Whole-brain imaging has become an increasingly important approach to investigate neural structures, such as somata distribution, dendritic morphology, and axonal projection patterns. Different structures require whole-brain imaging at different resolutions. Thus, it is highly desirable to perform whole-brain imaging at multiple scales. Imaging a complete mammalian brain at synaptic resolution is especially challenging, as it requires continuous imaging from days to weeks because of the large number of voxels to sample, and it is difficult to acquire a constant quality of imaging because of light scattering during in toto imaging. Here, we reveal that light-sheet microscopy has a unique advantage over wide-field microscopy in multi-scale imaging because of its decoupling of illumination and detection. Based on this observation, we have developed a multi-scale light-sheet microscope that combines tiling of light-sheet, automatic zooming, periodic sectioning, and tissue expansion to achieve a constant quality of brain-wide imaging from cellular (3 μm × 3 μm × 8 μm) to sub-micron (0.3 μm × 0.3 μm × 1 μm) spatial resolution rapidly (all within a few hours). We demonstrated the strength of the system by testing it using mouse brains prepared using different clearing approaches. We were able to track electrode tracks as well as axonal projections at sub-micron resolution to trace the full morphology of single medial prefrontal cortex (mPFC) neurons that have remarkable diversity in long-range projections.
In this paper, we investigate deep-learning-based image inpainting techniques for emergency remote sensing mapping. Image inpainting can generate fabricated targets to conceal real-world private structures and ensure informational privacy. However, casual inpainting outputs may seem incongruous within original contexts. In addition, the residuals of original targets may persist in the hiding results. A Residual Attention Target-Hiding (RATH) model has been proposed to address these limitations for remote sensing target hiding. The RATH model introduces the residual attention mechanism to replace gated convolutions, thereby reducing parameters, mitigating gradient issues, and learning the distribution of targets present in the original images. Furthermore, this paper modifies the fusion module in the contextual attention layer to enlarge the fusion patch size. We extend the edge-guided function to preserve the original target information and confound viewers. Ablation studies on an open dataset proved the efficiency of RATH for image inpainting and target hiding. RATH had the highest similarity, with a 90.44% structural similarity index metric (SSIM), for edge-guided target hiding. The training parameters had 1M fewer values than gated convolution (Gated Conv). Finally, we present two automated target-hiding techniques that integrate semantic segmentation with direct target hiding or edge-guided synthesis for remote sensing mapping applications.
Persistent activity underlying short-term memory encodes sensory information, or instructs specific future movement, and consequently plays a crucial role in cognition. Despite extensive study, how the same set of neurons respond differentially to form selective persistent activity remains unknown. Here we report that the cortico-basal ganglia-thalamo-cortical circuit supports the formation of selective persistent activity. Optogenetic activation or inactivation of the basal ganglia output nucleus SNr to thalamus pathway biased future licking choice, without affecting licking execution. This perturbation differentially affected persistent activity in the frontal cortex and selectively modulated neural trajectory encoding one choice but not the other. Recording showed that SNr neurons had selective persistent activity distributed across SNr, but with a hotspot in the lateral region. Optogenetic inactivation of the frontal cortex also differentially affected persistent activity in SNr. Together, these results reveal that a cortico-basal ganglia-thalamo-cortical channel functions to produce selective persistent activity underlying short-term memory.
Accumulating evidence has highlighted the effects of natural killer (NK) cells on shaping anti-tumor immunity. This study aimed to construct an NK cell marker gene signature (NKMS) to predict prognosis and therapeutic response of clear cell renal cell carcinoma (ccRCC) patients.Publicly available single-cell and bulk RNA profiles with matched clinical information of ccRCC patients were collected from Gene Expression Omnibus (GEO), The Cancer Genome Atlas (TCGA), ArrayExpress, and International Cancer Genome Consortium (ICGC) databases. A novel NKMS was constructed, and its prognostic value, associated immunogenomic features and predictive capability to immune checkpoint inhibitors (ICIs) and anti-angiogenic therapies were evaluated in ccRCC patients.We identified 52 NK cell marker genes by single-cell RNA-sequencing (scRNA-seq) analysis in GSE152938 and GSE159115. After least absolute shrinkage and selection operator (LASSO) and Cox regression, the most prognostic 7 genes (CLEC2B, PLAC8, CD7, SH3BGRL3, CALM1, KLRF1, and JAK1) composed NKMS using bulk transcriptome from TCGA. Survival and time-dependent receiver operating characteristic (ROC) analysis exhibited exceptional predictive capability of the signature in the training set and two independent validation cohorts (E-MTAB-1980 and RECA-EU cohorts). The seven-gene signature was able to identify patients within high Fuhrman grade (G3-G4) and American Joint Committee on Cancer (AJCC) stage (III-IV). Multivariate analysis confirmed the independent prognostic value of the signature, and a nomogram was built for clinical utility. The high-risk group was characterized by a higher tumor mutation burden (TMB) and greater infiltration of immunocytes, particularly CD8+ T cells, regulatory T (Treg) cells and follicular helper T (Tfh) cells, in parallel with higher expression of genes negatively regulating anti-tumor immunity. Moreover, high-risk tumors exhibited higher richness and diversity of T-cell receptor (TCR) repertoire. In two therapy cohorts of ccRCC patients (PMID32472114 and E-MTAB-3267), we demonstrated that high-risk group showed greater sensitivity to ICIs, whereas the low-risk group was more likely to benefit from anti-angiogenic therapy.We identified a novel signature that can be utilized as an independent predictive biomarker and a tool for selecting the individualized treatment for ccRCC patients.
Remote sensing imagery is of great significance for policy decisions, especially for disaster assessment and disaster relief. To ensure the privacy and inviolability of personal buildings, the information containing these buildings must be anonymized during the remote sensing mapping process. Traditional processing methods for these targets in remote sensing mapping are mainly based on manual retrieval and image editing tools, which are inefficient. Deep learning provides a new direction for target hiding. Although the image inpainting method based on deep learning is faster than the manual method, the cost of training calculation is a disadvantage. And the element-wise product operation used in the model increases the risk of vanished or exploded gradients. We propose a Residual Attention Target Hiding (RATH) model for remote sensing target hiding based on deep learning. RATH uses residual attention modules to replace gated convolutions, reducing parameters and mitigating gradient issues. The residual attention module preserves gated convolution performance but provides an adjustable kernel size. RATH retains gated convolutions for dynamic feature selection and balances model depth and width. Furthermore, this paper modifies the contextual attention layer by adjusting the fusion process to enlarge the fusion patch size. Finally, we extend the edge-guided function to preserve the original target information and confound viewers. Ablation studies on an open dataset prove RATH’s efficiency for image inpainting and target hiding. RATH achieves state-of-the-art results with lower complexity. And it has the highest similarity for edge-guided target hiding. RATH enables robust, efficient target hiding for privacy protection in remote sensing imagery while balancing performance and complexity. Experiments show RATH's superiority over existing methods in hiding arbitrary-shaped targets.
Along with the economic globalization, international competition is becoming more and more intense and internationalization of higher education has become an irresistible trend. Cultivation of international talents has become the only way for the development of a country. Internationalization of higher education in China is speeding up with the proposal of policy of “the Belt and Road” and “Double First-Class University Plan”. This study explores the relationship between internationalization of higher education and the further study trend of overseas students, and analyzes it with the case of university. The ways to promote internationalization of higher education are given. The results show that the total number of overseas students and the number of students studying in China have increased rapidly in recent years, and the number of Confucius Institutes established abroad has exceeded 550. China has made great efforts to improve the guarantee system for internationalization of higher education, the management system of overseas students and international academic exchanges, and has put forward many ways to solve the existing problems. This study will provide a theoretical basis for the realization of internationalization of higher education in China.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)