Search

This paper presents Diff-ReColor, a novel diffusion-based image colorization framework that addresses the challenges of colorizing grayscale images with high semantic fidelity and diversity. Recognizing the limitations of previous approaches, including CNN-based methods and GANs, which often resulted in ambiguous colorization and artifacts, we leverage the recent advancements in diffusion models known for their stable training and diverse output generation. Diff-ReColor integrates an edge-conditional Denoising Diffusion Probabilistic Model (DDPM) with Dual-Tier Attention Color Reference Encoder, and a segmentation-informed sampling technique to produce colorized images that are both semantically consistent and rich in detail. Our edge-conditional DDPM is specifically designed to handle intricate image details, while the segmentation-guided sampling technique ensures the retention of semantic nuances by utilizing edge and segmentation cues during the colorization process. Inputs are also channeled into a dual attention encoder, where spectral-attention filters semantic information enhanced by spatial-attention, acquiring coarse coloring images to assist the denoising process. Empirical validation on the ImageNet benchmark dataset demonstrates the superior performance of Diff-ReColor in terms of color richness, colorization accuracy, and semantic fidelity. The model's generalization capabilities are further highlighted through its performance on COCO-Stuff and CelebA-HQ datasets, where it surpasses the baseline models without the need for fine-tuning. [ABSTRACT FROM AUTHOR]

This paper introduces TrpViT, a novel triple attention vision transformer that efficiently captures both local and global features. The proposed architecture tackles global information acquisition by employing three complementary attention mechanisms in a unique attention block: Window, Dilated, and Channel attention. This attention block extracts spatially local features while expanding the receptive field to capture richer global context. By integrating this attention block with convolution, a new C-C-T-T architecture is formed. We rigorously evaluate TrpViT, demonstrating state-of-the-art performance on various computer vision tasks, including image classification, 2D and 3D object detection, instance segmentation, and low-level image colorization. Notably, TrpViT achieves strong accuracy across all parameter scales, highlighting its computational efficiency and effectiveness. • A Triple Attention Vision Transformer captures both global and local features. • TrpViT integrates convolution into transformer to provide induction bias. • TrpViT compensates non-local information using multiple complementary attentions. • TrpViT achieves state-of-the-art results across both high-level and low-level tasks. [ABSTRACT FROM AUTHOR]

Opioids and α2-agonists have been used as epidural adjuvants in local anesthetics for a long time, but the effect of the combination of opioids and α2-agonists as epidural adjuvants is not completely understood. In the present study, the combination of dexmedetomidine (Dex) and sufentanil as adjuvants to ropivacaine for epidural labor analgesia was investigated. A total of 108 parturient women receiving labor epidural analgesia were randomly divided into three groups: i) Group RD received 0.1% ropivacaine + 0.5 µg/ml Dex; ii) Group RS received 0.1% ropivacaine + 0.5 µg/ml sufentanil; and iv) Group RDS received 0.1% ropivacaine + 0.25 µg/ml Dex + 0.25 µg/ml sufentanil. Patients received a 10 ml loading dose followed by a maintenance by patient controlled epidural analgesia. The visual analog scale scores, onset time, local anesthetic requirements, motor blockage and adverse effects were recorded. Group RDS displayed an improved labor analgesia effect compared with Groups RD and RS. Group RDS displayed a shorter onset time compared with Groups RD and RS, and a reduced local anesthetic requirement compared with Group RS. The motor blockage in Groups RDS and RS was significantly lower compared with Group RD, and the incidence of pruritus in Groups RDS and RD was lower compared with Group RS. In conclusion, the combined use of 0.25 µg/ml Dex and 0.25 µg/ml sufentanil as adjuvants to 0.1% ropivacaine for epidural labor analgesia displayed an improved analgesia effect compared with the use of either 0.5 µg/ml sufentanil or 0.5 µg/ml Dex alone. The present study was registered with the Chinese Clinical Trial Registry Center on 23 February, 2018 (registration no. ChiCTR-IOR-1800014943). [ABSTRACT FROM AUTHOR]

The correction of exposure-related issues is a pivotal component in enhancing the quality of images, offering substantial implications for various computer vision tasks. Historically, most methodologies have predominantly utilized spatial domain recovery, offering limited consideration to the potentialities of the frequency domain. Additionally, there has been a lack of a unified perspective towards low-light enhancement, exposure correction, and multi-exposure fusion, complicating and impeding the optimization of image processing. In response to these challenges, this paper proposes a novel methodology that leverages the frequency domain to improve and unify the handling of exposure correction tasks. Our method introduces Holistic Frequency Attention and Dynamic Frequency Feed-Forward Network, which replace conventional correlation computation in the spatial-domain. They form a foundational building block that facilitates a U-shaped Holistic Dynamic Frequency Transformer as a filter to extract global information and dynamically select important frequency bands for image restoration. Complementing this, we employ a Laplacian pyramid to decompose images into distinct frequency bands, followed by multiple restorers, each tuned to recover specific frequency-band information. The pyramid fusion allows a more detailed and nuanced image restoration process. Ultimately, our structure unifies the three tasks of low-light enhancement, exposure correction, and multi-exposure fusion, enabling comprehensive treatment of all classical exposure errors. Benchmarking on mainstream datasets for these tasks, our proposed method achieves state-of-the-art results, paving the way for more sophisticated and unified solutions in exposure correction. • A holistic frequency attention to enhance image detail and color. • A dynamic frequency feed-forward network to dynamically filter crucial bands. • A Laplacian pyramid decomposition approach to refine frequency synthesis. • An integrated architecture to tackle exposure correction and image fusion tasks. [ABSTRACT FROM AUTHOR]

Pancreatic cancer exhibits a high degree of malignancy with a poor prognosis, lacking effective prognostic targets. Utilizing histopathological methodologies, this study endeavors to predict the expression of pathological features in pancreatic ductal adenocarcinoma (PAAD) and investigate their underlying molecular mechanisms. Pathological images, transcriptomic, and clinical data from TCGA-PAAD were collected for survival analysis. Image segmentation using unsupervised machine learning was employed to extract features, perform clustering, and establish models. The prognostic value of pathological features and associated clinical risk factors were evaluated; the correlation between pathological features and molecular mechanisms, gene mutations, and immune infiltration was analyzed. By clustering 45 effective pathological features, we divided PAAD patients into two groups: cluster 1 and cluster 2. Significant associations with poor prognosis were found for cluster 2 in both the training group (n = 113) and validation group (n = 75) (p = 0.006), with pathological stages II-IV identified as potential synergistic risk factors (HR = 2.421, 95% CI = 1.263–4.639, p = 0.008). Subsequently, through multi-omics correlation analysis, we further revealed a close association between cluster 2 and the oxidative phosphorylation mechanism. Within the cluster 2 group, 28 oxidative phosphorylation genes exhibited reduced expression, CDKN2A gene mutations were upregulated, and there was significant downregulation of Tregs infiltration and related immune gene expression. The pathomic model constructed using machine learning serves as a valuable prognostic target for PAAD. The histopathological features cluster 2 are closely associated with the downregulation of oxidative phosphorylation levels and Tregs immune infiltration. [ABSTRACT FROM AUTHOR]

Pancreatic cancer exhibits a high degree of malignancy with a poor prognosis, lacking effective prognostic targets. Utilizing histopathological methodologies, this study endeavors to predict the expression of pathological features in pancreatic ductal adenocarcinoma (PAAD) and investigate their underlying molecular mechanisms. Pathological images, transcriptomic, and clinical data from TCGA-PAAD were collected for survival analysis. Image segmentation using unsupervised machine learning was employed to extract features, perform clustering, and establish models. The prognostic value of pathological features and associated clinical risk factors were evaluated; the correlation between pathological features and molecular mechanisms, gene mutations, and immune infiltration was analyzed. By clustering 45 effective pathological features, we divided PAAD patients into two groups: cluster 1 and cluster 2. Significant associations with poor prognosis were found for cluster 2 in both the training group (n = 113) and validation group (n = 75) (p = 0.006), with pathological stages II-IV identified as potential synergistic risk factors (HR = 2.421, 95% CI = 1.263–4.639, p = 0.008). Subsequently, through multi-omics correlation analysis, we further revealed a close association between cluster 2 and the oxidative phosphorylation mechanism. Within the cluster 2 group, 28 oxidative phosphorylation genes exhibited reduced expression, CDKN2A gene mutations were upregulated, and there was significant downregulation of Tregs infiltration and related immune gene expression. The pathomic model constructed using machine learning serves as a valuable prognostic target for PAAD. The histopathological features cluster 2 are closely associated with the downregulation of oxidative phosphorylation levels and Tregs immune infiltration. [ABSTRACT FROM AUTHOR]

Previous studies have shown that microRNAs (miRs), such as miR-146a play an important role in the pathogenesis of intestinal ischemia/reperfusion (I/R)-induced injury; however, the role of miR-146a in intestinal I/R-induced acute lung injury has not been elucidated. An intestinal I/R-induced injury mouse model was established in the present study by clamping the superior mesenteric artery and expression levels of miR-146a in intestinal and lung tissue samples were evaluated using reverse transcription-quantitative PCR (RT-qPCR). Intestinal and lung histopathological characteristics in mice with intestinal I/R-induced injury were assessed by hematoxylin and eosin staining, and mRNA and protein expression levels in intestinal and lung tissue samples were evaluated using RT-qPCR and western blotting, respectively. miR-146a expression was significantly downregulated in the intestinal and lung tissue samples of mice with intestinal I/R-induced injury. Intestinal I/R injury-induced histopathological changes in the lung and intestines, and pulmonary edema in mice transduced with an adenoviral miR-146a-overexpression vector (the miR-146a overexpression group) were alleviated. mRNA expression levels of TNF-α, IL-1β, IFN-γ and TGF-β1, and protein expression levels of TNF receptor-associated factor 6, phosphorylated-p65 NF-κB, cleaved caspase-3 and cleaved caspase-9 in lung and intestinal tissue samples were downregulated in I/R-miR-146a-overexpressing mice, compared with those from the I/R-negative control group. Thus, the present study identified that pre-treatment with the miR-146a overexpression vector alleviated intestinal I/R-induced acute lung injury in mice. [ABSTRACT FROM AUTHOR]

Background: To study the incidence and risk factors of shivering in pregnant women during cesarean section. Methods: We performed a prospective nested case-control study involving parturients scheduled for cesarean sections between July 2018 and May 2021. The overall incidence of intraoperative shivering and its potential risk factors were investigated. The potential risk factors evaluated were pain, anxiety, emergency surgery, transfer from the delivery room, epidural labor analgesia, membrane rupture, labor, and the timing of the surgery. Shivering and body temperature at different time points during the cesarean section were also recorded. The selected seven time points were: entering the operating room, post-anesthesia, post-disinfection, post-delivery, post-oxytocin, post additional hysterotonics, and before leaving the operating room. Results: We analyzed 212 cesarean section parturients. The overall incidence of shivering was 89 (42.0%). Multivariate logistic regression showed that anxiety, emergency delivery, and transfer from the delivery room to the operating room increased the overall shivering incidence (odds ratio = 1.77, 2.90, and 3.83, respectively). The peak shivering incidence occurred after skin disinfection (63, 29.7%), and the lowest body temperature occurred after oxytocin treatment (36.24 ± 0.30 °C). Stratified analysis of surgery origin showed that emergency delivery was a risk factor for shivering (odds ratio = 2.99) in women transferred from the obstetric ward to the operating room. Conclusion: Shivering occurred frequently during cesarean sections, with the peak incidence occurring after skin disinfection. Anxiety, emergency delivery, and transfer from the delivery room to the operating room increased the risk of shivering development during cesarean sections. Trial registration: The study protocol was registered online at China Clinical Registration Center (registration number: ChiCTR-ROC-17010532, Registered on 29 January 2017). [ABSTRACT FROM AUTHOR]

[Display omitted] • Cotton fiber-reinforced chitosan/carbon nanofiber composites with varying epoxy and CNF contents (CS/CNF/EP) were prepared through a straightforward low-temperature curing process (40 °C, 4 h). • The CS/CNF/EP composites provided a rapid alarm (~1 s) and sustained a 263 s warning under flame. • The CS/CNF/EP composites provided a thermal response warning activation within 1s at 160 °C. • Flame-retardant mechanisms and fire warning behaviors were clarified. Enhancing fire safety performance in resin-based composites is critical for mitigating fire hazards. Composites that provide early warnings during initial fire stages and maintain prolonged alarm capabilities under flame exposure are essential for minimizing injuries and property damage. This study employs chitosan (CS) as a cross-linking and char-forming agent, while carbon nanofiber (CNF) acts as a flame retardant and conductive component. Cotton fiber-reinforced chitosan/carbon nanofiber composites with varying epoxy and CNF contents (CS/CNF/EP) were prepared through a straightforward low-temperature curing process (40 °C, 4 h). Crosslinking polymerization between chitosan and epoxy was confirmed via Fourier Transform Infrared Spectroscopy (FTIR). The presence of the tough structure of CS and the rigid structure of EP results in a synergistic toughening and reinforcement effect in the CS/CNF/EP composites. Additionally, due to the carbon facilitating properties of CS and the enhanced conductivity of CNF, the composites exhibit a sensitive fire alarm functionality. At 35 % epoxy content, the CS/CNF/EP composites demonstrated exceptional properties, including mechanical flexibility (tensile strength of 7.09 MPa and breaking elongation of 129.05 %) and outstanding flame retardancy (LOI of 32.3 % and UL-94 V-0 rating). These composites provided a rapid alarm (∼1 s), sustaining a 263 s warning under flame exposure and over 35 min post-flame removal. These findings indicate that CS/CNF/EP composites, integrating mechanical flexibility, flame retardancy, and superior fire warning properties, present significant potential applications in fire safety engineering. [ABSTRACT FROM AUTHOR]

This paper proposes a hybrid demand response (DR) mechanism considering three types of participants: power grid operator (PGO), retailers and end users. Different from the traditional price-based or incentive-based methods, this hybrid mechanism combines real-time pricing and real-time incentive together to implement demand response programs dispatched by PGO, i.e., the PGO provides incentives to retailers and the retailers set optimal real-time prices to users every 5 min. This hybrid DR mechanism can better motivate retailers to participate by providing them with monetary incentives from PGO for load shifting. We use a three-level Stackelberg game to model the proposed mechanism. The PGO first determines the optimal incentive rate to minimize its cost, then the retailers decide the optimal electricity price to maximize their profits, and the users finally choose the optimal power demand to maximize their welfare. The analytical solutions of the optimal decisions for every participant are given. We also propose a distributed algorithm to implement this mechanism in a practical application by considering information asymmetry. The simulation results verify its advantages over traditional demand response mechanisms. • A novel hybrid demand response (DR) mechanism combining real-time pricing and real-time incentive is proposed. • The hybrid DR mechanism is modeled by a three-level Stackelberg game. • The existence and uniqueness of the solution at game equilibrium is proved and analyzed. • A distributed algorithm is proposed to implement hybrid DR under information asymmetry. • The benefits of the power grid, retailers and users are improved under hybrid DR. [ABSTRACT FROM AUTHOR]