sequencing, multiplex cytokine profiling, and next-generation circulating tumor DNA (ctDNA) sequencing. TCR clonal expansion was determined using a published betabinomial model. Repertoire changes were assessed by Morisita's index, and dominant TCR repertoire motifs were characterized with ImmunoMap. Associations between blood markers and PFS were evaluated with Cox regression adjusted hazard ratios (aHR) accounting for randomization arm and stratifying for intermittent vs continuous HT.

RESULTS

Randomization to MDT+HT was associated with T-cell activation, proliferation, and clonal expansion. This response was first observed at end-MDT as upregulated expression of T-cell activation and inhibition markers (i.e., ICOS, Tim-3, and LAG-3) and increases in highly proliferative CD4+ and CD8+ Ki67hi T-cells (all P<0.05). TCR sequencing of 7,678,911 T-cells revealed that MDT+HT was associated with TCR clonal expansion, remodeling of the TCR repertoire, and changes in dominant TCR motifs at end-MDT and 3 mo F/U (all P<0.05). Observed T-cell priming could be driven by signaling networks of canonical T-cell stimulatory cytokines (IL-2, IL-12, and IL-15), which were upregulated at end-MDT and persisted at 3 mo F/U (all P<0.05). This modulation of T-cell phenotype, clonotype, and cytokine concentrations was not observed in the HT-monotherapy arm. At end-MDT, systemic T-cell responses were associated with improved PFS, most notably CD8+ T-cell expression of LAG-3 (aHR 0.22, 95% CI 0.03-0.91) and high TCR clonal expansion (aHR 0.13, 95% CI 0.02-0.52). High ctDNA burden at end-MDT correlated with worse PFS (aHR 1.41, 95% CI 1.04-2.54), as did CD8+ T-cell expression of inhibitory receptor TIGIT at 3 mo F/U (aHR 1.03, 95% CI 1.01-1.06).

CONCLUSION

The addition of MDT to HT induced systemic T-cell activation and expansion, which was not observed in the HT-only arm. This systemic immune response was independently associated with improved PFS. In addition to cytoreduction of macroscopic disease, MDT-induced immune education may be an important complementary mechanism of micrometastatic control in oligometastatic prostate cancer.

Addition of Metastasis-Directed Therapy to Standard of Care Systemic Therapy for Oligometastatic Breast Cancer (EXTEND): A Multicenter, Randomized Phase II Trial

PURPOSE/OBJECTIVE(S)

Prior retrospective and prospective evidence have suggested a potential survival benefit of adding metastasis-directed therapy (MDT) to standard of care systemic therapy for oligometastatic breast cancer. This has led to the increased utilization of MDT in this setting despite the lack of randomized evidence to support this approach. Furthermore, the recent presentation of NRG-BR002 has questioned the value of MDT. Thus, we evaluated whether the addition of MDT to systemic therapy improves PFS in oligometastatic breast cancer.

MATERIALS/METHODS

EXTEND (NCT03599765) is a phase II randomized basket trial for multiple solid tumors testing whether the addition of MDT improves PFS. The primary endpoint was pre-specified to be independently assessed and reported for the breast basket when a minimum of 6 months of follow-up had been reached. Patients with ≤5 metastases were randomized to standard of care systemic therapy with or without MDT. The choice of systemic therapy was at the discretion of the treating medical oncologist. Number of metastatic lesions and prior lines of systemic therapy for metastatic disease were used as stratification variables pre-randomization. The primary endpoint was progression-free survival (PFS) defined as time to randomization to date of clinical or radiographic progression or death. The study was designed to have 80% power to detect an improvement in median PFS from 18 to 36 months, with a type I error of 0.1.

RESULTS

Between September 2018 to July 2022, 43 patients were randomized. 22 patients were assigned to the MDT arm, and 21 patients to the no MDT arm. Three patients were not evaluable. The MDT arm patients were older vs the no-MDT arm patients (median 61.5 years vs 48 years, p = 0.01). Otherwise, the arms were well-balanced. Overall, 8 patients had triple negative disease (18.6%), and 12 patients (30%) had de novo metastatic disease. Of those patients with de novo presentation randomized to MDT, all except one had the primary tumor treated with surgery and radiation. At a median follow-up of 19.4 months, 20 events were observed. Among the 40 evaluable patients, there were 5 deaths (3 in the MDT arm and 2 in the no MDT arm). There was no difference in PFS between the MDT and no MDT arms (median 15.6 v 24.9 months, p = 0.66). Similarly, there was no difference in the secondary endpoint of time to new metastatic lesion appearance between the MDT and no MDT arms (median 15.6 months vs not reached, p = 0.09). Two grade 3 toxicities were observed in the MDT arm, and 1 in the no MDT arm. Further analysis of correlative translational biomarkers, including immune markers and ctDNA, are ongoing.

CONCLUSION

The addition of MDT to standard of care systemic therapy did not improve PFS or time to new metastatic lesion in patients with oligometastatic breast cancer. This data coupled with the recently presented NRG-BR002 results, suggests there is no benefit to MDT in an otherwise unselected oligometastatic breast cancer population.

Validation of the Prognostic Index for Spine Metastasis (PRISM) Score for Stratifying Survival in Patients Treated with Spinal Stereotactic Radiosurgery

PURPOSE/OBJECTIVE(S)

Stereotactic spinal radiosurgery (SSRS) has been increasingly utilized as a first-line treatment for the management of spine metastases due to its ability to prolong survival and improve symptom control. Studies have shown that SSRS is helpful for select patients; however, there is no universal scoring system utilized to predict patient response to treatment. The Prognostic Index for Spinal Metastases (PRISM) score was shown to predict the likelihood of patients benefiting from SSRS. We sought to further demonstrate its generalizability by performing validation with a large dataset from a second high-volume institution.

MATERIALS/METHODS

We performed a retrospective review from 2017-2019 of 424 patients treated with SSRS at a single institution. Patients were stratified on the previously described PRISM criteria: Female sex (+2), solitary bone disease (+3), performance status (0 through +3.5), prior surgery at the SSRS site (+1), number of other metastatic sites (-N), prior radiation at the SSRS site (-1), and latency to treatment ≥ 5 months (+3). Patients were grouped based on PRISM scores: >7, Group 1; 4-7, Group 2; 1-3, Group 3; <1 Group 4. There were 89, 188, 88, and 59 patients in Groups 1, 2, 3, and 4, respectively. Most patients were male (70%) with a performance status of 0 (53%). The most common tumor histologies were prostate (34%), renal (18%), and lung (11%). The median biological effective dose (BED10) was 60 Gy (interquartile range [IQR], 60-82). We performed Cox proportional hazards analysis on overall survival (OS) based on PRISM score and patient and tumor characteristics. Concordance indices created from PRISM criteria and the multivariate Cox proportional analysis were compared.

RESULTS

The median follow-up time was 50.5 months (95% confidence interval [CI], 45.8-54.7) with a median overall survival of 30.3 months (95% CI, 27.3-38.4). The median overall survivals for PRISM Groups 1, 2, 3, and 4 were 57.1, 37, 23.7, and 8.8 months, respectively. There were significant differences in overall survival among PRISM groups with hazard ratios of 0.49 (95% CI, 0.35-0.69; P<0.001) for Group 1, 0.71 (95% CI, 0.55-0.91); P<0.007) for Group 2, 1,45 (95% CI, 1.08-1.94); P = 0.010) for Group 3, and 3.47 (95% CI, 2.56-4.70; P<0.001) for Group 4. Multivariable Cox analysis for patient and tumor characteristics revealed only the number of organs involved and performance status as significant clinicopathologic prognostic attributes. However, the C-index using the PRISM criteria was 0.76, which was superior to the C-index when using the significant clinicopathologic attributes by themselves (0.71).

CONCLUSION

These data demonstrate robust validation of the PRISM score to stratify OS in patients treated with SSRS and may help guide optimal treatment selection in prospective trials and clinical settings.

Real-world implementation of non-endoscopic triage testing for Barrett's oesophagus during COVID-19

BACKGROUND

The Coronavirus pandemic (COVID-19) curtailed endoscopy services, adding to diagnostic backlogs. Building on trial evidence for a non-endoscopic oesophageal cell collection device coupled with biomarkers (Cytosponge), an implementation pilot was launched for patients on waiting lists for reflux and Barrett's oesophagus surveillance.

AIMS

(i) To review reflux referral patterns and Barrett's surveillance practices. (ii) To evaluate the range of Cytosponge findings and impact on endoscopy services.

DESIGN AND METHODS

Cytosponge data from centralized laboratory processing (trefoil factor 3 (TFF3) for intestinal metaplasia (IM), haematoxylin & eosin for cellular atypia and p53 for dysplasia) over a 2-year period were included.

RESULTS

A total of 10 577 procedures were performed in 61 hospitals in England and Scotland, of which 92.5% (N = 9784/10 577) were sufficient for analysis. In the reflux cohort (N = 4074 with gastro-oesophageal junction sampling), 14.7% had one or more positive biomarkers (TFF3: 13.6% (N = 550/4056), p53: 0.5% (21/3974), atypia: 1.5% (N = 63/4071)), requiring endoscopy. Among samples from individuals undergoing Barrett's surveillance (N = 5710 with sufficient gland groups), TFF3-positivity increased with segment length (odds ratio = 1.37 per cm (95% confidence interval: 1.33-1.41, P < 0.001)). Some surveillance referrals (21.5%, N = 1175/5471) had ≤1 cm segment length, of which 65.9% (707/1073) were TFF3 negative. Of all surveillance procedures, 8.3% had dysplastic biomarkers (4.0% (N = 225/5630) for p53 and 7.6% (N = 430/5694) for atypia), increasing to 11.8% (N = 420/3552) in TFF3+ cases with confirmed IM and 19.7% (N = 58/294) in ultra-long segments.

CONCLUSIONS

Cytosponge-biomarker tests enabled targeting of endoscopy services to higher-risk individuals, whereas those with TFF3 negative ultra-short segments could be reconsidered regarding their Barrett's oesophagus status and surveillance requirements. Long-term follow-up will be important in these cohorts.

Multi-view subspace clustering via adaptive graph learning and late fusion alignment

Multi-view subspace clustering has attracted great attention due to its ability to explore data structure by utilizing complementary information from different views. Most of existing methods learn a sample representation coefficient matrix or an affinity graph for each single view, then the final clustering result is obtained from the spectral embedding of a consensus graph using certain traditional clustering techniques, such as k-means. However, clustering performance will be degenerated if the early fusion of partitions cannot fully exploit relationships between all samples. Different from existing methods, we propose a multi-view subspace clustering method via adaptive graph learning and late fusion alignment (AGLLFA). For each view, AGLLFA learns an affinity graph adaptively to capture the similarity relationship among samples. Moreover, a spectral embedding learning term is designed to exploit the latent feature space of different views. Furthermore, we design a late fusion alignment mechanism to generate an optimal clustering partition by fusing view-specific partitions obtained from multiple views. An alternate updating algorithm with validated convergence is developed to solve the resultant optimization problem. Extensive experiments on several benchmark datasets are conducted to illustrate the effectiveness of the proposed method when compared with other state-of-the-art methods. The demo code of this work is publicly available at https://github.com/tangchuan2000/AGLLFA.

Resonant plasmonic detection of terahertz radiation in field-effect transistors with the graphene channel and the black-As[Formula: see text]P[Formula: see text] gate layer

We propose the terahertz (THz) detectors based on field-effect transistors (FETs) with the graphene channel (GC) and the black-Arsenic (b-As) black-Phosphorus (b-P), or black-Arsenic-Phosphorus (b-As[Formula: see text]P[Formula: see text]) gate barrier layer. The operation of the GC-FET detectors is associated with the carrier heating in the GC by the THz electric field resonantly excited by incoming radiation leading to an increase in the rectified current between the channel and the gate over the b-As[Formula: see text]P[Formula: see text] energy barrier layer (BLs). The specific feature of the GC-FETs under consideration is relatively low energy BLs and the possibility to optimize the device characteristics by choosing the barriers containing a necessary number of the b-As[Formula: see text]P[Formula: see text] atomic layers and a proper gate voltage. The excitation of the plasma oscillations in the GC-FETs leads to the resonant reinforcement of the carrier heating and the enhancement of the detector responsivity. The room temperature responsivity can exceed the values of [Formula: see text] A/W. The speed of the GC-FET detector's response to the modulated THz radiation is determined by the processes of carrier heating. As shown, the modulation frequency can be in the range of several GHz at room temperatures.

Re-recognition of innate immune memory as an integrated multidimensional concept

In the past decade, the concept of immunological memory, which has long been considered a phenomenon observed in the adaptive immunity of vertebrates, has been extended to the innate immune system of various organisms. This de novo immunological memory is mainly called "innate immune memory", "immune priming", or "trained immunity" and has received increased attention because of its potential for clinical and agricultural applications. However, research on different species, especially invertebrates and vertebrates, has caused controversy regarding this concept. Here we discuss the current studies focusing on this immunological memory and summarize several mechanisms underlying it. We propose "innate immune memory" as a multidimensional concept as an integration between the seemingly different immunological phenomena.

Mulberry-like porous-hollow AuPtAg nanorods for electrochemical immunosensing of biomarker myoglobin

Mulberry-like AuPtAg porous hollow nanorods (PHNR) were facilely synthesized for the first time via a wet chemical method, where Au nanorods (Au NR) behaved as sacrificed template. The anisotropic oriented growth and etching process are involved in this synthesis. Their structural and electronic characteristics were scrutinously examined by TEM, EDS, XPS, and electrochemical techniques. The AuPtAg PHNR provided a large specific surface area and exposed a large number of active sites, showing highly enhanced catalytic activity. On this foundation, a label-free electrochemical immunosensor was developed for myoglobin (Myo) assay based on the AuPtAg PHNR. Further, the built sensor exhibited fast and ultrasensitive responses in a linear range of 0.0001 ~ 1000 ng mL^-1 with a low limit of detection (LOD = 0.46 pg mL^-1, S/N = 3), and enabled efficient application to human serum samples with acceptable results. Consequently, the developed AuPtAg PHNR-based platform has a broad prospect in practically monitoring Myo and other biomarkers in clinics.

Bonding performance of experimental HEMA-free two-step universal adhesives to low C-factor flat dentin

OBJECTIVES

Experimental two-step universal adhesives (2-UAs) providing a particle-filled hydrophobic adhesive resin with a significant film thickness to hydrophobically seal the adhesive interface were designed and synthesized. This study aimed to characterize their interfacial interaction with dentin, to determine whether the 2-UA formulations achieve durable bonding to low C-factor flat dentin and to measure their water sorption.

METHODS

Bonding effectiveness of 2-UAs that combine a 10-MDP-based primer with hydrophobic adhesive resins differing only for filler (BZF-21, BZF-29, and BZF-29_hv) were comparatively investigated with the commercial adhesive Clearfil SE Bond 2 (C-SE2, Kuraray Noritake). Adhesive-dentin interfaces were characterized with TEM. Adhesive-resin disks were immersed in distilled water at 37 °C for 1 week, 6 months and 1 year to measure water sorption and solubility. 'Immediate' and 'aged' micro-tensile bond strength (μTBS) of the adhesives applied in etch-and-rinse (E&R) and self-etch (SE) bonding mode to low C-factor flat dentin were measured. Statistical analyses involved linear mixed-effects (LME) modelling and Kruskal-Wallis testing (p < 0.05).

RESULTS

TEM revealed that E&R hybrid layers were more sensitive to aging than SE hybrid layers. Lower water sorption was recorded for all UAs compared with C-SE2. The immediate μTBS of BZF-21 and BZF-29 was not significantly different from that of C-SE2. The 1-year aged μTBS of all 2-UAs was significantly lower than that of C-SE2, except for BZF-29 applied in E&R mode. A significant reduction in μTBS upon 1-year aging was recorded for BZF-21 and BZF-29 applied in E&R mode. A significant difference in μTBS between E&R and SE bonding modes was recorded for all adhesives except BZF-21.

SIGNIFICANCE

Experimental 2-UAs with a hydrophobic adhesive-resin design produced± 20-μm thick adhesive-resin layers, absorbed less water and resulted in bonding performance that was more aging-resistant when applied in SE than in E&R bonding mode. The silica-filled BZF-29 2-UA revealed the most comparable bonding performance with C-SE2 in a low C-factor condition (flat dentin).

Efficient and Effective One-Step Multiview Clustering

Multiview clustering algorithms have attracted intensive attention and achieved superior performance in various fields recently. Despite the great success of multiview clustering methods in realistic applications, we observe that most of them are difficult to apply to large-scale datasets due to their cubic complexity. Moreover, they usually use a two-stage scheme to obtain the discrete clustering labels, which inevitably causes a suboptimal solution. In light of this, an efficient and effective one-step multiview clustering (E ² OMVC) method is proposed to directly obtain clustering indicators with a small-time burden. Specifically, according to the anchor graphs, the smaller similarity graph of each view is constructed, from which the low-dimensional latent features are generated to form the latent partition representation. By introducing a label discretization mechanism, the binary indicator matrix can be directly obtained from the unified partition representation which is formed by fusing all latent partition representations from different views. In addition, by coupling the fusion of all latent information and the clustering task into a joint framework, the two processes can help each other and obtain a better clustering result. Extensive experimental results demonstrate that the proposed method can achieve comparable or better performance than the state-of-the-art methods. The demo code of this work is publicly available at https://github.com/WangJun2023/EEOMVC.

Fluorescent folic acid-capped copper nanoclusters for the determination of rifampicin based on inner filter effect

Development of excellent sensors to determine trace concentrations of rifampicin is of intense importance for medicine analysis and human health. Herein, a facile and green fluorescent probe was established for the determination of rifampicin by using folic acid protected copper nanoclusters (FA-Cu NCs). Many characterization methods were applied for the analysis of the as-prepared FA-Cu NCs including UV-visible absorption spectra, fluorescence spectra, Fourier-transform infrared spectroscopy (FT-IR), transmission electron microscope (TEM), fluorescence lifetime and X-ray photoelectron spectroscopy (XPS). The TEM image suggested that the as-prepared FA-Cu NCs were highly dispersed. The as-synthesized FA-Cu NCs emerged blue fluorescence under UV light and demonstrated maximum emission wavelength at 446 nm under the maximum excitation wavelength of 358 nm. After the addition of rifampicin, the FL intensities of FA-Cu NCs were uncommonly quenched. The related experimental data intimated that the quenching mechanisms were assumed to the inner filter effect (IFE) and static quenching. The as-proposed probe platform displayed an obvious linear relationship with rifampicin concentrations varying from 0.5 to 100 µM, and the corresponding detection limit (LOD) was 0.073 µM (S/N = 3). Finally, the as-established detection platform was successfully employed to analyze trace concentrations of rifampicin in real samples.

Adaptive Discriminative Regions Learning Network for Remote Sensing Scene Classification

As an auxiliary means of remote sensing (RS) intelligent interpretation, remote sensing scene classification (RSSC) attracts considerable attention and its performance has been improved significantly by the popular deep convolutional neural networks (DCNNs). However, there are still several challenges that hinder the practical applications of RSSC, such as complex composition of land cover, scale-variation of objects, and redundant and noisy areas for scene classification. In order to mitigate the impact of these issues, we propose an adaptive discriminative regions learning network for RSSC, referred as ADRL-Net briefly, which locates discriminative regions effectively for boosting the performance of RSSC by utilizing a novel self-supervision mechanism. Our proposed ADRL-Net consists of three main modules, including a discriminative region generator, a region discriminator, and a region scorer. Specifically, the discriminative region generator first generates some candidate regions which could be informative for RSSC. Then, the region discriminator evaluates the regions generated by region generator and provides feedback for the generator to update the informative regions. Finally, the region scorer makes prediction scores for the whole image by using the discriminative regions. In such a manner, the three modules of ADRL-Net can cooperate with each other and focus on the most informative regions of an image and reduce the interference of redundant regions for final classification, which is robust to the complex scene composition, object scales, and irrelevant information. In order to validate the efficacy of the proposed network, we conduct experiments on four widely used benchmark datasets, and the experimental results demonstrate that ADRL-Net consistently outperforms other state-of-the-art RSSC methods.

How does public environmental supervision affect the industrial structure optimization?

Based on the panel data of 278 prefecture-level cities in China from 2005 to 2017, this study takes the environmental information disclosure policy implemented in 2008 as a quasi-natural experiment and uses the difference-in-difference method to examine the impact of public environmental supervision on industrial structure upgrading. This study found that public environmental supervision has indeed significantly promoted the industrial structure optimization. It is noteworthy that, the impact of public environmental supervision on industrial structure optimization is heterogeneous across regions and corresponding to the intensity of environmental regulations. Public environmental supervision in the eastern region has significantly promoted the optimization of the industrial structure, while the promotion effect in the central and western regions is not obvious. Besides, the effect of public environmental supervision is more significant in the cities with relatively tight traditional environmental regulations. The mechanism analysis shows that public environmental supervision promotes industrial structure upgrading mainly by influencing the level of urban technological innovation, and the level of technological innovation plays an intermediary role between public environmental supervision and industrial structure optimization. This study has important reference significance for further improving the environmental information disclosure system and constructing an environmental governance system with the government as the leading body, enterprises as the main body, and the public as the common participation.

Study on Effects of Refining Slag on Properties and Hydration of Cemented Solid Waste-Based Backfill

This study used refining slag (RS), ground granulated blast furnace slag (GGBS), steel slag (SS), and desulfurized gypsum (DG) to prepare a mine-filling cementitious material. The developed cementitious material and tailings sand were mixed to prepare a novel mine backfill material with better performance and a lower cost. The macroscopic properties and hydration mechanism of the cemented solid waste-based backfill were investigated when RS content was 0, 5%, 10%, 15%, 20%, 30% and 40%. The results showed that introducing RS could reduce the bleeding rate and shorten the setting time of backfill slurry while significantly enhancing the 3-day compressive strength of backfill. Compared to JL-0, the bleeding rate decreased by 50.3% as the RS content was raised to 15%, while the setting time was shortened by 36.5%, and the 3-day compressive strength increased by 4.3 times. As the RS content did not exceed 20%, the 28-day compressive strength of the backfill was not lower than that of the cement backfill (4.3 MPa). The results of microanalysis (including XRD, FT-IR, SEM, TG-DSC, and heat of hydration) revealed that the hydration products of the RS-GGBS-SS-DG quaternary material are primarily C-(A)-S-H gels and AFt. The main effect of RS is to improve the content of aluminates, accelerating and increasing the production of AFt, thus leading to faster overall hydration. This research can provide data support for the application of RS in the mine-filling field. Applying quaternary solid waste-based cementitious materials in the mine-filling field has good economic benefits.

Transcriptome features of innate immune memory in Drosophila

Immune memory is the ability of organisms to elicit potentiated immune responses at secondary infection. Current studies have revealed that similar to adaptive immunity, innate immunity exhibits memory characteristics (called "innate immune memory"). Although epigenetic reprogramming plays an important role in innate immune memory, the underlying mechanisms have not been elucidated, especially at the individual level. Here, we established experimental systems for detecting innate immune memory in Drosophila melanogaster. Training infection with low-pathogenic bacteria enhanced the survival rate of the flies at subsequent challenge infection with high-pathogenic bacteria. Among low-pathogenic bacteria, Micrococcus luteus (Ml) and Salmonella typhimurium (St) exerted apparent training effects in the fly but exhibited different mechanisms of action. Ml exerted training effects even after its clearance from flies, while live St persisted in the flies for a prolonged duration. RNA sequencing (RNA-Seq) analysis revealed that Ml training enhanced the expression of the immune-related genes under the challenge condition but not under the non-challenge condition. In contrast, St training upregulated the expression of the immune-related genes independent of challenge. These results suggest that training effects with Ml and St are due to memory and persistence of immune responses, respectively. Furthermore, we searched for the gene involved in immune memory, and identified a candidate gene, Ada2b, which encodes a component of the histone modification complex. The Ada2b mutant suppressed Ml training effects on survival and disrupted the expression of some genes under the training + challenge condition. These results suggest that the gene expression regulated by Ada2b may contribute to innate immune memory in Drosophila.

White blood cell classification via a discriminative region detection assisted feature aggregation network

White blood cell (WBC) classification plays an important role in human pathological diagnosis since WBCs will show different appearance when they fight with various disease pathogens. Although many previous white blood cell classification have been proposed and earned great success, their classification accuracy is still significantly affected by some practical issues such as uneven staining, boundary blur and nuclear intra-class variability. In this paper, we propose a deep neural network for WBC classification via discriminative region detection assisted feature aggregation (DRFA-Net), which can accurately locate the WBC area to boost final classification performance. Specifically, DRFA-Net uses an adaptive feature enhancement module to refine multi-level deep features in a bilateral manner for efficiently capturing both high-level semantic information and low-level details of WBC images. Considering the fact that background areas could inevitably produce interference, we design a network branch to detect the WBC area with the supervision of segmented ground truth. The bilaterally refined features obtained from two directions are finally aggregated for final classification, and the detected WBC area is utilized to highlight the features of discriminative regions by an attention mechanism. Extensive experiments on several public datasets are conducted to validate that our proposed DRFA-Net can obtain higher accuracies when compared with other state-of-the-art WBC classification methods.

Region-Aware Hierarchical Latent Feature Representation Learning-Guided Clustering for Hyperspectral Band Selection

Hyperspectral band selection aims to identify an optimal subset of bands for hyperspectral images (HSIs). For most existing clustering-based band selection methods, they directly stretch each band into a single feature vector and employ the pixelwise features to address band redundancy. In this way, they do not take full consideration of the spatial information and deal with the importance of different regions in HSIs, which leads to a nonoptimal selection. To address these issues, a region-aware hierarchical latent feature representation learning-guided clustering (HLFC) method is proposed. Specifically, in order to fully preserve the spatial information of HSIs, the superpixel segmentation algorithm is adopted to segment HSIs into multiple regions first. For each segmented region, the similarity graph is constructed to reflect the bands-wise similarity, and its corresponding Laplacian matrix is generated for learning low-dimensional latent features in a hierarchical way. All latent features are then fused to form a unified feature representation of HSIs. Finally, k -means clustering is utilized on the unified feature representation matrix to generate multiple clusters from which the band with maximum information entropy is selected to form the final subset of bands. Extensive experimental results demonstrate that the proposed clustering method can achieve superior performance than the state-of-the-art representative methods on the band selection. The demo code of this work is publicly available at https://github.com/WangJun2023/HLFC.

Genetic dissection of innate immune memory in Drosophila melanogaster

Current studies have demonstrated that innate immunity possesses memory characteristics. Although the molecular mechanisms underlying innate immune memory have been addressed by numerous studies, genetic variations in innate immune memory and the associated genes remain unclear. Here, we explored innate immune memory in 163 lines of Drosophila melanogaster from the Drosophila Synthetic Population Resource. In our assay system, prior training with low pathogenic bacteria (Micrococcus luteus) increased the survival rate of flies after subsequent challenge with highly pathogenic bacteria (Staphylococcus aureus). This positive training effect was observed in most lines, but some lines exhibited negative training effects. Survival rates under training and control conditions were poorly correlated, suggesting that distinct genetic factors regulate training effects and normal immune responses. Subsequent quantitative trait loci analysis suggested that four loci containing 80 genes may be involved in regulating innate immune memory. Among them, Adgf-A, which encodes an extracellular adenosine deaminase-related growth factor, was shown to be associated with training effects. Our study findings help to elucidate the genetic architecture of innate immune memory in Drosophila and may provide insight for new therapeutic treatments aimed at boosting immunity.

Processed electroencephalography: impact of patient age and surgical position on intraoperative processed electroencephalogram monitoring of burst-suppression

We previously reported that processed EEG underestimated the amount of burst suppression compared to off-line visual analysis. We performed a follow-up study to evaluate the reasons for the discordance. Forty-five patients were monitored intraoperatively with processed EEG. A computer algorithm was used to convert the SedLine® (machine)-generated burst suppression ratio into a raw duration of burst suppression. The reference standard was a precise off-line measurement by two neurologists. We measured other potential variables that may affect machine accuracy such as age, surgery position, and EEG artifacts. Overall, the median duration of bust suppression for all study subjects was 15.4 min (Inter-quartile Range [IQR] = 1.0-20.1) for the machine vs. 16.1 min (IQR = 0.3-19.7) for the neurologists' assessment; the 95% limits of agreement fall within - 4.86 to 5.04 s for individual 30-s epochs. EEG artifacts did not affect the concordance between the two methods. For patients in prone surgical position, the machine estimates had significantly lower overall sensitivity (0.86 vs. 0.97; p = 0.038) and significantly wider limits of agreement ([- 4.24, 3.82] seconds vs. [- 1.36, 1.13] seconds, p = 0.001) than patients in supine position. Machine readings for younger patients (age < 65 years) had higher sensitivity (0.96 vs 0.92; p = 0.021) and specificity (0.99 vs 0.88; p = 0.007) for older patients. The duration of burst suppression estimated by the machine generally had good agreement compared with neurologists' estimation using a more precise off-line measurement. Factors that affected the concordance included patient age and position during surgery, but not EEG artifacts.

Dupilumab combined with low-dose systemic steroid therapy improves efficacy and safety for bullous pemphigoid

BACKGROUND

Routine systemic therapy for bullous pemphigoid (BP) has been challenged due to the inevitably adverse effects. According to the successful applications of dupilumab in BP cases reported, therefore, we investigate the real-life efficacy and safety of dupilumab combined with low-dose oral steroid for BP.

RESEARCH DESIGN AND METHODS

A cohort of BP patients who received either dupilumab plus low-dose methylprednisolone (dupilumab group) or merely methylprednisolone (control group) was retrospectively reviewed. The time to disease control was investigated. Additionally, the control dose and cumulative dosage of steroids, Bullous Pemphigoid Disease Area Index(BPDAI) scores, pruritus scores, and adverse events were assessed.

RESULTS

A total of 40 patients, with 20 in each group, were retrospectively studied. The time to disease control was shorter in the dupilumab group than the control group (14 days vs. 19 days, p=0.043). When the disease was controlled, the control dose and cumulative dosage of methylprednisolone in the dupilumab group were substantially lower than those of the control (24.6mg vs. 48.8mg, 376.8mg vs. 985.6mg, both p<0.01). Compared with the control, the percentage change from baseline in BPDAI scores and pruritus scores were both significantly reduced, and the adverse events were also less frequent in the dupilumab group.

CONCLUSIONS

The combination therapy of dupilumab plus low-dose methylprednisolone exhibits superior efficacy and safety in comparison with the current first-line systemic therapy for BP. This article is protected by copyright. All rights reserved.

Graph regularized spatial-spectral subspace clustering for hyperspectral band selection

Hyperspectral band selection, which aims to select a small number of bands to reduce data redundancy and noisy bands, has attracted widespread attention in recent years. Many effective clustering-based band selection methods have been proposed to accomplish the band selection task and have achieved satisfying performance. However, most of the previous methods reshape the original hyperspectral images (HSIs) into a set of stretched band vectors, which ignore the spatial information of HSIs and the difference between diverse regions. To address these issues, a graph regularized spatial-spectral subspace clustering method for hyperspectral band selection is proposed in this paper, referred to as GRSC. Specifically, the proposed method adopts superpixel segmentation to preserve the spatial information of HSIs by segmenting their first principal component into diverse homogeneous regions. Then the discriminative latent features are generated from each segmented region to represent the whole band, which can mitigate the effect of noise on the band selection. Finally, a self-representation subspace clustering model and an l_2,1-norm regularization are utilized to explore the spectral correlation among all bands. In addition, a similarity graph between region-aware latent features is adaptively learned to preserve the spatial structure of HSIs in the latent representation space. Extensive classification experimental results on three public datasets verify the effectiveness of GRSC over several state-of-the-art methods. The demo code of this work is publicly available at https://github.com/WangJun2023/GRSC.

High-Order Correlation Preserved Incomplete Multi-View Subspace Clustering

Incomplete multi-view clustering aims to exploit the information of multiple incomplete views to partition data into their clusters. Existing methods only utilize the pair-wise sample correlation and pair-wise view correlation to improve the clustering performance but neglect the high-order correlation of samples and that of views. To address this issue, we propose a high-order correlation preserved incomplete multi-view subspace clustering (HCP-IMSC) method which effectively recovers the missing views of samples and the subspace structure of incomplete multi-view data. Specifically, multiple affinity matrices constructed from the incomplete multi-view data are treated as a third-order low rank tensor with a tensor factorization regularization which preserves the high-order view correlation and sample correlation. Then, a unified affinity matrix can be obtained by fusing the view-specific affinity matrices in a self-weighted manner. A hypergraph is further constructed from the unified affinity matrix to preserve the high-order geometrical structure of the data with incomplete views. Then, the samples with missing views are restricted to be reconstructed by their neighbor samples under the hypergraph-induced hyper-Laplacian regularization. Furthermore, the learning of view-specific affinity matrices as well as the unified one, tensor factorization, and hyper-Laplacian regularization are integrated into a unified optimization framework. An iterative algorithm is designed to solve the resultant model. Experimental results on various benchmark datasets indicate the superiority of the proposed method. The code is implemented by using MATLAB R2018a and MindSpore library: https://github.com/ChangTang/HCP-IMSC.

DeFusionNET: Defocus Blur Detection via Recurrently Fusing and Refining Discriminative Multi-Scale Deep Features

Albeit great success has been achieved in image defocus blur detection, there are still several unsolved challenges, e.g., interference of background clutter, scale sensitivity and missing boundary details of blur regions. To deal with these issues, we propose a deep neural network which recurrently fuses and refines multi-scale deep features (DeFusionNet) for defocus blur detection. We first fuse the features from different layers of FCN as shallow features and semantic features, respectively. Then, the fused shallow features are propagated to deep layers for refining the details of detected defocus blur regions, and the fused semantic features are propagated to shallow layers to assist in better locating blur regions. The fusion and refinement are carried out recurrently. In order to narrow the gap between low-level and high-level features, we embed a feature adaptation module before feature propagating to exploit the complementary information as well as reduce the contradictory response of different feature layers. Since different feature channels are with different extents of discrimination for detecting blur regions, we design a channel attention module to select discriminative features for feature refinement. Finally, the output of each layer at last recurrent step are fused to obtain the final result. We collect a new dataset consists of various challenging images and their pixel-wise annotations for promoting further study. Extensive experiments on two commonly used datasets and our newly collected one are conducted to demonstrate both the efficacy and efficiency of DeFusionNet.

Morphological classification and clinical significance of medial malleolus based on computed tomography three-dimensional reconstruction

BACKGROUND

Internal malleolus fractures and postoperative functional limitations are serious complications of deltoid ligament repair, reconstruction, while studies on conducting beak. Anatomical structure classification of medial malleolus at home and abroad is reported rarely. Hence, this morphological study is mainly designed to investigate the anatomical morphological classification and clinical significance of medial malleolus based on computed tomography (CT) three-dimensional reconstruction.

MATERIALS AND METHODS

From October 2018 to January 2021, 373 patients who underwent CT examination of malleolus medialis joint in the Jiang'an Hospital of Traditional Chinese Medicine were observed. The medial malleolus was observed and classified; then, geometric parameters were measured according to different medial malleolus types.

RESULTS

According to the results of 373 cases, medial malleolus can be divided into four types: omega type (66%), radical sign type (16%), inverted triangle type (14%), and wave type (4%).

CONCLUSIONS

There are four main shapes: omega, inverted triangle, radical sign, and wave in the medial malleolus of all normal ankles. The measurement of medial malleolus parameters according to medial malleolus in different shapes was of importance to guide smooth operation of medial malleolus fixation and deltoid ligament reconstruction and epidemiological.