Early identification of stroke through deep learning with multi-modal human speech and movement data.

JOURNAL/nrgr/04.03/01300535-202501000-00031/figure1/v/2024-05-14T021156Z/r/image-tiff Early identification and treatment of stroke can greatly improve patient outcomes and quality of life. Although clinical tests such as the Cincinnati Pre-hospital Stroke Scale (CPSS) and the Face Arm Speech Test (FAST) are commonly used for stroke screening, accurate administration is dependent on specialized training. In this study, we proposed a novel multimodal deep learning approach, based on the FAST, for assessing suspected stroke patients exhibiting symptoms such as limb weakness, facial paresis, and speech disorders in acute settings. We collected a dataset comprising videos and audio recordings of emergency room patients performing designated limb movements, facial expressions, and speech tests based on the FAST. We compared the constructed deep learning model, which was designed to process multi-modal datasets, with six prior models that achieved good action classification performance, including the I3D, SlowFast, X3D, TPN, TimeSformer, and MViT. We found that the findings of our deep learning model had a higher clinical value compared with the other approaches. Moreover, the multi-modal model outperformed its single-module variants, highlighting the benefit of utilizing multiple types of patient data, such as action videos and speech audio. These results indicate that a multi-modal deep learning model combined with the FAST could greatly improve the accuracy and sensitivity of early stroke identification of stroke, thus providing a practical and powerful tool for assessing stroke patients in an emergency clinical setting.

Surface Plasmon Resonance-Enhanced CdS/FTO Heterojunction for Cu2+ Detection.

Copper ion (Cu2+) pollution poses a serious threat to marine ecology and fisheries. However, the complexity of seawater and its interference factors make the online detection of Cu2+ quite challenging. To address this issue, we introduce the concept of the photo-assisted adjustment barrier effect into electrochemical detection, using it as a driving force to generate electrochemical responses. The Schottky barrier demonstrates a remarkable regulatory influence on the electrochemical response under photoexcitation, facilitating the response through Cu2+ adsorption. We developed a 4-MBA-AuNPs/CdS/FTO composite that serves as a sensitive platform for Cu2+ detection, achieving a detection limit of 70 nM. Notably, the photo-assisted adjustment of the barrier effect effectively counters the interference posed by ions in seawater, ensuring accurate detection. Furthermore, the sensor exhibits a promising recovery rate (99.62-104.9%) in real seawater samples, highlighting its practical applications. This innovative approach utilizing the photo-assisted adjustment barrier effect offers a promising path for developing electrochemical sensors that can withstand interference.

Design and experimental evaluation of a variable pesticide application control system for the air-assisted rubber tree powder sprayer.

BACKGROUND: In order to address the issues of uneven pesticide deposition and low pesticide utilization in rubber gardens caused by the traditional diffuse plant protection spraying method, this study focuses on the air-assisted powder sprayer and proposes a variable pesticide application control system. A variable pesticide application decision-making model integrating the leaf area index (LAI) was designed based on powdery mildew control standards and individual rubber tree information. According to the target powder spraying accuracy requirements, a control model of the air velocity adjustment device was established and a fuzzy proportional-integral-differential (PID) air velocity control system was developed. RESULTS: The simulation results indicate that the wind speed control system exhibits a maximum overshoot of 2.18% and an average response time of 1.48 s. The field experiment conducted in a rubber plantation revealed that when the air-assisted powder sprayer operates in the variable powder spraying mode, the average response time of the control system is 2.5 s. The control accuracy of each executive mechanism exceeded 95.9%. The deposition coefficient of variation (CV) at different canopy heights was relatively consistent, with values of 35.38%, 36.26% and 36.90%. In comparison to the quantitative mode, the variable mode showed a significant 20.03% increase in the effective utilization rate of sulfur powder. CONCLUSION: These research findings provide valuable technical support for the advancement of mechanized variable powder spraying equipment in rubber tree cultivation. © 2024 Society of Chemical Industry.

Mechanistic Insight into the Synthesis and Morphological Evolution of Superhydrophobic Silica Nanotubes.

Silica nanotubes have significant applications in various fields, including thermal insulation, self-cleaning, and catalysis. Currently, the synthesis methods of silica nanotubes are mostly limited to the template method. In this work, a template-free strategy and vapor-phase approach were used to prepare silica nanotubes. Poly(methylhydrosiloxane) (PMHS) was hydrolyzed and condensed in a high-temperature closed reactor by using ammonia as a catalyst. The resulting product was then subjected to template-free self-assembly to synthesize silica nanotubes incorporating methyl groups. The silica nanotubes were synthesized under varying conditions, resulting in lengths ranging from 50 nm to several micrometers, exterior diameters between 40 and 120 nm, and wall thicknesses varying from 7 to 30 nm. The synthesized products underwent morphology analysis using TEM and FESEM for morphology analysis, elemental composition analysis using XPS, and chemical structure identification using FTIR, and the possible formation mechanism of silica nanotubes formation was also speculated. Furthermore, the coatings formed by silica nanotubes exhibited remarkable superhydrophobic self-cleaning properties with a water contact angle of 162° and a rolling angle of less than 1°.

HLAIImaster: a deep learning method with adaptive domain knowledge predicts HLA II neoepitope immunogenic responses.

While significant strides have been made in predicting neoepitopes that trigger autologous CD4+ T cell responses, accurately identifying the antigen presentation by human leukocyte antigen (HLA) class II molecules remains a challenge. This identification is critical for developing vaccines and cancer immunotherapies. Current prediction methods are limited, primarily due to a lack of high-quality training epitope datasets and algorithmic constraints. To predict the exogenous HLA class II-restricted peptides across most of the human population, we utilized the mass spectrometry data to profile >223 000 eluted ligands over HLA-DR, -DQ, and -DP alleles. Here, by integrating these data with peptide processing and gene expression, we introduce HLAIImaster, an attention-based deep learning framework with adaptive domain knowledge for predicting neoepitope immunogenicity. Leveraging diverse biological characteristics and our enhanced deep learning framework, HLAIImaster is significantly improved against existing tools in terms of positive predictive value across various neoantigen studies. Robust domain knowledge learning accurately identifies neoepitope immunogenicity, bridging the gap between neoantigen biology and the clinical setting and paving the way for future neoantigen-based therapies to provide greater clinical benefit. In summary, we present a comprehensive exploitation of the immunogenic neoepitope repertoire of cancers, facilitating the effective development of "just-in-time" personalized vaccines.

Divergent landscapes of A-to-I editing in postmortem and living human brain.

Adenosine-to-inosine (A-to-I) editing is a prevalent post-transcriptional RNA modification within the brain. Yet, most research has relied on postmortem samples, assuming it is an accurate representation of RNA biology in the living brain. We challenge this assumption by comparing A-to-I editing between postmortem and living prefrontal cortical tissues. Major differences were found, with over 70,000 A-to-I sites showing higher editing levels in postmortem tissues. Increased A-to-I editing in postmortem tissues is linked to higher ADAR and ADARB1 expression, is more pronounced in non-neuronal cells, and indicative of postmortem activation of inflammation and hypoxia. Higher A-to-I editing in living tissues marks sites that are evolutionarily preserved, synaptic, developmentally timed, and disrupted in neurological conditions. Common genetic variants were also found to differentially affect A-to-I editing levels in living versus postmortem tissues. Collectively, these discoveries offer more nuanced and accurate insights into the regulatory mechanisms of RNA editing in the human brain.

Integrated multi-omic analyses uncover the effects of aging on cell-type regulation in glucose-responsive tissues.

Aging significantly influences cellular activity and metabolism in glucose-responsive tissues, yet a comprehensive evaluation of the impacts of aging and associated cell-type responses has been lacking. This study integrates transcriptomic, methylomic, single-cell RNA sequencing, and metabolomic data to investigate aging-related regulations in adipose and muscle tissues. Through coexpression network analysis of the adipose tissue, we identified aging-associated network modules specific to certain cell types, including adipocytes and immune cells. Aging upregulates the metabolic functions of lysosomes and downregulates the branched-chain amino acids (BCAAs) degradation pathway. Additionally, aging-associated changes in cell proportions, methylation profiles, and single-cell expressions were observed in the adipose. In the muscle tissue, aging was found to repress the metabolic processes of glycolysis and oxidative phosphorylation, along with reduced gene activity of fast-twitch type II muscle fibers. Metabolomic profiling linked aging-related alterations in plasma metabolites to gene expression in glucose-responsive tissues, particularly in tRNA modifications, BCAA metabolism, and sex hormone signaling. Together, our multi-omic analyses provide a comprehensive understanding of the impacts of aging on glucose-responsive tissues and identify potential plasma biomarkers for these effects.

De novo biosynthesis of ß-arbutin in Corynebacterium glutamicum via pathway engineering and process optimization.

BACKGROUND: ß-Arbutin, a hydroquinone glucoside found in pears, bearberry leaves, and various plants, exhibits antioxidant, anti-inflammatory, antimicrobial, and anticancer effects. ß-Arbutin has wide applications in the pharmaceutical and cosmetic industries. However, the limited availability of high-performance strains limits the biobased production of ß-arbutin. RESULTS: This study established the ß-arbutin biosynthetic pathway in C. glutamicum ATCC13032 by introducing codon-optimized ubiC, MNX1, and AS. Additionally, the production titer of ß-arbutin was increased by further inactivation of csm and trpE to impede the competitive metabolic pathway. Further modification of the upstream metabolic pathway and supplementation of UDP-glucose resulted in the final engineered strain, C. glutamicum AR11, which achieved a ß-arbutin production titer of 7.94 g/L in the optimized fermentation medium. CONCLUSIONS: This study represents the first successful instance of de novo ß-arbutin production in C. glutamicum, offering a chassis cell for ß-arbutin biosynthesis.

Biosynthesis of ß-nicotinamide mononucleotide from glucose via a new pathway in Bacillus subtilis.

Introduction: ß-nicotinamide mononucleotide (ß-NMN) is an essential precursor of nicotinamide adenine dinucleotide (NAD+) and plays a key role in supplying NAD+ and maintaining its levels. Existing methods for NMN production have some limitations, including low substrate availability, complex synthetic routes, and low synthetic efficiency, which result in low titers and high costs. Methods: We constructed high-titer, genetically engineered strains that produce NMN through a new pathway. Bacillus subtilis WB600 was used as a safe chassis strain. Multiple strains overexpressing NadE, PncB, and PnuC in various combinations were constructed, and NMN titers of different strains were compared via shake-flask culture. Results: The results revealed that the strain B. subtilis PncB1-PnuC exhibited the highest total and extracellular NMN titers. Subsequently, the engineered strains were cultured in a 5-L fermenter using batch and fed-batch fermentation. B. subtilis PncB1-PnuC achieved an NMN titer of 3,398 mg/L via fed-batch fermentation and glucose supplementation, which was 30.72% higher than that achieved via batch fermentation. Discussion: This study provides a safe and economical approach for producing NMN on an industrial scale.

Distinct metabolites affect the phloem fungal communities in ash trees (Fraxinus spp.) native and nonnative to the highly invasive emerald ash borer (AGRILUS PLANIPENNIS).

Emerald ash borer (EAB, Agrilus planipennis) is an invasive killer of ash trees (Fraxinus spp.) in North America and Europe. Ash species co-evolved with EAB in their native range in Asia are mostly resistant, although the precise mechanism(s) remain unclear. Very little is also known about EAB or ash tree microbiomes. We performed the first joint comparison of phloem mycobiome and metabolites between a native and a nonnative ash species, infested and uninfested with EAB, in conjunction with investigation of larval mycobiome. Phloem mycobiome communities differed between the tree species, but both were unaffected by EAB infestation. Several indicator taxa in the larval gut shared a similarly high relative abundance only with the native host trees. Widely targeted metabolomics revealed 24 distinct metabolites in native trees and 53 metabolites in nonnative trees, respectively, that differed in relative content between infested and uninfested trees only in one species. Interestingly, four metabolites shared a strong relationship with the phloem mycobiomes, majority of which affected only the native trees. Collectively, our results demonstrate a complex interplay between host tree chemistry and mycobiome, and suggest the shared relationships between the mycobiomes of the native host tree and EAB may reflect their shared co-evolution.

CODC-v1: a quality-controlled and bias-corrected ocean temperature profile database from 1940-2023.

High-quality ocean in situ profile observations are fundamental for ocean and climate research and operational oceanographic applications. Here we describe a new global ocean subsurface temperature profile database named the Chinese Academy of Science (CAS) Oceanography Data Center version 1 (CODC-v1). This database contains over 17 million temperature profiles between 1940-2023 from all available instruments. The major data source is the World Ocean Database (WOD), but CODC-v1 also includes some data from some Chinese institutes which are not available in WOD. The data are quality-controlled (QC-ed) by a new QC system that considers the skewness of local temperature distributions, topographic barriers, and the shift of temperature distributions due to climate change. Biases in Mechanical Bathythermographs (MBTs), eXpendable Bathythermographs (XBTs), and Bottle data (OSD) are all corrected using recently proposed correction schemes, which makes CODC-v1 a bias-corrected dataset. These aspects ensure the data quality of the CODC-v1 database, making it suitable for a wide spectrum of ocean and climate research and applications.

An Assessment of Administration Route on MSC-sEV Therapeutic Efficacy.

Mesenchymal stem/stromal cell-derived small extracellular vesicles (MSC-sEVs) are promising therapeutic agents. In this study, we investigated how the administration route of MSC-sEVs affects their therapeutic efficacy in a mouse model of bleomycin (BLM)-induced skin scleroderma (SSc). We evaluated the impact of topical (TOP), subcutaneous (SC), and intraperitoneal (IP) administration of MSC-sEVs on dermal fibrosis, collagen density, and thickness. All three routes of administration significantly reduced BLM-induced fibrosis in the skin, as determined by Masson's Trichrome staining. However, only TOP administration reduced BLM-induced dermal collagen density, with no effect on dermal thickness observed for all administration routes. Moreover, SC, but not TOP or IP administration, increased anti-inflammatory profibrotic CD163+ M2 macrophages. These findings indicate that the administration route influences the therapeutic efficacy of MSC-sEVs in alleviating dermal fibrosis, with TOP administration being the most effective, and this efficacy is not mediated by M2 macrophages. Since both TOP and SC administration target the skin, the difference in their efficacy likely stems from variations in MSC-sEV delivery in the skin. Fluorescence-labelled TOP, but not SC MSC-sEVs when applied to skin explant cultures, localized in the stratum corneum. Hence, the superior efficacy of TOP over SC MSC-sEVs could be attributed to this localization. A comparison of the proteomes of stratum corneum and MSC-sEVs revealed the presence of >100 common proteins. Most of these proteins, such as filaggrin, were known to be crucial for maintaining skin barrier function against irritants and toxins, thereby mitigating inflammation-induced fibrosis. Therefore, the superior efficacy of TOP MSC-sEVs over SC and IP MSC-sEVs against SSc is mediated by the delivery of proteins to the stratum corneum to reinforce the skin barrier.

Robust k -Means-Type Clustering for Noisy Data.

Data clustering is a fundamental machine learning task that seeks to categorize a dataset into homogeneous groups. However, real data usually contain noise, which poses significant challenges to clustering algorithms. In this article, motivated by how the k -means algorithm is derived from a Gaussian mixture model (GMM), we propose a robust k -means-type algorithm, named k -means-type clustering based on t -distribution (KMTD), by assuming that the data points are drawn from a special multivariate t -mixture model (TMM). Compared to the Gaussian distribution, the t -distribution has a fatter tail. The proposed algorithm is more robust to noise. Like the k -means algorithm, the proposed algorithm is simpler than those based on a full TMM. Both synthetic and actual data are used to illustrate the proposed algorithm's performance and efficiency. The experimental results demonstrated that the proposed algorithm operates more quickly than other sophisticated algorithms and, in most cases, achieves higher accuracy than the other algorithms.

Distantly Supervised Biomedical Relation Extraction Via Negative Learning and Noisy Student Self-Training.

Biomedical relation extraction aims to identify underlying relationships among entities, such as gene associations and drug interactions, within biomedical texts. Despite advancements in relation extraction in general knowledge domains, the scarcity of labeled training data remains a significant challenge in the biomedical field. This paper provides a novel approach for biomedical relation extraction that leverages a noisy student self-training strategy combined with negative learning. This method addresses the challenge of data insufficiency by utilizing distantly supervised data to generate high-quality labeled samples. Negative learning, as opposed to traditional positive learning, offers a more robust mechanism to discern and relabel noisy samples, preventing model overfitting. The integration of these techniques ensures enhanced noise reduction and relabeling capabilities, leading to improved performance even with noisy datasets. Experimental results demonstrate the effectiveness of the proposed framework in mitigating the impact of noisy data and outperforming existing benchmarks.

Sustainable double-synergistic silver-hydroxyapatite composite catalyst derived from fish bones for efficient disinfection of Vibrio parahaemolyticus.

Vibrio parahaemolyticus is a food-borne pathogen that poses a serious threat to seafood safety and human health. An efficient, nontoxic, and sustainable disinfection material with a stable structure is urgently needed. Herein, silver (Ag)-hydroxyapatite (HAP) composite catalysts were prepared using HAP derived from waste fish bones. The Ag2.50%-HAP showed a 100% disinfection rate against V. parahaemolyticus, disinfecting nearly 7.0 lg CFU mL-1 within 15 min at a low concentration of 300 µg mL-1. This efficient disinfection activity could be attributed to the double-synergistic effect of Ag and superoxide radicals, which resulted in the destruction of bacterial cell structures and the leakage of intracellular proteins. Importantly, the composite also exhibited high activity in controlling the growth of pathogens during the storage process of Penaeus vannamei. These findings provided sustainable composite catalysts for disinfecting V. parahaemolyticus in seafood and a high-value utilization strategy for waste fish bones.

Analysis and experimental verification of the polarization characteristics of cube-corner reflectors.

The polarization effect of cube-corner reflectors (CCRs), which influences the performance of optical systems, requires comprehensive analysis. This study developed a model for the polarization state of uncoated solid and hollow CCRs using the Jones matrix derivation and Zemax software simulations. The accuracies of theoretical analyses and simulations were verified using an experimental setup. Theoretical analysis, simulation, and experimental results revealed that hollow CCRs are insensitive to the polarization state of the incident light, exhibiting average variations of 0.8° and 0.7° in the polarization direction and ellipticity, respectively. Contrastingly, the high sensitivity of solid CCRs to the polarization state of the incident light varied across different incident regions. The propagation paths 2-1-3 and 3-1-2 with minor polarization effects involved light that entered from one side of the CCR, traversed the bottom, and emitted from the other side. In these regions, the average variations in the polarization direction and ellipticity were 10.7° and 6.6°, respectively, whereas more affected regions exhibited corresponding values of 44.8° and 20.0°. These findings guide the enhancement and optimization of the performance of optical systems using CCRs.

Association between Pericoronary Fat Attenuation Index Values and Plaque Composition Volume Fraction Measured by Coronary Computed Tomography Angiography.

RATIONALE AND OBJECTIVES: The pericoronary fat attenuation index (FAI) values around plaques may reveal the relationship between periplaque vascular inflammation and different plaque component volume fractions. We aimed to evaluate the potential associations between periplaque FAI values and plaque component volume fractions. MATERIALS AND METHODS: 496 patients (1078 lesions) with coronary artery disease, who underwent computed tomography angiography (CCTA) between September 2022 and August 2023, were analyzed retrospectively. Each lesion was characterized and the plaque component volume fractions and periplaque FAI values were measured. Multiple linear regression, weighted quantile sum (WQS) regression, and quantile g-computation (Qgcomp) were used to explore the relationship between plaque component volume fractions and the risk of elevated periplaque FAI values. RESULTS: After adjusting for clinical characteristics, multiple linear regression identified that lipid components volume fraction (ß = 0.162, P < 0.001) were independent risk factors for elevated periplaque FAI values whereas calcified components volume fraction (ß = -0.066, P = 0.025) were independent protective factors. The WQS regression models indicated an increase in the overall confounding effect of the adjusted lipid indices and plaque composition volume fraction on the risk of elevated periplaque FAI values (P = 0.004). Qgcomp analysis indicated lipid component volume fraction and calcified component volume fraction was positively and negatively correlated with elevated plaque FAI values, respectively (all P < 0.05). CONCLUSIONS: Periplaque FAI values quantified by CCTA were strongly correlated with lipid and calcification component volume fractions.

Engineering Core/Ligands Interfacial Anchors of Nanoparticles for Efficiently Inhibiting Both Aß and Amylin Fibrillization.

Accurate construction of artificial nano-chaperones' structure is crucial for precise regulation of protein conformational transformation, facilitating effective treatment of proteopathy. However, how the ligand-anchors of nano-chaperones affect the spatial conformational changes in proteins remains unclear, limiting the development of efficient nano-chaperones. In this study, three types of gold nanoparticles (AuNPs) with different core/ligands interface anchor structures (Au─NH─R, Au─S─R, and Au─C≡C─R, R = benzoic acid) are synthesized as an ideal model to investigate the effect of interfacial anchors on Aß and amylin fibrillization. Computational results revealed that the distinct interfacial anchors imparted diverse distributions of electrostatic potential on the nanointerface and core/ligands bond strength of AuNPs, leading to differential interactions with amyloid peptides. Experimental results demonstrated that all three types of AuNPs exhibit site-specific inhibitory effects on Aß40 fibrillization due to preferential binding. For amylin, amino-anchored AuNPs demonstrate strong adsorption to multiple sites on amylin and effectively inhibit fibrillization. Conversely, thiol- and alkyne-anchored AuNPs adsorb at the head region of amylin, promoting folding and fibrillization. This study not only provided molecular insights into how core/ligands interfacial anchors of nanomaterials induce spatial conformational changes in amyloid peptides but also offered guidance for precisely engineering artificial-chaperones' nanointerfaces to regulate the conformational transformation of proteins.