Feasibility and effectiveness of a foam rolling intervention in pregnant women requiring bed rest for foetal protection: a randomised controlled trial.

BACKGROUND: Bed rest during pregnancy can lead to reduced physical activity, impairing lower limb venous blood flow and increasing the risk of deep vein thrombosis (DVT) and muscle atrophy. We investigated the clinical efficacy of foam rolling intervention (FRI) in enhancing lower limb venous blood flow, mitigating the risk of DVT and muscle atrophy in pregnant women on bed rest. METHODS: This single-blind, randomised controlled trial enrolled 86 pregnant women with long-term bed rest for foetal protection (≥ 7 days), gestational age ≥ 20 weeks, and maternal age < 40 years. Participants were randomly assigned to a control or experimental group using a random number table. The control group received standard care, whereas the experimental group underwent FRI. Researchers and statisticians were aware of the treatment groups, however, the participants were blinded. Lower limb blood flow velocity, D-dimer levels, incidence of DVT, and the extent of lower limb muscle atrophy were assessed in both groups at baseline and post-intervention (day 7). To account for a 5% attrition rate and potential sampling error, the estimated sample size for each experimental and control group was 40. RESULTS: Before the intervention, no significant differences were observed between the experimental and control groups in peak blood flow, mean flow velocity, D-dimer values, or leg circumference (P > 0.05), however, the peak blood velocities of the popliteal veins were significantly higher in the control group (P = 0.031). On the seventh day post-intervention, the experimental group had significantly higher mean and peak blood velocities in femoral and popliteal veins, significantly (P < 0.05), lower mean D-dimer levels (P = 0.035), and a significantly smaller reduction in thigh and calf circumference (P < 0.001). Consequently, the rate of thigh muscle atrophy was significantly slower in the experimental group (P = 0.011). CONCLUSIONS: FRI is an effective intervention for improving lower limb venous blood flow, mitigating the risk of DVT and muscle atrophy in pregnant women on bed rest. TRIAL REGISTRATION: This trial was retrospectively registered with the Chinese Clinical Trial Registry on June 18, 2024 (registration number: ChiCTR2400085770).

Neuroprotective effect and preparation methods of berberine.

Berberine (BBR) is a natural alkaloid, which has played an important role in the field of medicine since its discovery in the late 19th century. However, the low availability of BBR in vivo prevents its full effect. In recent years, a large number of studies confirmed that BBR has a protective effect on the nervous system through various functions, yet the issue of the inability to systematically understand the protection of BBR on the nervous system remains a gap that needs to be addressed. Many existing literature introductions about berberine in neurodegenerative diseases, but the role of berberine in the nervous system goes far beyond these. Different from these literatures, this review is divided into three parts: preparation method, mechanism, and therapeutic effect. Various dosage forms of BBR and their preparation methods are added, in order to provide a reasonable choice of BBR, and help to solve the problem of low bioavailability in treatment. More importantly, we more comprehensively summarize the mechanism of BBR to protect the nervous system, in addition to the treatment of neurodegenerative diseases (anti-oxidative stress, anti-neuroinflammation, regulation of apoptosis), two extra mechanisms of berberine for the protection of the nervous system were also introduced: bidirectional regulation of autophagy and promote angiogenesis. Also, we have clarified the precise mechanism by which BBR has a therapeutic effect not only on neurodegenerative illnesses but also on multiple sclerosis, gliomas, epilepsy, and other neurological conditions. To sum up, we hope that these can evoke more efforts to comprehensively utilize of BBR nervous system, and to promote the application of BBR in nervous system protection.

Effect of Low Temperature on the Fatigue Crack Propagation Behavior of Underwater Manned Vehicle Rudder Materials in Arctic Environments.

During polar navigation, adverse environmental conditions like cold temperatures, fatigue, and corrosion can affect surface and underwater manned vehicles (UMVs). Understanding the fatigue fracture growth behavior of polar ship steel is crucial for ensuring safety. This study investigates the mechanical properties and fatigue fracture propagation of steel used in underwater vehicle rudders under various low-temperature conditions through experimental research. It compares and analyzes the static mechanical characteristics, fatigue crack growth rate, and fracture morphology of underwater manned vehicle rudder steels at different low temperatures. Findings show enhancements in yield strength, tensile strength, elastic modulus, and fatigue crack propagation life of steel 925A, steel 20#, and their welded parts under low-temperature conditions. The tensile strength of 925A steel, 20# steel, and their welded parts increases by 6.87%, 14.61%, and 12.55%, respectively, as the temperature decreases from 20 to -60 °C. The yield strength also increases by 14.17%, 29.09%, and 15.76%, respectively. Fatigue crack propagation rate experiments were conducted under different constant low-temperature conditions. This study offers direction for future modeling and experimental testing.

Exploring the ability of novel choroidal biomarkers in predicting treatment outcomes of polypoidal choroidal vasculopathy.

INTRODUCTION: To explore the relationship between choroidal biomarkers and the response to anti- VEGF in PCV eyes. METHODS: A hospital-based retrospective study. We included 54 patients diagnosed with PCV who had received standard 3-monthly anti-VEGF monotherapy and had finished regular follow-ups. Choroidal thickness (CT), three-dimensional choroidal vascularity index (CVI), and the vascular density of choriocapillaris (CCVD) were measured utilizing SS-OCTA. Effective and poor responders were classified based on the changes of morphologic features. Multivariate linear regression models were performed for the outcomes to determine independent prognostic factors. Receiver operating characteristic (ROC) curves were used to compare the predictive ability of CT and CVI as biomarkers between effective and poor responders. RESULTS: A higher CVI at baseline was the only factor that correlated with the poor response after 3-monthly injections of anti-VEGF (P=0.038). The greater change of central macular thickness (CMT) was significantly correlated with increased CMT (P=0.030), decreased CT (P=0.042) and decreased CVI (p=0.038) at baseline. Using ROC curves, we found that the CVI value demonstrated superior predictive ability compared to the CT value, with AUC of 0.842 and the best cut-off value of 0.445. CONCLUSION: A higher three-dimensional CVI using SS-OCTA is a promising biomarker to predict the poor anatomical response to anti-VEGF treatment in PCV patients.

Bacillus cereus CGMCC 1.60196: a promising bacterial inoculant isolated from biological soil crusts for maize growth enhancement.

Soil microbial inoculants are widely recognized as an environmentally friendly strategy for promoting crop growth and increasing productivity. However, research on utilizing the microbial resources from desert biological soil crusts to enhance crop growth remains relatively unexplored. In the present work, a bacterial strain designated AC1-8 with high levels of amylase, protease, and cellulase activity was isolated from cyanobacterial crusts of the Tengger Desert and identified as Bacillus cereus (CGMCC 1.60196). The refinement of the fermentation parameters of B. cereus CGMCC 1.60196 determined that the most effective medium for biomass production was composed of 5 g/L glucose, 22 g/L yeast extract and 15 g/L MgSO4, and the optimal culture conditions were pH 6.0, temperature 37°C, inoculation quantity 3% and agitation speed 240 rpm. Furthermore, the utilization of B. cereus CGMCC 1.60196 has resulted in substantial improvements in various growth parameters of maize seedlings, including shoot length, shoot fresh and dry weights, root fresh and dry weights, and the contents of chlorophyll a, chlorophyll b, and total chlorophyll. The most pronounced growth promotion was observed at an application concentration of 1 × 109 CFU/m2. These results suggest that the novel B. cereus strain, isolated from cyanobacterial crusts, can be regarded as an exemplary biological agent for soil improvement, capable of enhancing soil conditions, promoting crop cultivation and supporting food production.

Unsupervised Dual Deep Hashing with Semantic-Index and Content-Code for Cross-Modal Retrieval.

Hashing technology has exhibited great cross-modal retrieval potential due to its appealing retrieval efficiency and storage effectiveness. Most current supervised cross-modal retrieval methods heavily rely on accurate semantic supervision, which is intractable for annotations with ever-growing sample sizes. By comparison, the existing unsupervised methods rely on accurate sample similarity preservation strategies with intensive computational costs to compensate for the lack of semantic guidance, which causes these methods to lose the power to bridge the semantic gap. Furthermore, both kinds of approaches need to search for the nearest samples among all samples in a large search space, whose process is laborious. To address these issues, this paper proposes an unsupervised dual deep hashing (UDDH) method with semantic-index and content-code for cross-modal retrieval. Deep hashing networks are utilized to extract deep features and jointly encode the dual hashing codes in a collaborative manner with a common semantic index and modality content codes to simultaneously bridge the semantic and heterogeneous gaps for cross-modal retrieval. The dual deep hashing architecture, comprising the head code on semantic index and tail codes on modality content, enhances the efficiency for cross-modal retrieval. A query sample only needs to search for the retrieved samples with the same semantic index, thus greatly shrinking the search space and achieving superior retrieval efficiency. UDDH integrates the learning processes of deep feature extraction, binary optimization, common semantic index, and modality content code within a unified model, allowing for collaborative optimization to enhance the overall performance. Extensive experiments are conducted to demonstrate the retrieval superiority of the proposed approach over the state-of-the-art baselines.

Differences Between French and English in the Use of Suprasegmental Cues for the Short-Term Recall of Word Lists.

PURPOSE: Greater recognition of the impact of hearing loss on cognitive functions has led speech/hearing clinics to focus more on auditory memory outcomes. Typically evaluated by scoring participants' recall on a list of unrelated words after they have heard the list read out loud, this method implies pitch and timing variations across words. Here, we questioned whether these variations could impact performance differentially in one language or another. METHOD: In a series of online studies evaluating auditory short-term memory in normally hearing adults, we examined how pitch patterns (Experiment 1), timing patterns (Experiment 2), and interactions between the two (Experiment 3) affected free recall of words, cued recall of forgotten words, and mental demand. Note that visual memory was never directly tested; written words were only used after auditory encoding in the cued recall part. Studies were administered in both French and English, always conducted with native listeners. RESULT: Confirming prior work, grouping mechanisms facilitated free recall, but not cued recall (the latter being only affected by longer presentation time) or ratings of mental demand. Critically, grouping by pitch provided more benefit for French than for English listeners, while grouping by time was equally beneficial in both languages. CONCLUSION: Pitch is more useful to French- than to English-speaking listeners for encoding spoken words in short-term memory, perhaps due to the syllable-based versus stress-based rhythms inherent to each language. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.27048328.

Preparation of accelerated-wound-healing lignin/dopamine-based nano-Fe3O4 hydrogels in sensing.

Flexible conductive hydrogels hold great promise for applications in motion and medical detection. It is difficult to produce conductive hydrogel epidermal sensors in wearable hydrogels with dependable adhesion, sensing, and wound-healing properties. Nano-Fe3O4 was used as physical cross-linking points in the polyacrylamide/polyvinyl alcohol double network (PP) to increase the strain capacity of the hydrogel. The conductive lignin-dopamine (LD) was immobilized on the surface of Fe3O4 particles, and the LD-coated Fe3O4 was then incorporated into the double network hydrogel to create the PP/LD/Fe3O4 hydrogel. This work was done to look into the possibility of using Fe3O4 hydrogels as flexible strain sensors. The addition of LD/Fe3O4 caused the composite hydrogel to strain up to 124 %, with a modulus of elasticity of 21,308 Pa and electrical conductivity as high as 2.3 S•m-1 following the introduction of LD/Fe3O4. Moreover, the PP/LD/Fe3O4 hydrogel's adhesive qualities offered adequate antimicrobial properties and promoted wound healing. These results indicate that the developed electricity-responsive and tissue-adhesive hydrogel dressing offers a candidate to serve as a tissue sealant for wound healing.

Focusing on CD8+ T-cell phenotypes: improving solid tumor therapy.

Vigorous CD8+ T cells play a crucial role in recognizing tumor cells and combating solid tumors. How T cells efficiently recognize and target tumor antigens, and how they maintain the activity in the "rejection" of solid tumor microenvironment, are major concerns. Recent advances in understanding of the immunological trajectory and lifespan of CD8+ T cells have provided guidance for the design of more optimal anti-tumor immunotherapy regimens. Here, we review the newly discovered methods to enhance the function of CD8+ T cells against solid tumors, focusing on optimizing T cell receptor (TCR) expression, improving antigen recognition by engineered T cells, enhancing signal transduction of the TCR-CD3 complex, inducing the homing of polyclonal functional T cells to tumors, reversing T cell exhaustion under chronic antigen stimulation, and reprogramming the energy and metabolic pathways of T cells. We also discuss how to participate in the epigenetic changes of CD8+ T cells to regulate two key indicators of anti-tumor responses, namely effectiveness and persistence.

Mediator complex subunit 1 promotes oral squamous cell carcinoma progression by activating MMP9 transcription and suppressing CD8+ T cell antitumor immunity.

BACKGROUND: The role of Mediator complex subunit 1 (MED1), a pivotal transcriptional coactivator implicated in diverse biological pathways, remains unexplored in the context of oral squamous cell carcinoma (OSCC). This study aims to elucidate the contributory mechanisms and potential impact of MED1 on the progression of OSCC. METHODS: The expression and clinical significance of MED1 in OSCC tissues were evaluated through the bioinformatics analyses. The effects of MED1 on the biological behavior of OSCC cancer cells were assessed both in vitro and in vivo. Dual-luciferase reporter assay, chromatin immunoprecipitation (ChIP) assay, bioinformatic analysis, CD8+ T cell isolation experiment, coculture experiment, enzyme-linked immunosorbent assay (ELISA), and flow cytometric analysis were employed to elucidate the underlying mechanism through which MED1 operates in the progression of OSCC. RESULTS: MED1 exhibited upregulation in both OSCC tissues and multiple OSCC cell lines, which correlated with decreased overall survival in patients. In vitro experiments demonstrated that knockdown of MED1 in metastatic OSCC cell lines SCC-9 and UPCI-SCC-154 hindered cell migration and invasion, while overexpression of MED1 promoted these processes. Whereas, MED1 knockdown had no impact on proliferation of cell lines mentioned above. In vivo studies further revealed that downregulation of MED1 effectively suppressed distant metastasis in OSCC. Mechanistically, MED1 enhanced the binding of transcription factors c-Jun and c-Fos to the matrix metalloprotein 9 (MMP9) promoters, resulting in a significant upregulation of MMP9 transcription. This process contributes to the migration and invasion of SCC-9 and UPCI-SCC-154 cells. Furthermore, MED1 modulated the expression of programmed death-ligand 1 (PD-L1) through the Notch signaling pathway, consequently impacting the tumor-killing capacity of CD8+ T cells in the tumor microenvironment. CONCLUSIONS: Our findings indicate that MED1 plays a pivotal role in OSCC progression through the activation of MMP9 transcription and suppression of CD8+ T cell antitumor immunity, suggesting that MED1 may serve as a novel prognostic marker and therapeutic target in OSCC.

Electrospun Lithium Porous Nanosorbent Fibers for Enhanced Lithium Adsorption and Sustainable Applications.

Electrospun nanosorbent fibers specifically designed for efficient lithium extraction were developed, exhibiting superior physicochemical properties. These fibers were fabricated using a polyacrylonitrile/dimethylformamide matrix, with viscosity and dynamic mechanical analysis showing that optimal interactions were achieved at lower contents of layered double hydroxide. This meticulous adjustment in formulation led to the creation of lithium porous nanosorbent fibers (Li-PNFs-1). Li-PNFs-1 exhibited outstanding mechanical attributes, including a yield stress of 0.09 MPa, a tensile strength of 2.48 MPa, and an elongation at a break of 19.7%. Additionally, they demonstrated pronounced hydrophilicity and hierarchical porous architecture, which greatly favor rapid wetting kinetics and lithium adsorption. Morphologically, they exhibited uniform smoothness with a diameter averaging 546 nm, indicative of orderly crystalline growth and a dense molecular arrangement. X-ray photoelectron spectroscopy and density functional theory using Cambridge Serial Total Energy Package revealed modifications in the spatial and electronic configurations of polyacrylonitrile due to hydrogen bonding, facilitating lithium adsorption capacity up to 13.45 mg/g under optimal conditions. Besides, kinetics and isotherm showed rapid equilibrium within 60 min and confirmed the chemical and selective nature of Li+ uptake. These fibers demonstrated consistent adsorption performance across multiple cycles, highlighting their potential for sustainable use in industrial applications.

Cloud-Edge Collaborative Defect Detection Based on Efficient Yolo Networks and Incremental Learning.

Defect detection constitutes one of the most crucial processes in industrial production. With a continuous increase in the number of defect categories and samples, the defect detection model underpinned by deep learning finds it challenging to expand to new categories, and the accuracy and real-time performance of product defect detection are also confronted with severe challenges. This paper addresses the problem of insufficient detection accuracy of existing lightweight models on resource-constrained edge devices by presenting a new lightweight YoloV5 model, which integrates four modules, SCDown, GhostConv, RepNCSPELAN4, and ScalSeq. Here, this paper abbreviates it as SGRS-YoloV5n. Through the incorporation of these modules, the model notably enhances feature extraction and computational efficiency while reducing the model size and computational load, making it more conducive for deployment on edge devices. Furthermore, a cloud-edge collaborative defect detection system is constructed to improve detection accuracy and efficiency through initial detection by edge devices, followed by additional inspection by cloud servers. An incremental learning mechanism is also introduced, enabling the model to adapt promptly to new defect categories and update its parameters accordingly. Experimental results reveal that the SGRS-YoloV5n model exhibits superior detection accuracy and real-time performance, validating its value and stability for deployment in resource-constrained environments. This system presents a novel solution for achieving efficient and accurate real-time defect detection.

The mechanism of action and chemical synthesis of FDA newly approved drug molecules.

In 2023, the U.S. Food and Drug Administration has approved 29 small molecule drugs. These newly approved small molecule drugs possess the distinct scaffolds, thereby exhibiting diverse mechanisms of action and binding modalities. Moreover, the marketed drugs have always been an important source of new drug development and creative inspiration, thereby fostering analogous endeavors in drug discovery that potentially extend to the diverse clinical indications. Therefore, conducting a comprehensive evaluation of drug approval experience and associated information will facilitate the expedited identification of highly potent drug molecules. In this review, we comprehensively summarized the relevant information regarding the clinical applications, mechanisms of action and chemical synthesis of 29 small molecule drugs, with the aim of providing a promising structural basis and design inspiration for pharmaceutical chemists.

Three-dimensional atomic insights into the metal-oxide interface in Zr-ZrO2 nanoparticles.

Metal-oxide interfaces with poor coherency have specific properties comparing to bulk materials and offer broad applications in heterogeneous catalysis, battery, and electronics. However, current understanding of the three-dimensional (3D) atomic metal-oxide interfaces remains limited because of their inherent structural complexity and the limitations of conventional two-dimensional imaging techniques. Here, we determine the 3D atomic structure of metal-oxide interfaces in zirconium-zirconia nanoparticles using atomic-resolution electron tomography. We quantitatively analyze the atomic concentration and the degree of oxidation, and find the coherency and translational symmetry of the interfaces are broken. Atoms at the interface have low structural ordering, low coordination, and elongated bond length. Moreover, we observe porous structures such as Zr vacancies and nano-pores, and investigate their distribution. Our findings provide a clear 3D atomic picture of metal-oxide interface with direct experimental evidence. We anticipate this work could encourage future studies on fundamental problems of oxides, such as interfacial structures in semiconductor and atomic motion during oxidation process.

[Early classification and recognition algorithm for sudden cardiac arrest based on limited electrocardiogram data trained with a two-stages convolutional neural network].

Sudden cardiac arrest (SCA) is a lethal cardiac arrhythmia that poses a serious threat to human life and health. However, clinical records of sudden cardiac death (SCD) electrocardiogram (ECG) data are extremely limited. This paper proposes an early prediction and classification algorithm for SCA based on deep transfer learning. With limited ECG data, it extracts heart rate variability features before the onset of SCA and utilizes a lightweight convolutional neural network model for pre-training and fine-tuning in two stages of deep transfer learning. This achieves early classification, recognition and prediction of high-risk ECG signals for SCA by neural network models. Based on 16 788 30-second heart rate feature segments from 20 SCA patients and 18 sinus rhythm patients in the international publicly available ECG database, the algorithm performance evaluation through ten-fold cross-validation shows that the average accuracy (Acc), sensitivity (Sen), and specificity (Spe) for predicting the onset of SCA in the 30 minutes prior to the event are 91.79%, 87.00%, and 96.63%, respectively. The average estimation accuracy for different patients reaches 96.58%. Compared to traditional machine learning algorithms reported in existing literatures, the method proposed in this paper helps address the requirement of large training datasets for deep learning models and enables early and accurate detection and identification of high-risk ECG signs before the onset of SCA.

[Study on direct ventricular assist loading mode based on a finite element method].

To investigate the biomechanical effects of direct ventricular assistance and explore the optimal loading mode, this study established a left ventricular model of heart failure patients based on the finite element method. It proposed a loading mode that maintains peak pressure compression, and compared it with the traditional sinusoidal loading mode from both hemodynamic and biomechanical perspectives. The results showed that both modes significantly improved hemodynamic parameters, with ejection fraction increased from a baseline of 29.33% to 37.32% and 37.77%, respectively, while peak pressure, stroke volume, and stroke work parameters also increased. Additionally, both modes showed improvements in stress concentration and excessive fiber strain. Moreover, considering the phase error of the assist device's working cycle, the proposed assist mode in this study was less affected. Therefore, this research may provide theoretical support for the design and optimization of direct ventricular assist devices.

Make or Break: The Influence of Expected Challenges and Rewards on the Motivation and Experience Associated with Cognitive Effort Exertion.

Challenging goals can induce harder work but also greater stress, in turn potentially undermining goal achievement. We sought to examine how mental effort and subjective experiences thereof interact as a function of the challenge level and the size of the incentives at stake. Participants performed a task that rewarded individual units of effort investment (correctly performed Stroop trials) but only if they met a threshold number of correct trials within a fixed time interval (challenge level). We varied this challenge level (Study 1, n = 40) and the rewards at stake (Study 2, n = 79) and measured variability in task performance and self-reported affect across task intervals. Greater challenge and higher rewards facilitated greater effort investment but also induced greater stress, whereas higher rewards (and lower challenge) simultaneously induced greater positive affect. Within intervals, we observed an initial speed up then slowdown in performance, which could reflect dynamic reconfiguration of control. Collectively, these findings further our understanding of the influence of task demands and incentives on mental effort exertion and well-being.

Changes in the medial prefrontal cortex metabolites after 6 months of medication therapy for patients with bipolar disorder: A 1H-MRS study.

AIMS: The study aimed to assess brain metabolite differences in the medial prefrontal cortex (mPFC) between acute and euthymic episodes of bipolar disorder (BD) with both mania and depression over a 6-month medication treatment period. METHODS: We utilized 1H-MRS technology to assess the metabolite levels in 53 individuals with BD (32 in depressive phase, 21 in manic phase) and 34 healthy controls (HCs) at baseline. After 6 months of medication treatment, 40 subjects underwent a follow-up scan in euthymic state. Metabolite levels, including N-acetyl aspartate (NAA), glutamate (Glu), and Glutamine (Gln), were measured in the mPFC. RESULTS: Patients experiencing depressive and manic episodes exhibited a notable reduction in NAA/Cr + PCr ratios at baseline compared to healthy controls (p = 0.004; p = 0.006) in baseline, compared with HCs. Over the 6-month follow-up period, the manic group displayed a significant decrease in Gln/Cr + PCr compared to the initial acute phase (p = 0.03). No significant alterations were found in depressed group between baseline and follow-up. CONCLUSION: This study suggests that NAA/Cr + PCr ratios and Gln/Cr + PCr ratios in the mPFC may be associated with manic and depressive episodes, implicating that Gln and NAA might be useful biomarkers for distinguishing mood phases in BD and elucidating its mechanisms.

Semantic Interaction Meta-Learning Based on Patch Matching Metric.

Metric-based meta-learning methods have demonstrated remarkable success in the domain of few-shot image classification. However, their performance is significantly contingent upon the choice of metric and the feature representation for the support classes. Current approaches, which predominantly rely on holistic image features, may inadvertently disregard critical details necessary for novel tasks, a phenomenon known as "supervision collapse". Moreover, relying solely on visual features to characterize support classes can prove to be insufficient, particularly in scenarios involving limited sample sizes. In this paper, we introduce an innovative framework named Patch Matching Metric-based Semantic Interaction Meta-Learning (PatSiML), designed to overcome these challenges. To counteract supervision collapse, we have developed a patch matching metric strategy based on the Transformer architecture to transform input images into a set of distinct patch embeddings. This approach dynamically creates task-specific embeddings, facilitated by a graph convolutional network, to formulate precise matching metrics between the support classes and the query image patches. To enhance the integration of semantic knowledge, we have also integrated a label-assisted channel semantic interaction strategy. This strategy merges word embeddings with patch-level visual features across the channel dimension, utilizing a sophisticated language model to combine semantic understanding with visual information. Our empirical findings across four diverse datasets reveal that the PatSiML method achieves a classification accuracy improvement of 0.65% to 21.15% over existing methodologies, underscoring its robustness and efficacy.

Human-in-the-Loop Trajectory Optimization Based on sEMG Biofeedback for Lower-Limb Exoskeleton.

Lower-limb exoskeletons (LLEs) can provide rehabilitation training and walking assistance for individuals with lower-limb dysfunction or those in need of functionality enhancement. Adapting and personalizing the LLEs is crucial for them to form an intelligent human-machine system (HMS). However, numerous LLEs lack thorough consideration of individual differences in motion planning, leading to subpar human performance. Prioritizing human physiological response is a critical objective of trajectory optimization for the HMS. This paper proposes a human-in-the-loop (HITL) motion planning method that utilizes surface electromyography signals as biofeedback for the HITL optimization. The proposed method combines offline trajectory optimization with HITL trajectory selection. Based on the derived hybrid dynamical model of the HMS, the offline trajectory is optimized using a direct collocation method, while HITL trajectory selection is based on Thompson sampling. The direct collocation method optimizes various gait trajectories and constructs a gait library according to the energy optimality law, taking into consideration dynamics and walking constraints. Subsequently, an optimal gait trajectory is selected for the wearer using Thompson sampling. The selected gait trajectory is then implemented on the LLE under a hybrid zero dynamics control strategy. Through the HITL optimization and control experiments, the effectiveness and superiority of the proposed method are verified.