Search

The motor imagery brain-computer interface (MI-BCI) based on electroencephalography (EEG) is a widely used human-machine interface paradigm. However, due to the non-stationarity and individual differences among subjects in EEG signals, the decoding accuracy is limited, affecting the application of the MI-BCI. In this paper, we propose the EISATC-Fusion model for MI EEG decoding, consisting of inception block, multi-head self-attention (MSA), temporal convolutional network (TCN), and layer fusion. Specifically, we design a DS Inception block to extract multi-scale frequency band information. And design a new cnnCosMSA module based on CNN and cos attention to solve the attention collapse and improve the interpretability of the model. The TCN module is improved by the depthwise separable convolution to reduces the parameters of the model. The layer fusion consists of feature fusion and decision fusion, fully utilizing the features output by the model and enhances the robustness of the model. We improve the two-stage training strategy for model training. Early stopping is used to prevent model overfitting, and the accuracy and loss of the validation set are used as indicators for early stopping. The proposed model achieves within-subject classification accuracies of 84.57% and 87.58% on BCI Competition IV Datasets 2a and 2b, respectively. And the model achieves cross-subject classification accuracies of 67.42% and 71.23% (by transfer learning) when training the model with two sessions and one session of Dataset 2a, respectively. The interpretability of the model is demonstrated through weight visualization method.

Aiming at the problem that more and more kinds of radar compound jamming signals and too few training samples were difficult to make the deep learning model reach the optimal state, a semi-supervised transfer learning recognition method for radar compound jamming under small samples was proposed, which solved the problem of low network training accuracy caused by the difficulty in obtaining labeled samples through unlabeled samples.The feature extractor and classifier obtained after pre-training of single jamming data set were transferred to small-scale compound jamming data set, and the model was fine-tuning by using weight imprinting and semi-supervised learning.The model parameters were optimized by the proposed nearest neighbor correlation loss nearest neighbor correlation loss (NNCL).The experimental results show that the recognition accuracy of the model can reach 93.20% when the jamming-to-noise ratio is 10 dB and there are only 5 labeled samples of the new class of compound jamming signals.

Abstract Immune checkpoints play key roles in maintaining self-tolerance. Targeted potentiation of the checkpoint molecule PD-L1 through in situ manipulation offers clinical promise for patients with autoimmune diseases. However, the therapeutic effects of these approaches are often compromised by limited specificity and inadequate expression. Here, we report a two-step dual-anchor coupling strategy for enhanced immobilization of PD-L1 on target endogenous cells by integrating bioorthogonal chemistry and physical insertion of the cell membrane. In both type 1 diabetes and rheumatoid arthritis mouse models, we demonstrate that this approach leads to elevated and sustained conjugation of PD-L1 on target cells, resulting in significant suppression of autoreactive immune cell activation, recruitment of regulatory T cells, and systematic reshaping of the immune environment. Furthermore, it restores glucose homeostasis in type 1 diabetic mice for over 100 days. This specific in situ bioengineering approach potentiates the functions of PD-L1 and represents its translational potential.

Abstract Chemotherapy remains as the first-choice treatment option for triple-negative breast cancer (TNBC). However, the limited tumor penetration and low cellular internalization efficiency of current nanocarrier-based systems impede the access of anticancer drugs to TNBC with dense stroma and thereby greatly restricts clinical therapeutic efficacy, especially for TNBC bone metastasis. In this work, biomimetic head/hollow tail nanorobots were designed through a site-selective superassembly strategy. We show that nanorobots enable efficient remodeling of the dense tumor stromal microenvironments (TSM) for deep tumor penetration. Furthermore, the self-movement ability and spiky head markedly promote interfacial cellular uptake efficacy, transvascular extravasation, and intratumoral penetration. These nanorobots, which integrate deep tumor penetration, active cellular internalization, near-infrared (NIR) light-responsive release, and photothermal therapy capacities into a single nanodevice efficiently suppress tumor growth in a bone metastasis female mouse model of TNBC and also demonstrate potent antitumor efficacy in three different subcutaneous tumor models.

Abstract Boron neutron capture therapy (BNCT) was clinically approved in 2020 and exhibits remarkable tumour rejection in preclinical and clinical studies. It is binary radiotherapy that may selectively deposit two deadly high-energy particles (4He and 7Li) within a cancer cell. As a radiotherapy induced by localized nuclear reaction, few studies have reported its abscopal anti-tumour effect, which has limited its further clinical applications. Here, we engineer a neutron-activated boron capsule that synergizes BNCT and controlled immune adjuvants release to provoke a potent anti-tumour immune response. This study demonstrates that boron neutron capture nuclear reaction forms considerable defects in boron capsule that augments the drug release. The following single-cell sequencing unveils the fact and mechanism that BNCT heats anti-tumour immunity. In female mice tumour models, BNCT and the controlled drug release triggered by localized nuclear reaction causes nearly complete regression of both primary and distant tumour grafts.

Abstract Objectives The early detection and identification of stroke are essential to the prognosis of patients with suspected stroke symptoms out-of-hospital. We aimed to develop a risk prediction model based on the FAST score to identify the different types of strokes early for emergency medical services (EMS). Methods This retrospective observational study enrolled 394 stroke patients at a single center from January 2020 to December 2021. Demographic data, clinical characteristics, and stroke risk factors with patients were collected from the EMS record database. Univariate and multivariate logistic regression analysis was used to identify the independent risk predictors. The nomogram was developed based on the independent predictors, in which the discriminative value and calibration of the nomogram were verified by the receiver operator characteristic (ROC) curve and calibration plots. Results A total of 31.90% (88/276) of patients were diagnosed with hemorrhagic stroke in the training set, while 36.40% (43/118) in the validation set. The nomogram was developed based on the multivariate analysis, including age, systolic blood pressure, hypertension, vomiting, arm weakness, and slurred speech. The area under the curve (AUC) of the ROC with nomogram was 0.796 (95% CI: 0.740–0.852, P

This paper studies the P-wave first-arrival picking based on UNet++. First, UNet++ is dimensionally reduced, and the original network is improved from the depth of the network structure and the operation of a single block. Then, the model selects a loss function and an optimizer to let the model find optimization goals and directions; then data preprocessing is performed, and the data with a signal-to-noise ratio less than 20db are screened out, and wavelet threshold denoising and normalization are performed on them; finally, training and validation are performed, and the performance on the validation set is selected. The optimal model is used as the final model. After the test of 150 test set data, it is proved that the proposed method is superior to STA/LTA and AR-AIC in the three indicators of mean, variance, and hit rate. The precision of the P-wave first arrival picked up by the method is as high as 98.00%. This work provides a new idea for automatic pickup of P-wave first arrivals.

This paper investigates the three-dimensional guidance and control problem of missile intercepting highly maneuvering target, whose acceleration information is difficult to accurately predict. With the three-dimensional guidance model for intercepting single target established by using the principle of zeroing the rate of line-of-sight (LOS), a novel intelligence guidance law has been designed through backstepping sliding mode control method, radial basis function (RBF) neural network and adaptive control technique. Then, a Lyapunov-based stability analysis demonstrates that all the signals are bounded, and the LOS rates ultimately converge to a neighborhood of the origin. Following advantages are highlighted in this paper: (i) the target information is online estimated and compensated by the RBF neural network, which indicates that the proposed guidance law is easily put into practice only relying on the position information of target. (ii) an adaptive gain term is designed in the control system, which greatly reduces the inherent chattering of sliding mode method. At last, simulations are conducted, and results illustrate the effectiveness and superiority of the designed guidance law.

In recent years, in order to achieve the goal of “carbon peaking and carbon neutralization”, many countries have focused on the development of clean energy, and the prediction of photovoltaic power generation has become a hot research topic. However, many traditional methods only use meteorological factors such as temperature and irradiance as the features of photovoltaic power generation, and they rarely consider the multi-features fusion methods for power prediction. This paper first preprocesses abnormal data points and missing values in the data from 18 power stations in Northwest China, and then carries out correlation analysis to screen out 8 meteorological features as the most relevant to power generation. Next, the historical generating power and 8 meteorological features are fused in different ways to construct three types of experimental datasets. Finally, traditional time series prediction methods, such as Recurrent Neural Network (RNN), Convolution Neural Network (CNN) combined with eXtreme Gradient Boosting (XGBoost), are applied to study the impact of different feature fusion methods on power prediction. The results show that the prediction accuracy of Long Short-Term Memory (LSTM), stacked Long Short-Term Memory (stacked LSTM), Bi-directional LSTM (Bi-LSTM), Temporal Convolutional Network (TCN), and XGBoost algorithms can be greatly improved by the method of integrating historical generation power and meteorological features. Therefore, the feature fusion based photovoltaic power prediction method proposed in this paper is of great significance to the development of the photovoltaic power generation industry.

Macroencapsulation has been widely used in cell therapy due to its capability to provide immune-privileged sites for implanted allogeneic or xenogeneic cells. Macroencapsulation also serves to provide mechanical and physiochemical support for maintaining cell expansion and promoting therapeutic functions. Macroencapsulation devices such as membrane-controlled release systems, hydrogels, microneedle (MN) array patches, and three-dimensional (3D) stents have shown promising in-lab and preclinical results in the maintenance of long-term cell survival and the strengthening of treatment efficacy. Recent studies focus on expanding the applications of these devices to new cell-based areas such as chimeric antigen receptor (CAR)-T cell delivery, cardiovascular disease therapy, and the exploration of new materials, construction methods, and working principles to augment treatment efficacy and prolong therapy duration. Here, we survey innovative platforms and approaches, as well as translation outcomes, for advancing the performance and applications of macrodevices for cell-based therapies. A discussion and critique regarding future opportunities and challenges is also provided.

The inability of new users to adapt quickly to the surface electromyography (sEMG) interface has greatly hindered the development of sEMG in the field of rehabilitation. This is due mainly to the large differences in sEMG signals produced by muscles when different people perform the same motion. To address this issue, a multi-user sEMG framework is proposed, using discriminative canonical correlation analysis and adaptive dimensionality reduction (ADR). The interface projects the feature sets for training users and new users into a low-dimensional uniform style space, overcoming the problem of individual differences in sEMG. The ADR method removes the redundant information in sEMG features and improves the accuracy of system motion recognition. The presented framework was validated on eight subjects with intact limbs, with an average recognition accuracy of 92.23% in 12 categories of upper-limb movements. In rehabilitation laboratory experiments, the average recognition rate reached 90.52%. The experimental results suggest that the framework offers a good solution to enable new rehabilitation users to adapt quickly to the sEMG interface.

Cancer metastasis is the leading cause of death in patients, here, the authors show disrupting tumor vasculature could recruit and activate anti-PD-L1 engineered platelet at metastatic tumor sites to block the PD-1/PD-L1 crosstalk and enhance the anticancer immunotherapy.

Abstract Deep learning (DL) has exhibited its exceptional performance in fields like intrusion detection. Various augmentation methods have been proposed to improve data quality and eventually to enhance the performance of DL models. However, the classic augmentation methods cannot be applied to those DL models which exploit the system-call sequences to detect intrusion. Previously, the seq2seq model has been explored to augment system-call sequences. Following this work, we propose a gated convolutional neural network (GCNN) model to thoroughly extract the potential information of augmented sequences. Also, in order to enhance the model’s robustness, we adopt adversarial training to reduce the impact of adversarial examples on the model. Adversarial examples used in adversarial training are generated by the proposed adversarial sequence generation algorithm. The experimental results on different verified models show that GCNN model can better obtain the potential information of the augmented data and achieve the best performance. Furthermore, GCNN with adversarial training can enhance robustness significantly.

Background Bronchoalveolar lavage fluid (BALF) exosomes possess different properties in different diseases, which are mediated through microRNAs (miRNAs) and long noncoding RNAs (lncRNAs), among others. By sequencing the differentially expressed lncRNAs in BALF exosomes, we seek potential targets for the diagnosis and treatment of acute lung injury (ALI). Methods Considering that human and rat genes are about 80% similar, ALI was induced using lipopolysaccharide in six male Wistar rats, with six rats as control (all weighing 200 ± 20 g and aged 6–8 weeks). BALF exosomes were obtained 24 h after ALI. The exosomes in BALF were extracted by ultracentrifugation. The differential expression of BALF exosomal lncRNAs in BALF was analyzed by RNA sequencing. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to predict the functions of differentially expressed lncRNAs, which were confirmed by reverse transcription–polymerase chain reaction. Results Compared with the control group, the ALI group displayed a higher wet/dry ratio, tumor necrosis factor-α levels, and interleukin-6 levels (all P < 0.001). The airway injection of exosomes in rats led to significant infiltration by neutrophils. A total of 2,958 differentially expressed exosomal lncRNAs were identified, including 2,524 upregulated and 434 downregulated ones. Five lncRNAs confirmed the reliability of the sequencing data. The top three GO functions were phagocytic vesicle membrane, regulation of receptor biosynthesis process, and I-SMAD binding. Salmonella infection, Toll-like receptor signaling pathway, and osteoclast differentiation were the most enriched KEGG pathways. The lncRNA–miRNA interaction network of the five confirmed lncRNAs could be predicted using miRDB. Conclusions BALF-derived exosomes play an important role in ALI development and help identify potential therapeutic targets related to ALI.

This paper is concerned with the problem of trajectory tracking control of underactuated autonomous surface vehicles subject to parameter uncertainties and nonlinear external disturbances. A robust control scheme is presented by employing backstepping method, neural network and sliding mode control. In addition, the overall signals are guaranteed the uniformly ultimate boundness by the Lyapunov stability theory. These advantages are highlighted as follows: (i) The derivations of virtual variables are obtained by a second-order filter. A compensation loop is proposed to reduce the filtered errors between the filtered variables and virtual variables. (ii) The neural network is combined with low-frequency learning techniques to estimate and approximate unknown functions of system.(iii) An anti-windup design is employed to restrict the amplitude of control inputs. Finally, simulation results show the strong robustness and tracking effectiveness of the designed control scheme under the nonlinear external disturbances.

Illeis bistigmosa (Mulsant, 1850) is a potential biological control agent of powdery mildews. We have determined the first mitochondrial genome of I. bistigmosa. The circular mitogenome of I. bistigmosa consists of 17,840 bp including 13 protein-coding genes, 22 tRNAs, 2 rRNAs, and a control region (D-loop). The base composition was AT-biased (78.44%). Maximum-likelihood phylogenetic trees strongly supported the monophyly of Coccinellinae. Illeis bistigmosa is the sister group of Halyzia sedecimguttata and Halyziini species (unclassified Halyziini), within fungivorous coccinellids. Illeis bistigmosa mitochondrial genome will be a fundamental resource for understanding the molecular phylogenetic relationships of the species-rich family Coccinellidae of Coleoptera.

The mechanical behaviors of lattice structures can be tuned by arranging or adjusting their geometric parameters. Once fabricated, the lattice’s mechanical behavior is generally fixed and cannot adapt to environmental change. In this paper, we developed a shape reconfigurable, highly stretchable lattice structure with tunable Poisson’s ratio. The lattice is built based on a hexagonal honeycomb structure. By replacing the straight beam with curled microstructure, the stretchability of the lattice is significantly improved. The Poisson’s ratio is adjusted using a geometric angle. The lattice is 3D printed using a shape memory polymer. Using its shape memory effect, the lattice demonstrates tunable shape reconfigurability as the ambient temperature changes. To capture its high stretchability, tunable Poisson’s ratio and shape reconfigurability, a phase evolution model for lattice structure is used. In the theoretical model, the effects of temperature on the material’s nonlinearity and geometric nonlinearity due to the lattice structure are assumed to be decoupled. The theoretical shape change agrees well with the Finite element results, while the theoretical model significantly reduces the computational cost. Numerical results show that the geometrical parameters and the ambient temperature can be manipulated to transform the lattice into target shapes with varying Poisson’s ratios. This work provides a design method for the 3D printed lattice structures and has potential applications in flexible electronics, soft robotics, and biomedicine.

Shape-shifting structures have gained growing interest recently and found wide applications in areas such as soft robotics, biomedical devices and self-folding origami, attributed to their ability to construct complicated shapes directly from simple structures. However, an efficient method to design and fabricate programmable 3D shape-shifting structures from 2D polymer films still lacks. In this work, we design programmable shape-shifting 3D structures via the release of internal gradient stress using the frontal photopolymerization (FPP) method. First, the relation between the non-uniformly distributed material and loading parameters, and the geometric and fabrication parameters are established theoretically. The finite element (FE) model is then developed based on the theoretically obtained material and loading parameters. Next, the elastic instability in the shape-shifting behaviors of a cured film is captured through an elastic energy minimization. Furthermore, by using grayscale light patterns, it is shown that we can selectively manipulate the geometric and fabrication parameters to improve the design freedom of various complex 3D structures.

This paper investigates the problem of path following control of the underactuated autonomous underwater vehicles in the presence of model uncertainties and external disturbances. With the three-dimensional path following error model established based on virtual guidance method, a path following robust control system is proposed using the command filtered backstepping control, neural networks, and adaptive control techniques. Then, a Lyapunov-based stability analysis demonstrates that all the signals are bounded and path following errors ultimately converge to a neighborhood of the origin. Following advantages are highlighted in this paper: 1) the derivative of virtual control is obtained via a second-order filter, which avoids explosion of complexity in the traditional backstepping design, and filters out high frequency measurement noise to keep the control system more robust, and a filtered error compensation loop is developed to guarantee the approximation precision between the virtual control signals and the filtered signals and 2) the presented controller is easily put into practice without any former knowledge of vehicle parameters and external environmental disturbances. Finally, the simulations are conducted, and results illustrate the effectiveness and good robustness of the proposed control system through a new class of flying wing autonomous underwater vehicle.

To improve the recognition rate of lower limb actions based on surface electromyography (sEMG), an effective weighted feature method is proposed, and an improved genetic algorithm support vector machine (IGA-SVM) is designed in this paper. First, for the problem of high feature redundancy and low discrimination in the surface electromyography feature extraction process, the weighted feature method is proposed based on the correlation between muscles and actions. Second, to solve the problem of the genetic algorithm selection operator easily falling into a local optimum solution, the improved genetic algorithm-support vector machine is designed by championship with sorting method. Finally, the proposed method is used to recognize six types of lower limb actions designed, and the average recognition rate reaches 94.75%. Experimental results indicate that the proposed method has definite potentiality in lower limb action recognition.

People with type 1 or advanced type 2 diabetes are highly dependent on insulin. However, insulin cannot match its function to daily varying blood glucose levels, putting people with diabetes at risk of hypoglycemia. Glucose-responsive insulin is capable of improving blood glucose manipulation and elevating the life quality of people with diabetes. Therefore, enormous endeavors have been devoted to developing glucose-responsive insulin formulations, including glucose-responsive insulin delivery systems and glucose-responsive insulin analogs. In this review, we focus on glucose-responsive insulin analogs, especially three representatives, including phenylboronic acid-mediated, glucose transporter-mediated and lectin-mediated glucose-responsive insulin analogs. Based on the published studies, the opportunities and challenges to engineering glucose-responsive insulin analogs are also discussed. Keywords: Diabetes, Insulin, Glucose-responsive, Drug delivery, Drug discovery

Abstract Background Schistosoma japonicum is one of the major causative agents of schistosomiasis. The pairing of males and females leads to female sexual maturation and maintains this mature state. However, the mechanisms by which pairing facilitates sexual maturation are yet to be investigated. Methods Parasites isolated from single- and double-sex cercariae-infected mice were analyzed by Solexa to uncover pair-regulated miRNA profiles. To reveal the biological functions of differentially expressed miRNAs among the samples, we predicted the target genes of these differentially expressed miRNAs and compared the gene expression between 23-d-old female schistosomula from double-sex infections (23DSI) and 23-d-old female schistosomula from single-sex infections (23SSI) by analyzing digital gene expression profiling (DGE). KEGG pathway analysis was used to investigate the relevant biological processes of these target genes to understand the significance of differentially expressed miRNAs after pairing. Results The differentially expressed miRNA profiles of female 18- and 23-d post-single- and double-sex infections were analysed by Solexa. Similar miRNA profiles were observed in 18SSI and 18DSI, with the presence of identically expressed high-abundance miRNA, such as miRNA-1, miRNA-71b-5p and let-7. By contrast, in 23DSI and 23SSI, most of these high-abundance miRNAs were down-regulated. Furthermore, among all samples, bantam was distinctly up-regulated in 23 DSI, and miR-1, miR-71, miR-7-5p, and miR-7 were distinctly up-regulated in 23SSI. The transcriptomes of 23DSI and 23SSI revealed that the predicted target genes of miRNA-1, miRNA-71, miRNA-7, and miR-7-5p were associated with the ribonucleoprotein complex assembly and microtubule-based process. Conversely, the predicted target genes of bantam were related to the embryo development, development of primary sexual characteristics and regulation of transcription. KEGG pathway analysis revealed that in unpaired females, the highly-expressed miRNA-1, miRNA-71, miRNA-7, and miR-7-5p only inhibited the limited pathways, such as proteasome and ribosome assembly. Meanwhile, in paired mature females, highly-expressed bantam inhibited more biological pathways, such as the citrate cycle, glycolysis, fatty acid biosynthesis and RNA degradation. Conclusions The differentially expressed miRNAs between 23SSI and 23DSI and their different functions indicated that more genes or metabolic pathways in paired mature females were inhibited than those in unpaired ones. The results suggested that after pairing, specific miRNAs regulated gene expression to lead to female sexual maturation.

Multiple preclinical evidences have supported the potential value of mesenchymal stem cells (MSCs) for treatment of acute lung injury (ALI). However, few studies focus on the dynamic tropism of MSCs in animals with acute lung injury. In this study, we track systemically transplanted human bone marrow-derived mesenchymal stem cells (hBMSCs) in NOD/SCID mice with smoke inhalation injury (SII) through bioluminescence imaging (BLI). The results showed that hBMSCs systemically delivered into healthy NOD/SCID mouse initially reside in the lungs and then partially translocate to the abdomen after 24 h. Compared with the uninjured control group treated with hBMSCs, higher numbers of hBMSCs were found in the lungs of the SII NOD/SCID mice. In both the uninjured and SII mice, the BLI signals in the lungs steadily decreased over time and disappeared by 5 days after treatment. hBMSCs significantly attenuated lung injury, elevated the levels of KGF, decreased the levels of TNF-α in BALF, and inhibited inflammatory cell infiltration in the mice with SII. In conclusion, our findings demonstrated that more systemically infused hBMSCs localized to the lungs in mice with SII. hBMSC xenografts repaired smoke inhalation-induced lung injury in mice. This repair was maybe due to the effect of anti-inflammatory and secreting KGF of hMSCs but not associated with the differentiation of the hBMSCs into alveolar epithelial cells.

In order to analyze the influence of sub-national actors on international organizations, this paper uses a questionnaire to survey the level of participation and influence of local governments. As substate actors, local governments have their responsibilities and obligations to influence state behaviour in a variety of ways. Local governments need to do the following: the influence and status of substate actors on international organizations, whether local governments are sub-state actors and their influence on state behavior, and the uncertain influence of local governments and states on international organizations. The results show that local actors play a secondary role in influencing international organizations, and that local governments are one of the local actors and have a supervisory and binding role on state behavior. The impact of local government and national uncertainty on international organizations is largely based on the outcome of local government involvement. Therefore, it is the state actors who are able to fulfil the responsibilities of international organizations and are participants in the relevant work of international organizations, in which the influence of local government participation on international organizations is more significant. [ABSTRACT FROM AUTHOR]