Transformers have quickly shined in the computer vision world since the emergence of Vision Transformers (ViTs). The dominant role of convolutional neural networks (CNNs) seems to be challenged by increasingly effective transformer-based models. Very recently, a couple of advanced convolutional models strike back with large kernels motivated by the local but large attention mechanism, showing appealing performance and efficiency. While one of them, i.e. RepLKNet, impressively manages to scale the kernel size to 31×31 with improved performance, the performance starts to saturate as the kernel size continues growing, compared to the scaling trend of advanced ViTs such as Swin Transformer. In this paper, we explore the possibility of training extreme convolutions larger than 31×31 and test whether the performance gap can be eliminated by strategically enlarging convolutions. This study ends up with a recipe for applying extremely large kernels from the perspective of sparsity, which can smoothly scale up kernels to 61×61 with better performance. Built on this recipe, we propose Sparse Large Kernel Network (SLaK), a pure CNN architecture equipped with 51×51 kernels that can perform on par with or better than state-of-the-art hierarchical Transformers and modern ConvNet architectures like ConvNeXt and RepLKNet, on ImageNet classification as well as typical downstream tasks.Codes: https://github.com/VITA-Group/SLaK * Equal contribution.Preprint. Under review.
No abstract
The polarization of macrophages and its anti-inflammatory and proinflammatory properties play a significant role in host response after implant placement to determine the outcome of osseointegration and long-term survival. In the previous study, we immobilized an antimicrobial peptide, GL13K, onto titanium surfaces to provide immune regulation property. In the herein presented study, we aimed at investigating whether GL13K immobilized titanium surface could improve osteogenesis and reduce the inflammatory reaction around the biomaterials by altering macrophage response. We evaluated the cell proliferation of the different phenotypes of macrophages seeded in GL13K-coated titanium surface, which indicated an inhibition of M1 macrophages and a good cytocompatibility to M2 macrophages. Then, we measured the inflammatory and anti-inflammatory activity of the M1 and M2 macrophages seeded on the GL13K-coated titanium surfaces. The results of the enzyme-linked immunosorbent assay and quantitative reverse transcription-polymerase chain reaction showed that the group with the GL13K modified surface had a downregulation in the expression level of the tumor necrosis factor-α and interleukin-1β in M1 macrophages and an upregulation of IL-10 and transforming growth factor-β3 (TGF-β3) levels in M2 macrophages. This study demonstrated that the GL13K modified titanium surfaces can regulate macrophages’ polarization and the expression of inflammatory and anti-inflammatory effects, reducing the effects of the inflammatory process, which may promote the process of bone regeneration and osseointegration.
Creating a pro-regenerative immune microenvironment around implant biomaterial surfaces is significant to osseointegration. Immune cells, especially macrophages that participate in the osseointegration, including osteogenesis, osteoclastogenesis, and angiogenesis, should be considered when testing biomaterials. In this study, we immobilized an antimicrobial peptide GL13K with immunomodulatory properties onto a titanium surface via silanization. The modified surfaces show good biocompatibility with bone mesenchymal stromal cells (BMSCs), human umbilical vein endothelial cells (HUVECs), and RAW264.7. By co-culturing BMSCs with RAW264.7, we found that the GL13K-coated titanium surfaces could promote late-stage osteogenesis as demonstrated by the upregulated expression of recombinant collagen type I alpha 1 (COL-1α1) and more extracellular matrix mineralization, while the early phase remained unchanged. The surfaces inhibited the osteoclastogenic differentiation of RAW264.7 cells by restraining nuclear factor-activated T cells, cytoplasmic 1 (NFATc1), the main factor of the receptor activator of nuclear factor-κ B, and the receptor activator of the nuclear factor-κ B ligand signaling pathway, from entering the nucleus and further reduced the expression of the activating osteoclastogenic tartrate-resistant acid phosphatase gene. Moreover, the GL13K-coated titanium surface demonstrated significant promotion of angiogenesis differentiation of HUVECs as indicated by the upregulated expression of essential angiogenesis function genes, including hypoxia-inducible factor-1α, endothelial nitric oxide synthase, kinase insert domain receptor, and vascular endothelial growth factor A (HIF-1α, eNOS, KDR, and VEGF-A). Taken together, these results demonstrated that the GL13K coating had properties of osteogenesis, angiogenesis, and anti-osteoclastogenesis via its immunomodulatory potential.
Background Natural products, especially those with high contents of phytochemicals, are promising alternative medicines owing to their antitumor properties and few side effects. In this study, the effects of a plant-based medicinal food (PBMF) composed of six medicinal and edible plants, namely, Coix seed, Lentinula edodes, Asparagus officinalis L., Houttuynia cordata, Dandelion, and Grifola frondosa, on gastric cancer and the underlying molecular mechanisms were investigated in vivo. Methods A subcutaneous xenograft model of gastric cancer was successfully established in nude mice inoculated with SGC-7901 cells. The tumor-bearing mice were separately underwent with particular diets supplemented with three doses of PBMF (43.22, 86.44, and 172.88 g/kg diet) for 30 days. Tumor volumes were recorded. Histopathological changes in and apoptosis of the xenografts were evaluated by hematoxylin and eosin staining and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling staining, respectively. Serum levels of TNF-α, MMP-2, and MMP-9 were detected by enzyme-linked immunosorbent assay. The mRNA expression levels of β-catenin, GSK-3β, E-cadherin, N-cadherin, MMP-2/9, Snail, Bax, Bcl-2, Caspase-3/9, and Cyclin D1 were evaluated via real-time quantitative polymerase chain reaction. The protein expression levels of GSK-3β, E-cadherin, N-cadherin, and Ki-67 were determined by immunohistochemistry staining. Results PBMF treatment efficiently suppressed neoplastic growth, induced apoptosis, and aggravated necrosis in the xenografts of SGC-7901 cells. PBMF treatment significantly decreased the serum levels of MMP-2 and MMP-9 and significantly increased that of TNF-α. Furthermore, PBMF treatment notably upregulated the mRNA expression levels of GSK-3β, E-cadherin, Bax, Caspase-3, and Caspase-9 but substantially downregulated those of β-catenin, N-cadherin, MMP-2, MMP-9, Snail, and Cyclin D1 in tumor tissues. The Bax/Bcl-2 ratio was upregulated at the mRNA level. Moreover, PBMF treatment remarkably increased the protein expression levels of GSK-3β and E-cadherin but notably reduced those of Ki-67 and N-cadherin in tumor tissues. Conclusions The PBMF concocted herein exerts anti-gastric cancer activities via epithelial–mesenchymal transition reversal, apoptosis induction, and proliferation inhibition. The underlying molecular mechanisms likely rely on suppressing the Wnt/β-catenin signaling pathway.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.