The impact of fine-tuning paradigms on unknown plant diseases recognition.

Plant diseases pose significant threats to agriculture, impacting both food safety and public health. Traditional plant disease detection systems are typically limited to recognizing disease categories included in the training dataset, rendering them ineffective against new disease types. Although out-of-distribution (OOD) detection methods have been proposed to address this issue, the impact of fine-tuning paradigms on these methods has been overlooked. This paper focuses on studying the impact of fine-tuning paradigms on the performance of detecting unknown plant diseases. Currently, fine-tuning on visual tasks is mainly divided into visual-based models and visual-language-based models. We first discuss the limitations of large-scale visual language models in this task: textual prompts are difficult to design. To avoid the side effects of textual prompts, we futher explore the effectiveness of purely visual pre-trained models for OOD detection in plant disease tasks. Specifically, we employed five publicly accessible datasets to establish benchmarks for open-set recognition, OOD detection, and few-shot learning in plant disease recognition. Additionally, we comprehensively compared various OOD detection methods, fine-tuning paradigms, and factors affecting OOD detection performance, such as sample quantity. The results show that visual prompt tuning outperforms fully fine-tuning and linear probe tuning in out-of-distribution detection performance, especially in the few-shot scenarios. Notably, the max-logit-based on visual prompt tuning achieves an AUROC score of 94.8 % in the 8-shot setting, which is nearly comparable to the method of fully fine-tuning on the full dataset (95.2 % ), which implies that an appropriate fine-tuning paradigm can directly improve OOD detection performance. Finally, we visualized the prediction distributions of different OOD detection methods and discussed the selection of thresholds. Overall, this work lays the foundation for unknown plant disease recognition, providing strong support for the security and reliability of plant disease recognition systems. We will release our code at https://github.com/JiuqingDong/PDOOD to further advance this field.

Chitooligosaccharides and Arbuscular Mycorrhizal fungi alleviate the damage by Phytophthora nicotianae to tobacco seedlings by inducing changes in rhizosphere microecology.

Arbuscular mycorrhizal fungi (AMF) and Chitooligosaccharide (COS) can increase the resistance of plants to disease. COS can also promote the symbiosis between AMF and plants. However, the effects of AMF & COS combined application on the rhizosphere soil microbial community of tobacco and the improvement of tobacco's resistance to black shank disease are poorly understood.·We treated tobacco with AMF, COS, and combined application of AMF & COS (AC), respectively. Then studied the incidence, physio-biochemical changes, root exudates, and soil microbial diversity of tobacco seedling that was inoculated with Phytophthora nicotianae. The antioxidant enzyme activity and root vigor of tobacco showed a regular of AC > AMF > COS > CK, while the severity of tobacco disease showed the opposite regular. AMF and COS enhance the resistance to black shank disease by enhancing root vigor, and antioxidant capacity, and inducing changes in the rhizosphere microecology of tobacco. We have identified key root exudates and critical soil microorganisms that can inhibit the growth of P. nicotianae. The presence of caprylic acid in root exudates and Bacillus (WdhR-2) in rhizosphere soil microorganisms is the key factor that inhibits P. nicotianae growth. AC can significantly increase the content of caprylic acid in tobacco root exudates compared to AMF and COS. Both AMF and COS can significantly increase the abundance of Bacillus in tobacco rhizosphere soil, but the abundance of Bacillus in AC is significantly higher than that in AMF and COS. This indicates that the combined application of AMF and COS is more effective than their individual use. These findings suggest that exogenous stimuli can induce changes in plant root exudates, regulate plant rhizosphere microbial community, and then inhibit the growth of pathogens, thereby improving plant resistance to diseases.

Genome-scale CRISPR-Cas9 screening in stem cells: theories, applications and challenges.

Due to the rapid development of stem cell technology, there have been tremendous advances in molecular biological and pathological research, cell therapy as well as organoid technologies over the past decades. Advances in genome editing technology, particularly the discovery of clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-related protein 9 (Cas9), have further facilitated the rapid development of stem cell researches. The CRISPR-Cas9 technology now goes beyond creating single gene editing to enable the inhibition or activation of endogenous gene loci by fusing inhibitory (CRISPRi) or activating (CRISPRa) domains with deactivated Cas9 proteins (dCas9). These tools have been utilized in genome-scale CRISPRi/a screen to recognize hereditary modifiers that are synergistic or opposing to malady mutations in an orderly and fair manner, thereby identifying illness mechanisms and discovering novel restorative targets to accelerate medicinal discovery investigation. However, the application of this technique is still relatively rare in stem cell research. There are numerous specialized challenges in applying large-scale useful genomics approaches to differentiated stem cell populations. Here, we present the first comprehensive review on CRISPR-based functional genomics screening in the field of stem cells, as well as practical considerations implemented in a range of scenarios, and exploration of the insights of CRISPR-based screen into cell fates, disease mechanisms and cell treatments in stem cell models. This review will broadly benefit scientists, engineers and medical practitioners in the areas of stem cell research.

Prevalence of homologous recombination biomarkers in multiple tumor types: an observational study.

Aim: To determine the prevalence of deleterious mutations in BRCA1 and BRCA2 and in 13 genes involved in homologous recombination repair (HRR), the prevalence of genomic loss of heterozygosity and the allelic and hereditary status of BRCA1, BRCA2 and other HRR gene mutations in multiple solid tumor types. Patients & methods: This was a retrospective observational study of patients with an advanced/metastatic diagnosis in one of 15 solid tumor types, who were identified in a real-world clinico-genomic database. Results: Tumor tissue samples from 9457 patients were analyzed, among which 4.7% had known or suspected deleterious BRCA1/2 mutations. The prevalence (range) of mutations in HRR genes was 13.6% (2.4%-26.0%) and genomic loss of heterozygosity ≥16% was 20.6% (2.6-34.4%) across all tumor types. Conclusion: The prevalence of mutations varied significantly depending on the type of tumor.

The integrity of the human genome is maintained via multiple pathways of DNA repair, one of the most important of which is homologous recombination repair (HRR), which uses a sister chromatid as a template for high-fidelity restoration of altered DNA sequences. This study aimed to determine the prevalence of deleterious mutations, i.e., changes in the genetic code that interfere with proper cellular function, in the breast cancer genes BRCA1 and BRCA2 and in 13 other genes involved in HRR in various types of solid tumors in patients with advanced or metastatic cancer. The researchers found that 4.7% of tumor samples had BRCA1/2 mutations, 13.6% had mutations in any of the HRR genes and 20.6% had genomic loss of heterozygosity (gLOH) of at least 16% i.e., loss of sections of chromosomes affecting 16% or more of the genome. BRCA1/2 mutations were most common in ovarian cancer (13.1%), prostate cancer (9.3%), breast cancer (8.2%) and pancreatic cancer (4.9%). Prevalence for mutations in HRR genes ranges from 2.4 to 26.0% and gLOH ≥16% ranged from 2.6 to 34.4% depending on the tumor type. In conclusion, the prevalence of mutations in the BRCA1/2 genes, HRR genes and gLOH ≥16% varied widely across 15 tumor types.

Sustainable Production of Biofuels from Biomass Feedstocks Using Modified Montmorillonite Catalysts.

The rampant exploitation of fossil fuels has led to the significant energy scarcity and environmental disruption, affecting the sound momentum of development and progress of human civilization. To build a closed-loop anthropogenic carbon cycle, development of biofuels employing sustainable biomass feedstocks stands at the forefront of advancing carbon neutrality, yet its widespread adoption is mainly hampered by the high production costs. Montmorillonite, however, has garnered considerable attention serving as an efficient heterogeneous catalyst of ideal economic feasibility for biofuel production, primarily due to its affordability, accessibility, stability, and excellent plasticity. Up to now, nevertheless, it has merely received finite concerns and interests in production of various biofuels using montmorillonite-based catalysts. There is no timely and comprehensive review that addresses this latest relevant progress. This review fills the gap by providing a systematically review and summary in controllable synthesis, performance enhancement, and applications related to different kinds of biofuels including biodiesel, biohydrogenated diesel, levulinate, γ-valerolactone, 5-ethoxymethylfurfural, gaseous biofuels (CO, H2), and cycloalkane, by using montmorillonite catalysts and its modified forms. Particularly, this review critically depicts the design strategies for montmorillonite, illustrates the relevant reaction mechanisms, and assesses their economic viability, realizing sustainable biofuels production via efficient biomass valorization.

A new subtype of artificial cell-derived vesicles from dental pulp stem cells with the bioequivalence and higher acquisition efficiency compared to extracellular vesicles.

Extracellular vesicles (EVs) derived from dental pulp stem cells (DPSC) have been shown an excellent efficacy in a variety of disease models. However, current production methods fail to meet the needs of clinical treatment. In this study, we present an innovative approach to substantially enhance the production of 'Artificial Cell-Derived Vesicles (ACDVs)' by extracting and purifying the contents released by the DPSC lysate, namely intracellular vesicles. Comparative analysis was performed between ACDVs and those obtained through ultracentrifugation. The ACDVs extracted from the cell lysate meet the general standard of EVs and have similar protein secretion profile. The new ACDVs also significantly promoted wound healing, increased or decreased collagen regeneration, and reduced the production of inflammatory factors as the EVs. More importantly, the extraction efficiency is improved by 16 times compared with the EVs extracted using ultracentrifuge method. With its impressive attributes, this new subtype of ACDVs emerge as a prospective candidate for the future clinical applications in regenerative medicine.

Unsupervised domain adaptation with weak source domain labels via bidirectional subdomain alignment.

Unsupervised domain adaptation (UDA) enables knowledge transfer from a labeled source domain to an unlabeled target domain. However, UDA performance often relies heavily on the accuracy of source domain labels, which are frequently noisy or missing in real applications. To address unreliable source labels, we propose a novel framework for extracting robust, discriminative features via iterative pseudo-labeling, queue-based clustering, and bidirectional subdomain alignment (BSA). The proposed framework begins by generating pseudo-labels for unlabeled source data and constructing codebooks via iterative clustering to obtain label-independent class centroids. Then, the proposed framework performs two main tasks: rectifying features from both domains using BSA to match subdomain distributions and enhance features; and employing a two-stage adversarial process for global feature alignment. The feature rectification is done before feature enhancement, while the global alignment is done after feature enhancement. To optimize our framework, we formulate BSA and adversarial learning as maximizing a log-likelihood function, which is implemented via the Expectation-Maximization algorithm. The proposed framework shows significant improvements compared to state-of-the-art methods on Office-31, Office-Home, and VisDA-2017 datasets, achieving average accuracies of 91.5%, 76.6%, and 87.4%, respectively. Compared to existing methods, the proposed method shows consistent superiority in unsupervised domain adaptation tasks with both fully and weakly labeled source domains.

Efficiency, mechanism and application prospect of ammonium adsorption and desorption over a sodium-acetate-modified synthetic zeolite.

Adsorption is an effective approach for remediating ammonium pollution, and zeolite has exceptional efficacy for the adsorption of ammonium. The investigation of ammonium adsorption using coal-fly-ash-based zeolite has gained remarkable attention in contemporary research. In this work, a sodium-acetate-modified synthetic zeolite (MSZ) was used to absorb ammonium in simulated wastewater. The MSZ had an adsorption capacity for ammonium of 27.46 mg g-1, and the adsorption process followed the Langmuir isotherm model and pseudo-second-order kinetics model. The adsorption and desorption of ammonium were controlled by ion exchange, pore diffusion, and electrostatic attraction processes. Ion exchange was responsible for 77.90% of the adsorption process and 80.16% of the desorption process. The MSZ was capable of continuously removing large amounts of ammonium from wastewater through fixed bed adsorption. After 5 regeneration cycles, MSZ still maintained 75% adsorption characteristics for ammonium. Using MSZ adsorbed with ammonium as a soil amendment increased the germination rate of mung beans by 10%. Furthermore, it also increased the stem length, root length, and fresh weight by 20-30%. These findings suggest that MSZ provides a promising application prospect to mitigate ammonium pollution and recycle ammonium resources.

Distinctive tumorigenic significance and innovative oncology targets of SUMOylation.

Protein SUMOylation, a post-translational modification, intricately regulates diverse biological processes including gene expression, cell cycle progression, signaling pathway transduction, DNA damage response, and RNA metabolism. This modification contributes to the acquisition of tumorigenicity and the maintenance of cancer hallmarks. In malignancies, protein SUMOylation is triggered by various cellular stresses, promoting tumor initiation and progression. This augmentation is orchestrated through its specific regulatory mechanisms and characteristic biological functions. This review focuses on elucidating the fundamental regulatory mechanisms and pathological functions of the SUMO pathway in tumor pathogenesis and malignant evolution, with particular emphasis on the tumorigenic potential of SUMOylation. Furthermore, we underscore the potential therapeutic benefits of targeting the SUMO pathway, paving the way for innovative anti-tumor strategies by perturbing this dynamic and reversible modifying process.

Two-dimensionally high AR performance beam scanning utilizing randomly-rotated single-PIN-diode elements for circularly-polarized programmable metasurface.

In this paper, we introduce a novel technique that utilizes randomly rotated elements (RREs) for the cross-polarization and axial ratio (AR) control of a circularly polarized programmable metasurface (CPPMS). We evaluate the CPPMS performance by comparing RREs layout with uniform elements (UEs) layout, and analyze far-field radiation parameters for 50 groups of CPPMS with different RREs layouts. Simulation results demonstrate consistent and improved performance across various RREs layouts, showcasing reduced cross-polarization and enhanced AR beamwidth. To validate these findings, we design a 1-bit CPPMS in Ku-band comprising 20 × 20 elements with the optimal RREs layout, and conduct measurements in an anechoic chamber. The CPPMS prototype achieves high gain (22.34 dBi), low cross-polarization (-20.5 dB), and a narrow 3 dB AR beamwidth (8.93°). Notably, it offers wide-angle beam scanning capabilities of up to ±60°. The gain bandwidth at -3 dB ranges from 14.54 to 16.65 GHz, with a relative bandwidth of 7.3%, while the 3 dB AR bandwidth extends from 14.24 to 16.07 GHz. Consequently, the proposed 1-bit CPPMS exhibits high-performance two-dimensional AR beam scanning, presenting promising applications in satellite communications.

Effect of RAS and BRAF mutations on peritoneal metastasis risk and cytoreductive surgery/hyperthermic intraperitoneal chemotherapy efficacy in colorectal cancer: A systematic review and meta-analysis.

BACKGROUND: Colorectal cancer (CRC) patients with peritoneal metastasis (CRC-PM) have a worse prognosis than those with liver and lung metastases. Cytoreductive surgery (CRS) followed by hyperthermic intraperitoneal chemotherapy (HIPEC) is an effective locoregional treatment for CRC-PM. To date, the prognostic analysis of CRS/HIPEC mostly focuses on clinical and pathological characteristics; however, genetic characteristics, such as RAS/BRAF mutation status, are not sufficient. This study aimed to systematically assess the correlation between RAS/BRAF status and PM risk, as well as the prognostic efficacy of CRS/HIPEC for CRC. METHOD: This study was written in accordance with the 2020 guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols. We searched PubMed, EMBASE, and the Cochrane library with the following keywords: "Peritoneal Neoplasms," "raf Kinases" and "ras Proteins". The fixed-effects model and inverse variance method were used for analysis. Odds ratios (OR) and 95 % confidence intervals (CI) were used to reflect the risk of PM associated with RAS/BRAF mutations. Hazard ratios (HR) and 95 % CI were used to evaluate the effects of RAS/BRAF mutations on the prognosis of CRS/HIPEC. RESULT: Eighteen articles included 5567 patients. In the risk analysis of PM, patients with BRAF mutation were more likely to have PM than those with wild-type BRAF (OR = 2.28, 95 % CI = 1.73-3.01, P < 0.001, I2 = 0 %). In contrast, there was no significant difference in the effect of RAS mutation and wild-type on PM of CRC (OR = 1.28, 95 % CI = 0.99-1.66, P = .06, I2 = 0 %). In a prognostic analysis of CRS/HIPEC, RAS mutation predicted poor overall survival (HR = 1.68, 95 % CI = 1.39-2.02, P < 0.001, I2 = 1 %) and disease-free survival (HR = 1.61, 95 % CI = 1.34-1.94, P < 0.001, I2 = 42 %). The results for BRAF mutation was consistent with the prognostic impact of RAS mutation's overall survival (HR = 2.57, 95 % CI = 1.93-3.44, P < 0.001, I2 = 0 %) and disease-free survival (HR = 1.90, 95 % CI = 1.40-2.56, P < 0.001, I2 = 82 %). CONCLUSION: BRAF mutation, rather than RAS mutation, was a high-risk factor for CRC-PM. And both BRAF and RAS mutations negatively affected the prognosis of CRS/HIPEC in CRC-PM patients. Our results could provide suggestions for the selection of comprehensive treatment for CRC-PM with RAS/BRAF mutations.

Macrophage and fibroblast trajectory inference and crosstalk analysis during myocardial infarction using integrated single-cell transcriptomic datasets.

BACKGROUND: Cardiac fibrosis after myocardial infarction (MI) has been considered an important part of cardiac pathological remodeling. Immune cells, especially macrophages, are thought to be involved in the process of fibrosis and constitute a niche with fibroblasts to promote fibrosis. However, the diversity and variability of fibroblasts and macrophages make it difficult to accurately depict interconnections. METHODS: We collected and reanalyzed scRNA-seq and snRNA-seq datasets from 12 different studies. Differentiation trajectories of these subpopulations after MI injury were analyzed by using scVelo, PAGA and Slingshot. We used CellphoneDB and NicheNet to infer fibroblast-macrophage interactions. Tissue immunofluorescence staining and in vitro experiments were used to validate our findings. RESULTS: We discovered two subsets of ECM-producing fibroblasts, reparative cardiac fibroblasts (RCFs) and matrifibrocytes, which appeared at different times after MI and exhibited different transcriptional profiles. We also observed that CTHRC1+ fibroblasts represent an activated fibroblast in chronic disease states. We identified a macrophage subset expressing the genes signature of SAMs conserved in both human and mouse hearts. Meanwhile, the SPP1hi macrophages were predominantly found in the early stages after MI, and cell communication analysis indicated that SPP1hi macrophage-RCFs interactions are mainly involved in collagen deposition and scar formation. CONCLUSIONS: Overall, this study comprehensively analyzed the dynamics of fibroblast and macrophage subsets after MI and identified specific subsets of fibroblasts and macrophages involved in scar formation and collagen deposition.

System-wide identification of novel de-ubiquitination targets for USP10 in gastric cancer metastasis through multi-omics screening.

OBJECTIVE: Ubiquitin-specific peptidase 10 (USP10), a typical de-ubiquitinase, has been found to play a double-edged role in human cancers. Previously, we reported that the expression of USP10 was negatively correlated with the depth of gastric wall invasion, lymph node metastasis, and prognosis in gastric cancer (GC) patients. However, it remains unclear whether USP10 can regulate the metastasis of GC cells through its de-ubiquitination function. METHODS: In this study, proteome, ubiquitinome, and transcriptome analyses were conducted to comprehensively identify novel de-ubiquitination targets for USP10 in GC cells. Subsequently, a series of validation experiments, including in vitro cell culture studies, in vivo metastatic tumor models, and clinical sample analyses, were performed to elucidate the regulatory mechanism of USP10 and its de-ubiquitination targets in GC metastasis. RESULTS: After overexpression of USP10 in GC cells, 146 proteins, 489 ubiquitin sites, and 61 mRNAs exhibited differential expression. By integrating the results of multi-omics, we ultimately screened 9 potential substrates of USP10, including TNFRSF10B, SLC2A3, CD44, CSTF2, RPS27, TPD52, GPS1, RNF185, and MED16. Among them, TNFRSF10B was further verified as a direct de-ubiquitination target for USP10 by Co-IP and protein stabilization assays. The dysregulation of USP10 or TNFRSF10B affected the migration and invasion of GC cells in vitro and in vivo models. Molecular mechanism studies showed that USP10 inhibited the epithelial-mesenchymal transition (EMT) process by increasing the stability of TNFRSF10B protein, thereby regulating the migration and invasion of GC cells. Finally, the retrospective clinical sample studies demonstrated that the downregulation of TNFRSF10B expression was associated with poor survival among 4 of 7 GC cohorts, and the expression of TNFRSF10B protein was significantly negatively correlated with the incidence of distant metastasis, diffuse type, and poorly cohesive carcinoma. CONCLUSIONS: Our study established a high-throughput strategy for screening de-ubiquitination targets for USP10 and further confirmed that inhibiting the ubiquitination of TNFRSF10B might be a promising therapeutic strategy for GC metastasis.

Repair spinal cord injury with a versatile anti-oxidant and neural regenerative nanoplatform.

Spinal cord injury (SCI) often results in motor and sensory deficits, or even paralysis. Due to the role of the cascade reaction, the effect of excessive reactive oxygen species (ROS) in the early and middle stages of SCI severely damage neurons, and most antioxidants cannot consistently eliminate ROS at non-toxic doses, which leads to a huge compromise in antioxidant treatment of SCI. Selenium nanoparticles (SeNPs) have excellent ROS scavenging bioactivity, but the toxicity control problem limits the therapeutic window. Here, we propose a synergistic therapeutic strategy of SeNPs encapsulated by ZIF-8 (SeNPs@ZIF-8) to obtain synergistic ROS scavenging activity. Three different spatial structures of SeNPs@ZIF-8 were synthesized and coated with ferrostatin-1, a ferroptosis inhibitor (FSZ NPs), to achieve enhanced anti-oxidant and anti-ferroptosis activity without toxicity. FSZ NPs promoted the maintenance of mitochondrial homeostasis, thereby regulating the expression of inflammatory factors and promoting the polarization of macrophages into M2 phenotype. In addition, the FSZ NPs presented strong abilities to promote neuronal maturation and axon growth through activating the WNT4-dependent pathways, while prevented glial scar formation. The current study demonstrates the powerful and versatile bioactive functions of FSZ NPs for SCI treatment and offers inspiration for other neural injury diseases.

The role and molecular mechanism of CTHRC1 in fibrosis.

Fibrosis, a pathological state characterized by the excessive accumulation of extracellular matrix components, is primarily driven by the overactivation of fibroblasts. This condition becomes particularly pronounced under chronic inflammatory conditions. Fibrosis can occur in several tissues throughout the body. Among the notable discoveries in the study of fibrosis is the role of Collagen Triple Helix Repeat Containing-1 (CTHRC1), a protein that has emerged as a critical regulator in the fibrotic process. CTHRC1 is rapidly expressed on the outer membrane of fibroblasts and intimal smooth muscle cells following vascular injury, such as that induced by balloon angioplasty. This expression denotes the organism efforts to repair and restructure compromised tissue, signifying a critical component of the tissue repair mechanism in reaction to fibrosis. It plays a pivotal role in promoting cell migration and aiding tissue repair post-injury, contributing significantly to various pathophysiological processes including revascularization, bone formation, developmental morphological changes, inflammatory arthritis, and the progression of cancer. Significantly, researchers have observed marked expression of CTHRC1 across a variety of fibrotic conditions, closely associating it with the progression of the disease. Intervention with CTHRC1 can affect the occurrence and progression of fibrosis. This review aims to comprehensively explore the role and underlying mechanisms of CTHRC1 in fibrotic diseases, highlighting its potential as a key target for therapeutic interventions.

Human dental pulp stem cells mitigate the neuropathology and cognitive decline via AKT-GSK3ß-Nrf2 pathways in Alzheimer's disease.

Oxidative stress is increasingly recognized as a major contributor to the pathophysiology of Alzheimer's disease (AD), particularly in the early stages of the disease. The multiplicity advantages of stem cell transplantation make it fascinating therapeutic strategy for many neurodegenerative diseases. We herein demonstrated that human dental pulp stem cells (hDPSCs) mediated oxidative stress improvement and neuroreparative effects in in vitro AD models, playing critical roles in regulating the polarization of hyperreactive microglia cells and the recovery of damaged neurons. Importantly, these therapeutic effects were reflected in 10-month-old 3xTg-AD mice after a single transplantation of hDPSCs, with the treated mice showing significant improvement in cognitive function and neuropathological features. Mechanistically, antioxidant and neuroprotective effects, as well as cognitive enhancements elicited by hDPSCs, were at least partially mediated by Nrf2 nuclear accumulation and downstream antioxidant enzymes expression through the activation of the AKT-GSK3ß-Nrf2 signaling pathway. In conclusion, our findings corroborated the neuroprotective capacity of hDPSCs to reshape the neuropathological microenvironment in both in vitro and in vivo AD models, which may be a tremendous potential therapeutic candidate for Alzheimer's disease.

Multi-Strategy Improved Dung Beetle Optimization Algorithm and Its Applications.

The dung beetle optimization (DBO) algorithm, a swarm intelligence-based metaheuristic, is renowned for its robust optimization capability and fast convergence speed. However, it also suffers from low population diversity, susceptibility to local optima solutions, and unsatisfactory convergence speed when facing complex optimization problems. In response, this paper proposes the multi-strategy improved dung beetle optimization algorithm (MDBO). The core improvements include using Latin hypercube sampling for better population initialization and the introduction of a novel differential variation strategy, termed "Mean Differential Variation", to enhance the algorithm's ability to evade local optima. Moreover, a strategy combining lens imaging reverse learning and dimension-by-dimension optimization was proposed and applied to the current optimal solution. Through comprehensive performance testing on standard benchmark functions from CEC2017 and CEC2020, MDBO demonstrates superior performance in terms of optimization accuracy, stability, and convergence speed compared with other classical metaheuristic optimization algorithms. Additionally, the efficacy of MDBO in addressing complex real-world engineering problems is validated through three representative engineering application scenarios namely extension/compression spring design problems, reducer design problems, and welded beam design problems.

Eco-Friendly Production of Boron Nitride Nanosheets via Deep Eutectic Solvents and Their Application in Enhancing Thermal Conductivity of PVDF Composites.

Boron nitride nanosheets (BNNS) are expected to be ideal fillers because of their high thermal conductivity and excellent electrical insulation. However, it is still an open challenge to produce BNNS on a large scale using ecofriendly solvents. Here, first, we demonstrate an effective liquid exfoliation method for producing BNNS via utilizing deep eutectic solvents (DES) composed of D,L-menthol and various acids with the assistance of ultrasonication. The results show that the BNNSs with sizes of 1-2 µm in width and 6-8 nm in thickness were successfully exfoliated with a DES formulation of D,L-menthol and decanoic acid. Second, the obtained BNNSs were used for fabricating 1,6-hexanediol diacrylate@polydopamine functionalized BNNS (HDDA@BNNSs-PDA) core-shell microspheres via a Pickering emulsion method. Furthermore, these microspheres were incorporated into a polyvinylidene fluoride (PVDF) matrix to construct 3D thermally conductive networks, leading to a substantial enhancement in the thermal conductivity of the resulting composites. Impressively, the composites with only 25 wt % of BNNS loading reach a high thermal conductivity of 3.20 W/m K, which is a 1500% increase over the pure polymer matrix. This work not only provides a significant way for producing BNNSs ecofriendly but also demonstrates a tactic for constructing 3D thermally conductive networks.

Quality-driven deep cross-supervised learning network for semi-supervised medical image segmentation.

Semi-supervised medical image segmentation presents a compelling approach to streamline large-scale image analysis, alleviating annotation burdens while maintaining comparable performance. Despite recent strides in cross-supervised training paradigms, challenges persist in addressing sub-network disagreement and training efficiency and reliability. In response, our paper introduces a novel cross-supervised learning framework, Quality-driven Deep Cross-supervised Learning Network (QDC-Net). QDC-Net incorporates both an evidential sub-network and an vanilla sub-network, leveraging their complementary strengths to effectively handle disagreement. To enable the reliability and efficiency of semi-supervised training, we introduce a real-time quality estimation of the model's segmentation performance and propose a directional cross-training approach through the design of directional weights. We further design a truncated form of sample-wise loss weighting to mitigate the impact of inaccurate predictions and collapsed samples in semi-supervised training. Extensive experiments on LA and Pancreas-CT datasets demonstrate that QDC-Net surpasses other state-of-the-art methods in semi-supervised medical image segmentation. Code release is available at https://github.com/Medsemiseg.