The ability to quickly recognize and learn new visual concepts from limited samples enable humans to quickly adapt to new tasks and environments. This ability is enabled by the semantic association of novel concepts with those that have already been learned and stored in memory. Computers can start to ascertain similar abilities by utilizing a semantic concept space. A concept space is a high-dimensional semantic space in which similar abstract concepts appear close and dissimilar ones far apart. In this paper, we propose a novel approach to one-shot learning that builds on this core idea. Our approach learns to map a novel sample instance to a concept, relates that concept to the existing ones in the concept space and, using these relationships, generates new instances, by interpolating among the concepts, to help learning. Instead of synthesizing new image instance, we propose to directly synthesize instance features by leveraging semantics using a novel auto-encoder network called dual TriNet. The encoder part of the TriNet learns to map multi-layer visual features from CNN to a semantic vector. In semantic space, we search for related concepts, which are then projected back into the image feature spaces by the decoder portion of the TriNet. Two strategies in the semantic space are explored. Notably, this seemingly simple strategy results in complex augmented feature distributions in the image feature space, leading to substantially better performance.

Tumor-infiltrating immune cells (TIICs) play essential roles in cancer development and progression. However, the association of TIICs with prognosis in colorectal cancer (CRC) patients remains elusive. Infiltration of TIICs was assessed using ssGSEA and CIBERSORT tools. The association of TIICs with prognosis was analyzed in 1,802 CRC data downloaded from the GEO (https://www.ncbi.nlm.nih.gov/geo/) and TCGA (https://portal.gdc.cancer.gov/) databases. Three populations of TIICs, including CD66b+ tumor-associated neutrophils (TANs), FoxP3+ Tregs, and CD163+ tumor-associated macrophages (TAMs) were selected for immunohistochemistry (IHC) validation analysis in 1,008 CRC biopsies, and their influence on clinical features and prognosis of CRC patients was analyzed. Prognostic models were constructed based on the training cohort (359 patients). The models were further tested and verified in testing (249 patients) and validation cohorts (400 patients). Based on ssGSEA and CIBERSORT analysis, the correlation between TIICs and CRC prognosis was inconsistent in different datasets. Moreover, the results with disease-free survival (DFS) and overall survival (OS) data in the same dataset also differed. The high abundance of TIICs found by ssGSEA or CIBERSORT tools can be used for prognostic evaluation effectively. IHC results showed that TANs, Tregs, TAMs were significantly correlated with prognosis in CRC patients and were independent prognostic factors (P_DFS ≤ 0.001; P_OS ≤ 0.023). The prognostic predictive models were constructed based on the numbers of TANs, Tregs, TAMs (C-index_DFS&OS = 0.86; AIC_DFS = 448.43; AIC_OS = 184.30) and they were more reliable than traditional indicators for evaluating prognosis in CRC patients. Besides, TIICs may affect the response to chemotherapy. In conclusion, TIICs were correlated with clinical features and prognosis in patients with CRC and thus can be used as markers.

The 5-methylcytosine (m5C) RNA methyltransferase NSUN2 is involved in the regulation of cell proliferation and metastasis formation and is upregulated in multiple cancers. However, the biological significance of NSUN2 in gastric cancer (GC) and the modification of NSUN2 itself have not been fully investigated. Here, we analyzed the expression level of NSUN2 in tissue microarrays containing 403 GC tissues by immunohistochemistry. NSUN2 was upregulated in GC, and that it was a predictor of poor prognosis. NSUN2 promotes the proliferation, migration, and invasion of GC cells in vitro. We also demonstrated that small ubiquitin-like modifier (SUMO)-2/3 interacts directly with NSUN2 by stabilizing it and mediating its nuclear transport. This facilitates the carcinogenic activity of NSUN2. Furthermore, m5C bisulfite sequencing (Bis-seq) in NSUN2-deficient GC cells showed that m5C-methylated genes are involved in multiple cancer-related signaling pathways. PIK3R1 and PCYT1A may be the target genes that participate in GC progression. Our findings revealed a novel mechanism by which NSUN2 functions in GC progression. This may provide new treatment options for GC patients.

This paper aims at learning to score the figure skating sports videos. To address this task, we propose a deep architecture that includes two complementary components, i.e., Self-Attentive LSTM and Multi-scale Convolutional Skip LSTM. These two components can efficiently learn the local and global sequential information in each video. Furthermore, we present a large-scale figure skating sports video dataset FisV dataset. This dataset includes 500 figure skating videos with the average length of 2 minutes and 50 seconds. Each video is annotated by two scores of nine different referees, i.e., Total Element Score(TES) and Total Program Component Score (PCS). Our proposed model is validated on FisV and MIT-skate datasets. The experimental results show the effectiveness of our models in learning to score the figure skating videos. The codes and datasets would be downloaded from https://github.com/loadder/MS_LSTM.git.

The induction of cuproptosis, a recently identified form of copper‐dependent immunogenic cell death, is a promising approach for antitumor therapy. However, sufficient accumulation of intracellular copper ions (Cu²⁺) in tumor cells is essential for inducing cuproptosis. Herein, an intelligent cuproptosis‐inducing nanosystem is constructed by encapsulating copper oxide (CuO) nanoparticles with the copper ionophore elesclomol (ES). After uptake by tumor cells, ES@CuO is degraded to release Cu²⁺ and ES to synergistically trigger cuproptosis, thereby significantly inhibiting the tumor growth of murine B16 melanoma cells. Moreover, ES@CuO further promoted cuproptosis‐mediated immune responses and reprogrammed the immunosuppressive tumor microenvironment by increasing the number of tumor‐infiltrating lymphocytes and secreted inflammatory cytokines. Additionally, combining ES@CuO with programmed cell death‐1 (PD‐1) immunotherapy substantially increased the antitumor efficacy in murine melanoma. Overall, the findings of this study can lead to the use of a novel strategy for cuproptosis‐mediated antitumor therapy, which may enhance the efficacy of immune checkpoint inhibitor therapy.