A multi-scale feature fusion neural network for multi-class disease classification on the maize leaf images.

Maize is a globally important cereal crop, however, maize leaf disease is one of the most common and devastating diseases that afflict it. Artificial intelligence methods face challenges in identifying and classifying maize leaf diseases due to variations in image quality, similarity among diseases, disease severity, limited dataset availability, and limited interpretability. To address these challenges, we propose a residual-based multi-scale network (MResNet) for classifying multi-type maize leaf diseases from maize images. MResNet consists of two residual subnets with different scales, enabling the model to detect diseases in maize leaf images at different scales. We further utilize a hybrid feature weight optimization method to optimize and fuse the feature mapping weights of two subnets. We validate MResNet on a maize leaf diseases dataset. MResNet achieves 97.45% accuracy. The performance of MResNet surpasses other state-of-the-art methods. Various experiments and two additional datasets confirm the generalization performance of our model. Furthermore, thermodynamic diagram analysis increases the interpretability of the model. This study provides technical support for the disease classification of agricultural plants.

Macrophage-derived ectosomal miR-350-3p promotes osteoarthritis progression through downregulating chondrocyte H3K36 methyltransferase NSD1.

Mechanical overloading can promote cartilage senescence and osteoarthritis (OA) development, but its impact on synovial macrophages and the interaction between macrophages and chondrocytes remain unknown. Here, we found that macrophages exhibited M1 polarization under mechanical overloading and secreted ectosomes that induced cartilage degradation and senescence. By performing miRNA sequencing on ectosomes, we identified highly expressed miR-350-3p as a key factor mediating the homeostatic imbalance of chondrocytes caused by M1-polarized macrophages, this result being confirmed by altering the miR-350-3p level in chondrocytes with mimics and inhibitor. In experimental OA mice, miR-350-3p was increased in synovium and cartilage, while intra-articular injection of antagomir-350-3p inhibited the increase of miR-350-3p and alleviated cartilage degeneration and senescence. Further studies showed that macrophage-derived ectosomal miR-350-3p promoted OA progression by inhibiting nuclear receptor binding SET domain protein 1(NSD1) in chondrocytes and regulating histone H3 lysine 36(H3K36) methylation. This study demonstrated that the targeting of macrophage-derived ectosomal miRNAs was a potential therapeutic method for mechanical overload-induced OA.

Joint global and local interpretation method for CIN status classification in breast cancer.

Breast cancer is among the cancer types with the highest numbers of new cases. The study of this disease from a microscopic perspective has been a prominent research topic. Previous studies have shown that microRNAs (miRNAs) are closely linked to chromosomal instability (CIN). Correctly predicting CIN status from miRNAs can help to improve the survival of breast cancer patients. In this study, a joint global and local interpretation method called GL_XGBoost is proposed for predicting CIN status in breast cancer. GL_XGBoost integrates the eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanation (SHAP) methods. XGBoost is used to predict CIN status from miRNA data, whereas SHAP is used to select miRNA features that have strong relationships with CIN. Furthermore, SHAP's rich visualization strategies enhance the interpretability of the entire model at the global and local levels. The performance of GL_XGBoost is validated on the TCGA-BRCA dataset, and it is shown to have an accuracy of 78.57% and an area under the curve value of 0.87. Rich visual analysis is used to explain the relationships between miRNAs and CIN status from different perspectives. Our study demonstrates an intuitive way of exploring the relationship between CIN and cancer from a microscopic perspective.

RPL35 downregulated by mechanical overloading promotes chondrocyte senescence and osteoarthritis development via Hedgehog-Gli1 signaling.

Objectives: To investigate the potential role of Ribosomal protein L35 (RPL35) in regulating chondrocyte catabolic metabolism and to examine whether osteoarthritis (OA) progression can be delayed by overexpressing RPL35 in a mouse compression loading model. Methods: RNA sequencing analysis was performed on chondrocytes treated with or without 20 % elongation strain loading for 24 h. Experimental OA in mice was induced by destabilization of the medial meniscus and compression loading. Mice were randomly assigned to a sham group, an intra-articular adenovirus-mediated overexpression of the negative group, and an intra-articular adenovirus-mediated overexpression of the RPL35 operated group. The Osteoarthritis Research Society International score was used to evaluate cartilage degeneration. Immunostaining and western blot analyses were conducted to detect relative protein levels. Primary mouse chondrocytes were treated with 20 % elongation strain loading for 24 h to investigate the role of RPL35 in modulating chondrocyte catabolic metabolism and regulating cellular senescence in chondrocytes. Results: The protein expression of RPL35 in mouse chondrocytes was significantly reduced when excessive mechanical loading was applied, while elevated protein levels of RPL35 protected articular chondrocytes from degeneration. In addition, the RPL35 knockdown alone induced chondrocyte senescence, decreased the expression of anabolic markers, and increased the expression of catabolic markers in vitro in part through the hedgehog (Hh) pathway. Conclusions: These findings demonstrated a functional pathway important for OA development and identified intra-articular injection of RPL35 as a potential therapy for OA prevention and treatment. The translational potential of this article: It is necessary to develop new targeted drugs for OA due to the limitations of conventional pharmacotherapy. Our study explores and demonstrates the protective effect of RPL35 against excessive mechanical stress in OA models in vivo and in vitro in animals. These findings might provide novel insights into OA pathogenesis and show its translational potential for OA therapy.

Co single atom coupled oxygen vacancy on W18O49 nanowires surface to construct asymmetric active site enhanced peroxymonosulfate activation.

Enhancing the activation of peroxymonosulfate (PMS) is essential for generating more reactive oxygen species in advanced oxidation process (AOPs). Nevertheless, improving PMS adsorption and expediting interfacial electron transfer to enhance reaction kinetics pose significant challenges. Herein, we construct confined W18O49 nanowires with asymmetric active centers containing Co-Vo-W (Vo: oxygen vacancy). The design incorporates surface-rich Vo and single-atom Co, and the resulting material is employed for PMS activation in water purification. By coupling unsaturated coordinated electrons in Vo with low-valence Co single atoms to construct an the "electron fountainhead", the adsorption and activation of PMS are enhanced. This results in the generation of more active free radicals (SO4•-, •OH, •O2-) and non-free radicals (1O2) for the decomposition of micropollutants. Thereinto, the degradation rate of bisphenol A (BPA) by Co-W18O49 is 32.6 times faster that of W18O49 monomer, which is also much higher than those of other transition-metal-doped W18O49 composites. This work is expected to help to elucidate the rational design and efficient PMS activation of catalysts with asymmetric active centers.

Clinically-observed FOXA1 mutations upregulate SEMA3C through transcriptional derepression in prostate cancer.

FOXA1 is a pioneer transcription factor that is frequently mutated in prostate, breast, bladder, and salivary gland malignancies. Indeed, metastatic castration-resistant prostate cancer (mCRPC) commonly harbour FOXA1 mutations with a prevalence of 35%. However, despite the frequent recurrence of FOXA1 mutations in prostate cancer, the mechanisms by which FOXA1 variants drive its oncogenic effects are still unclear. Semaphorin 3C (SEMA3C) is a secreted autocrine growth factor that drives growth and treatment resistance of prostate and other cancers and is known to be regulated by both AR and FOXA1. In the present study, we characterize FOXA1 alterations with respect to its regulation of SEMA3C. Our findings reveal that FOXA1 alterations lead to elevated levels of SEMA3C both in prostate cancer specimens and in vitro. We further show that FOXA1 negatively regulates SEMA3C via intronic cis elements, and that mutations in FOXA1 forkhead domain attenuate its inhibitory function in reporter assays, presumably by disrupting DNA binding of FOXA1. Our findings underscore the key role of FOXA1 in prostate cancer progression and treatment resistance by regulating SEMA3C expression and suggest that SEMA3C may be a driver of growth and tumor vulnerability of mCRPC harboring FOXA1 alterations.

A Coordination-Derived Cerium-Based Amorphous-Crystalline Heterostructure with High Electrocatalytic Oxygen Evolution Activity.

The rational design of heterogeneous catalysts is crucial for achieving optimal physicochemical properties and high electrochemical activity. However, the development of new amorphous-crystalline heterostructures is significantly more challenging than that of the existing crystalline-crystalline heterostructures. To overcome these issues, a coordination-assisted strategy that can help fabricate an amorphous NiO/crystalline NiCeOx (a-NiO/c-NiCeOx ) heterostructure is reported herein. The coordination geometry of the organic ligands plays a pivotal role in permitting the formation of coordination polymers with high Ni contents. This consequently provides an opportunity for enabling the supersaturation of Ni in the NiCeOx structure during annealing, leading to the endogenous spillover of Ni from the depths of NiCeOx to its surface. The resulting heterostructure, featuring strongly coupled amorphous NiO and crystalline NiCeOx , exhibits harmonious interactions in addition to low overpotentials and high catalytic stability in the oxygen evolution reaction (OER). Theoretical calculations prove that the amorphous-crystalline interfaces facilitate charge transfer, which plays a critical role in regulating the local electron density of the Ni sites, thereby promoting the adsorption of oxygen-based intermediates on the Ni sites and lowering the dissociation-related energy barriers. Overall, this study underscores the potential of coordinating different metal ions at the molecular level to advance amorphous-crystalline heterostructure design.

Cloning, Expression, and Functional Analysis of the MYB Transcription Factor SlMYB86-like in Tomato.

MYB transcription factors (TFs) have been shown to play a key role in plant growth and development and are in response to various types of biotic and abiotic stress. Here, we clarified the structure, expression patterns, and function of a MYB TF, SlMYB86-like (Solyc06g071690) in tomato using an inbred tomato line exhibiting high resistance to bacterial wilt (Hm 2-2 (R)) and one susceptible line (BY 1-2 (S)). The full-length cDNA sequence of this gene was 1226 bp, and the open reading frame was 966 bp, which encoded 321 amino acids; its relative molecular weight was 37.05055 kDa; its theoretical isoelectric point was 7.22; it was a hydrophilic nonsecreted protein; and it had no transmembrane structures. The protein also contains a highly conserved MYB DNA-binding domain and was predicted to be localized to the nucleus. Phylogenetic analysis revealed that SlMYB86-like is closely related to SpMYB86-like in Solanum pennellii and clustered with other members of the family Solanaceae. Quantitative real-time PCR (qRT-PCR) analysis revealed that the expression of the SlMYB86-like gene was tissue specific and could be induced by Ralstonia solanacearum, salicylic acid, and jasmonic acid. The results of virus-induced gene silencing (VIGS) revealed that SlMYB86-like silencing decreased the resistance of tomato plants to bacterial wilt, suggesting that it positively regulates the resistance of tomatoes to bacterial wilt. Overall, these findings indicate that SlMYB86-like plays a key role in regulating the resistance of tomatoes to bacterial wilt.

Collaborative Transfer Network for Multi-Classification of Breast Cancer Histopathological Images.

The incidence of breast cancer is increasing rapidly around the world. Accurate classification of the breast cancer subtype from hematoxylin and eosin images is the key to improve the precision of treatment. However, the high consistency of disease subtypes and uneven distribution of cancer cells seriously affect the performance of multi-classification methods. Furthermore, it is difficult to apply existing classification methods to multiple datasets. In this article, we propose a collaborative transfer network (CTransNet) for multi-classification of breast cancer histopathological images. CTransNet consists of a transfer learning backbone branch, a residual collaborative branch, and a feature fusion module. The transfer learning branch adopts the pre-trained DenseNet structure to extract image features from ImageNet. The residual branch extracts target features from pathological images in a collaborative manner. The feature fusion strategy of optimizing these two branches is used to train and fine-tune CTransNet. Experiments show that CTransNet achieves 98.29% classification accuracy on the public BreaKHis breast cancer dataset, exceeding the performance of state-of-the-art methods. Visual analysis is carried out under the guidance of oncologists. Based on the training parameters of the BreaKHis dataset, CTransNet achieves superior performance on other two public breast cancer datasets (breast-cancer-grade-ICT and ICIAR2018_BACH_Challenge), indicating that CTransNet has good generalization performance.

Ultrahigh-Q and angle-robust chiroptical resonances beyond BIC splitting.

Chiroptical resonances inspired by bound states in the continuum (BICs) open a new, to the best of our knowledge, avenue to enhance chiral light-matter interaction. Symmetry breaking is the widely employed way, wherein the circularly polarized states (CPSs) arise from BIC splitting. Here, we utilize a far-field interference mechanism to create ultrahigh-Q (typically, 2.36 × 106) chiroptical resonance beyond BIC splitting, in which CPSs coexist with BICs in the momentum space. Accordingly, the spin-selective absorption with ultranarrow linewidth is achieved at the CPS points, which can be regulated by monolayer transition metal dichalcogenides (TMDCs). In addition, the chiral response of our scheme exhibits the incident-direction robustness and flexible tunability. Our findings may facilitate potential applications in light manipulation, spin-valley interaction, and chiral sensing.

Fructose-bisphosphatase1 (FBP1) alleviates experimental osteoarthritis by regulating Protein crumbs homolog 3 (CRB3).

PURPOSE: To identify the role of gluconeogenesis in chondrocytes in osteoarthritis (OA). MATERIALS AND METHODS: Cartilage samples were collected from OA patients and C57 mice and were stained with Safranin O-Fast Green to determine the severity of OA. Periodic acid Schiff staining was used to characterize the contents of polysaccharides and SA-ßGal staining was used to characterize the aging of chondrocytes. Immunohistochemistry and western blotting were used to detect fructose-bisphosphatase1 (FBP1), SOX9, MMP13, P21, and P16 in cartilage or chondrocyte. The mRNA levels of fbp1, mmp13, sox9, colX, and acan were analyzed by qPCR to evaluate the role of FBP1 in chondrocytes. RESULTS: The level of polysaccharides in cartilage was reduced in OA and the expression of FBP1 was also reduced. We treated the chondrocytes with IL-1ß to cause OA in vitro, and then made chondrocytes overexpress FBP1 with plasma. It shows that FBP1 alleviated the degeneration and senescence of chondrocytes in vitro and that it also showed the same effects in vivo experiments. To further understand the mechanism of FBP1, we screened the downstream protein of FBP1 and found that CRB3 was significantly downregulated. And we confirmed that CRB3 suppressed the degeneration and delayed senescence of chondrocytes. CONCLUSIONS: FBP1 promoted the polysaccharide synthesis in cartilage and alleviated the degeneration of cartilage by regulating CRB3, so FBP1 is a potential target in treating OA.

MGGAN: A multi-generator generative adversarial network for breast cancer immunohistochemical image generation.

The immunohistochemical technique (IHC) is widely used for evaluating diagnostic markers, but it can be expensive to obtain IHC-stained section. Translating the cheap and easily available hematoxylin and eosin (HE) images into IHC images provides a solution to this challenge. In this paper, we propose a multi-generator generative adversarial network (MGGAN) that can generate high-quality IHC images based on the HE of breast cancer. Our MGGAN approach combines the low-frequency and high-frequency components of the HE image to improve the translation of breast cancer image details. We use the multi-generator to extract semantic information and a U-shaped architecture and patch-based discriminator to collect and optimize the low-frequency and high-frequency components of an image. We also include a cross-entropy loss as a regularization term in the loss function to ensure consistency between the synthesized image and the real image. Our experimental and visualization results demonstrate that our method outperforms other state-of-the-art image synthesis methods in terms of both quantitative and qualitative analysis. Our approach provides a cost-effective and efficient solution for obtaining high-quality IHC images.

Oriented Ultrathin π-complexation MOF Membrane for Ethylene/Ethane and Flue Gas Separations.

Rational design and engineering of high-performance molecular sieve membranes towards C2 H4 /C2 H6 and flue gas separations remain a grand challenge to date. In this study, through combining pore micro-environment engineering with meso-structure manipulation, highly c-oriented sub-100 nm-thick Cu@NH2 -MIL-125 membrane was successfully prepared. Coordinatively unsaturated Cu ions immobilized in the NH2 -MIL-125 framework enabled high-affinity π-complexation interactions with C2 H4 , resulting in an C2 H4 /C2 H6 selectivity approaching 13.6, which was 9.4 times higher than that of pristine NH2 -MIL-125 membrane; moreover, benefiting from π-complexation interactions between CO2 and Cu(I) sites, our membrane displayed superior CO2 /N2 selectivity of 43.2 with CO2 permeance of 696 GPU, which far surpassed the benchmark of other pure MOF membranes. The above multi-scale structure optimization strategy is anticipated to present opportunities for significantly enhancing the separation performance of diverse molecular sieve membranes.

Research on machine vision and deep learning based recognition of cotton seedling aphid infestation level.

Aphis gossypii Glover is a major insect pest in cotton production, which can cause yield reduction in severe cases. In this paper, we proposed the A. gossypii infestation monitoring method, which identifies the infestation level of A. gossypii at the cotton seedling stage, and can improve the efficiency of early warning and forecasting of A. gossypii, and achieve precise prevention and cure according to the predicted infestation level. We used smartphones to collect A. gossypii infestation images and compiled an infestation image data set. And then constructed, trained, and tested three different A. gossypii infestation recognition models based on Faster Region-based Convolutional Neural Network (R-CNN), You Only Look Once (YOLO)v5 and single-shot detector (SSD) models. The results showed that the YOLOv5 model had the highest mean average precision (mAP) value (95.7%) and frames per second (FPS) value (61.73) for the same conditions. In studying the influence of different image resolutions on the performance of the YOLOv5 model, we found that YOLOv5s performed better than YOLOv5x in terms of overall performance, with the best performance at an image resolution of 640×640 (mAP of 96.8%, FPS of 71.43). And the comparison with the latest YOLOv8s showed that the YOLOv5s performed better than the YOLOv8s. Finally, the trained model was deployed to the Android mobile, and the results showed that mobile-side detection was the best when the image resolution was 256×256, with an accuracy of 81.0% and FPS of 6.98. The real-time recognition system established in this study can provide technical support for infestation forecasting and precise prevention of A. gossypii.

Food safety news events classification via a hierarchical transformer model.

In light of the significance of regulatory authorities and the rising demand for information disclosure, a vast amount of information on food safety news reports is readily accessible on the Internet. The extraction of such information for precise classification and provision of appropriate safety alerts based on their respective categories has emerged as a challenging problem for academic research. Given that most food safety-related events in news reports comprise lengthy text, the pre-trained language models currently employed for text analysis are generally limited in their capability to handle long documents. This paper proposes a long-text classification model utilising hierarchical Transformers. We categorise information in long documents into two distinct types: (1) multiple text chunks meeting the length constraint and (2) essential sentences within long documents, such as headings, paragraph start and end sentences, etc. Initially, our proposed model utilises the text chunks as input to the BERT model. Then, it concatenates the output of the BERT model with the important sentences from the document and use them as input to the Transformer model for feature transformation. Finally, we utilise a classifier for food safety news classification. We conducted several comparative experiments with various commonly used text classification models on a dataset constructed from publicly available information on food regulatory websites. Our proposed method outperforms existing methods, establishing itself as the leading approach in terms of performance.

A correlation graph attention network for classifying chromosomal instabilities from histopathology whole-slide images.

The chromosome instability (CIN) is one of the hallmarks of cancer and is closely related to tumor metastasis. However, the sheer size and resolution of histopathology whole-slide images (WSIs) already challenges the capabilities of computational pathology. In this study, we propose a correlation graph attention network (MLP-GAT) that can construct graphs for classifying multi-type CINs from the WSIs of breast cancer. We construct a WSIs dataset of breast cancer from the Cancer Genome Atlas Breast Invasive Carcinoma (TCGA-BRCA). Extensive experiments show that MLP-GAT far outperforms accepted state-of-the-art methods and demonstrate the advantages of the constructed graph networks for analyzing WSI data. The visualization shows the difference among the tiles in a WSI. Furthermore, the generalization performance of the proposed method was verified on the stomach cancer. This study provides guidance for studying the relationship between CIN and cancer from the perspective of image phenotype.

Phytochemical content and antioxidant activity of different anatomical parts of Corchorus olitorius and C. capsularis during different phenological stages.

Different parts of Corchorus olitorius and C. capsularis possess different antioxidant compounds. This study investigated the phytochemical components and antioxidant capacities of ultrasound-assisted extraction of different plant parts of both species using spectrophotometry at various phenological stages. Results also indicate that leaves, stems and roots of C. olitorius at various growth stages showed higher phytochemical components and antioxidant potential compared to C. capsularis. The phytochemical components from roots to leaves in C. olitorius including total polyphenol 0.97-11.11 mg GAE/g DW, total flavonoid 0.99-7.78 mg QE/g DW and total tannin 4.02-26.89 TA E/g DW, whereas C. capsularis total polyphenol 1.04-7.93 mg GAE/g DW, total flavonoid 0.77-5.5.92 mg QE/g DW and total tannin content 3.17-22.73 TA E/g DW. C. olitorius produced overall 22.23%, 13.61%, 12.24% higher total polyphenol, total flavonoid and total tannin, respectively compare to C. capsularis. Different parts extract also significantly affected antioxidant capacities including DPPH, ABTS, and FRAP activity with values of 22.03-79.46% inhibition, 10.84-104.10 µmol TE/g DW, and 10.84-104.10 µmol Fe2+/g DW respectively for C. olitorius, while C. capsularis demonstrated 14.03-70.97% of DPPH inhibition, 9.16-95.60 µmol TE/g DW of ABTS and 5.31-71.82 µmol Fe2+/g DW of FRAP activity. Moreover, leaves of the flowering stage, young stems and aged roots of both species displayed a higher content of phytochemical and antioxidant activities than other growth stages. A positive correlation between the phytochemical and antioxidant potential indicated that phenolic constituents solely affected antioxidant activity. Thus, this study established that the plant's parts and phenological growth stages significantly influence the concentration of phytoconstituents and antioxidant activities, and determine the harvesting stages of the different organs of C. olitorius and C. capsularis for considerable medicinal importance as folk and industry.

Pathogenicity and innate immune responses induced by fowl adenovirus serotype 8b in specific pathogen-free chicken.

Fowl adenovirus serotype 8b (FAdV-8b), as causative agent of inclusion body hepatitis (IBH), poses a great threat to the poultry industry. Considering the importance of innate immune response in host against viral infections, we investigated pathogenicity of a FAdV-8b strain HLJ/151129 in 1-mo-old specific pathogen-free (SPF) chickens and immune responses of host to FAdV-8b infection in this study. The results demonstrated that no obvious clinical signs were observed in infected birds. Neither mobility nor mortality was observed in both FAdV-8b infected and control chickens, as well. However, hepatic necrosis and a small amount of inflammatory cell infiltration were observed by pathological analysis. Viral load was detected in bursa of Fabricius, cecal tonsils, liver, heart, spleen, Harderian glands, and thymus. Virus shedding and viremia generated as early as 3 days postinfection (dpi) (9/10) and reached the peak at 7 dpi (10/10). In addition, the infected birds had developed FAdV-specific antibodies at 7 dpi, and the antibody titers reached the peak at 14 dpi. Furthermore, the results demonstrated that the mRNA expression levels of most of toll-like receptors (TLRs), most of avian ß-defensins (AvBDs), and cytokines [interleukin (IL)-2, IL-6, and interferon (IFN)-γ], were significantly upregulated in most tissues at early phases of FAdV-8b infection, especially in liver and spleen. In contrast, FAdV-8b infection results in downregulation of TLR4, TLR5, and TLR21 expressions in some tissues of infected chickens. In addition, FAdV-8b infection upregulated myeloid differentiation factor 88 (MyD88), nuclear factor-kappa B (NF-κB) p65, and TIR-domain-containing adapter inducing interferon-ß (TRIF) expression in some tissues, while decreased NF-κBp65 and TRIF in spleen at both 72 hpi and 21 dpi. Taken together, these results confirmed that FAdV-8b could replicate in all investigated tissues of infected birds, and then, result in production of FAdV-specific antibody titers. Meanwhile, the FAdV-8b infection induces strong innate immune responses at early stage in chickens, which may associate with the viral pathogenesis.

Hybrid Contextual Semantic Network for Accurate Segmentation and Detection of Small-Size Stroke Lesions From MRI.

Stroke is a cerebrovascular disease with high mortality and disability rates. The occurrence of the stroke typically produces lesions of different sizes, with the accurate segmentation and detection of small-size stroke lesions being closely related to the prognosis of patients. However, the large lesions are usually correctly identified, the small-size lesions are usually ignored. This article provides a hybrid contextual semantic network (HCSNet) that can accurately and simultaneously segment and detect small-size stroke lesions from magnetic resonance images. HCSNet inherits the advantages of the encoder-decoder architecture and applies a novel hybrid contextual semantic module that generates high-quality contextual semantic features from the spatial and channel contextual semantic features through the skip connection layer. Moreover, a mixing-loss function is proposed to optimize HCSNet for unbalanced small-size lesions. HCSNet is trained and evaluated on 2D magnetic resonance images produced from the Anatomical Tracings of Lesions After Stroke challenge (ATLAS R2.0). Extensive experiments demonstrate that HCSNet outperforms several other state-of-the-art methods in its ability to segment and detect small-size stroke lesions. Visualization and ablation experiments reveal that the hybrid semantic module improves the segmentation and detection performance of HCSNet.