Search

Background Specific factors correlated with hypothyroidism in systemic lupus erythematosus (SLE) patients remain unclear. Therefore, we aim to evaluate the prevalence of thyroid dysfunction in Chinese patients with SLE and the relationship between clinical hypothyroidism and SLE. Methods We conducted a cross sectional study of the prevalence of thyroid dysfunction in 672 patients with SLE and 605 age- and sex-matched healthy controls. Demographic, clinical, and biochemical data were compared between 58 patients with SLE with hypothyroidism and 197 patients with SLE with euthyroidism. Multivariate analysis was performed using binomial logistic regression analysis. Spearman's rank correlation was used to identify an association between thyroid function and disease activity. Results The prevalence of thyroid dysfunction was significantly higher in patients with SLE than in controls (70.7% vs 19.7%). SLE was associated with higher rates of hypothyroidism (9.6%, P ≤ 0.001) and euthyroid sick syndrome (49.6%, P ≤ 0.001) compared with control subjects. Further analyses showed that hypothyroidism in patients with SLE was associated with high blood pressure, renal disorder, high serum creatinine, high uric acid, hyperlipidaemia, low C3 and C4, positive anti-dsDNA antibodies, and high SLE disease activity index (SLEDAI) score. In multiple logistic regression models, albumin, platelet count, serum creatinine, and anti-dsDNA antibodies were associated with hypothyroidism. Finally, free tri-iodothyronine was significantly negatively correlated with SLEDAI score. Conclusions Hypothyroidism was more prevalent in patients with SLE. There was a relationship between hypothyroidism with renal disorder and lupus activity. Albumin, platelet count, serum creatinine, and anti-dsDNA antibodies were correlated with hypothyroidism.

Pyroptosis, a proinflammatory form of programmed cell death, plays important roles in the pathogenesis of many diseases. Inflammasome activation, which has been shown in hepatic ischemia–reperfusion injury (IRI), is demonstrated to be closely associated with pyroptosis, indicating that pyroptosis may occur and perform functions in hepatic IRI. However, there is no direct evidence showing the function of pyroptosis in hepatic IRI. In this study, by detecting the pyroptosis markers, we showed that pyroptosis may be induced during hepatic IRI. Furthermore, by adopting caspase-1 inhibitors, we showed that inhibition of pyroptosis could significantly ameliorate liver injury and suppress inflammatory response during hepatic IRI. Interestingly, caspase-1 inhibitors have no protective effects on in vitro hepatocytes under hypoxic reoxygenation condition. To investigate pyroptosis induced in which specific cell types may affect hepatic IRI, we generated hepatocyte-specific Gsdmd-knockout (Hep-Gsdmd−/−) and myeloid-specific Gsdmd-knockout (LysmCre+Gsdmdf/f) mice. Functional experiments showed that compared to control mice (Gsdmdf/f), there were alleviated liver injury and inflammation in LysmCre+Gsdmdf/f mice, but not in AlbCre+Gsdmdf/f mice. In parallel in vitro studies, cytokine expression and production decreased in bone-marrow-derived macrophages and Kupffer cells from LysmCre+Gsdmdf/f mice compared to their controls. Our findings demonstrated that pyroptosis in innate immune cells aggravates hepatic IRI and implied that hepatic IRI could be protected by blocking pyroptosis, which may become a potential therapeutic target in the clinic.

Cholangiocarcinoma (CCA) is an epithelial cancer and has high death and recurrence rates, current methods cannot satisfy the need for predicting cancer relapse effectively. So, we aimed at conducting a multi-mRNA signature to improve the relapse prediction of CCA. We analyzed mRNA expression profiling in large CCA cohorts from the Gene Expression Omnibus (GEO) database (GSE76297, GSE32879, GSE26566, GSE31370, and GSE45001) and The Cancer Genome Atlas (TCGA) database. The Least absolute shrinkage and selection operator (LASSO) regression model was used to establish a 7-mRNA-based signature that was significantly related to the recurrence-free survival (RFS) in two test series. Based on the 7-mRNA signature, the cohort TCGA patients could be divided into high-risk or low-risk subgroups with significantly different RFS [p < 0.001, hazard ratio (HR): 48.886, 95% confidence interval (CI): 6.226-3.837E+02]. Simultaneously, the prognostic value of the 7-mRNA signature was confirmed in clinical samples of Ren Ji hospital (p < 0.001, HR: 4.558, 95% CI: 1.829-11.357). Further analysis including multivariable and sub-group analyses revealed that the 7-mRNA signature was an independent prognostic value for recurrence of patients with CCA. In conclusion, our results might provide an efficient tool for relapse prediction and were beneficial to individualized management for CCA patients.

Recognizing human expression in videos is a challenging task due to dynamic changes in facial actions and diverse visual appearances. The key to design a reliable video-based expression recognition system is to extract robust spatial features and make full use of temporal modality characteristics. In this paper, we present a novel network architecture called Cascaded Attention Network (CAN) which is a cascaded spatiotemporal model incorporating with both spatial and temporal attention, tailored to video-level facial expression recognition. The cascaded fundamental model consists of a transfer convolutional network and Bidirectional Long Short-Term Memory (BiLSTM) network. Spatial attention is designed from the facial landmarks since facial expressions depend on the actions of key regions (eyebrows, eyes, nose, and mouth) on the face. Focusing on these key regions can help to decrease the effect of person-specific attributes. Meanwhile, the temporal attention is applied to automatically select the peak of expressions and aggregate the video-level representation. Our proposed CAN achieves the state-of-the-art performance on the three most widely used facial expression datasets: CK+ (99.03%), Oulu-CASIA (88.33%), and MMI (83.55%). Moreover, we conduct an extended experiment on a much more complex wild dataset AFEW and the experimental results further verify the generality of our attention mechanisms.

Recent deep networks have convincingly demonstrated high capability in crowd counting, which is a critical task attracting widespread attention due to its various industrial applications. Despite such progress, trained data-dependent models usually can not generalize well to unseen scenarios because of the inherent domain shift. To facilitate this issue, this paper proposes a novel adversarial scoring network (ASNet) to gradually bridge the gap across domains from coarse to fine granularity. In specific, at the coarse-grained stage, we design a dual-discriminator strategy to adapt source domain to be close to the targets from the perspectives of both global and local feature space via adversarial learning. The distributions between two domains can thus be aligned roughly. At the fine-grained stage, we explore the transferability of source characteristics by scoring how similar the source samples are to target ones from multiple levels based on generative probability derived from coarse stage. Guided by these hierarchical scores, the transferable source features are properly selected to enhance the knowledge transfer during the adaptation process. With the coarse-to-fine design, the generalization bottleneck induced from the domain discrepancy can be effectively alleviated. Three sets of migration experiments show that the proposed methods achieve state-of-the-art counting performance compared with major unsupervised methods., Accepted by ACMMM2021

Crowd counting is a challenging task due to the large variations in crowd distributions. Previous methods tend to tackle the whole image with a single fixed structure, which is unable to handle diverse complicated scenes with different crowd densities. Hence, we propose the Adaptive Capacity Multi-scale convolutional neural networks (ACM-CNN), a novel crowd counting approach which can assign different capacities to different portions of the input. The intuition is that the model should focus on important regions of the input image and optimize its capacity allocation conditioning on the crowd intensive degree. ACM-CNN consists of three types of modules: a coarse network, a fine network, and a smooth network. The coarse network is used to explore the areas that need to be focused via count attention mechanism, and generate a rough feature map. Then the fine network processes the areas of interest into a fine feature map. To alleviate the sense of division caused by fusion, the smooth network is designed to combine two feature maps organically to produce high-quality density maps. Extensive experiments are conducted on five mainstream datasets. The results demonstrate the effectiveness of the proposed model for both density estimation and crowd counting tasks., Comment: Accepted to Neurocomputing, code will be released soon

Video-based person re-identification aims to retrieve video sequences of the same person in the multi-camera system. In this paper, we propose a k -reciprocal harmonious attention network (KHAN) to jointly learn discriminative spatiotemporal features and the similarity metrics. In KHAN, the harmonious attention module adaptively calibrates response at each spatial position and each channel by explicitly inspecting position-wise and channel-wise interactions over feature maps. Besides, the k -reciprocal attention module attends key features from all frame-level features with a discriminative feature selection algorithm; thus, useful temporal information within contextualized key features can be assimilated to produce more robust clip-level representation. Compared with commonly used local-context based approaches, the proposed KHAN captures long dependency of different spatial regions and visual patterns while incorporating informative context at each time-step in a non-parametric manner. The extensive experiments on three public benchmark datasets show that the performance of our proposed approach outperforms the state-of-the-art methods.

Background The right posterior segment (RPS) graft was introduced to overcome graft size discrepancy in living donor liver transplantation (LDLT). However, it was very rarely used in pediatric patients. Here we presented 4 pediatric LDLT cases receiving RPS graft between January 2015 and April 2020 in our center. A total of 1868 LDLT procedures were performed in this period. Methods Recipients included 1 boy and 3 girls with a median age of 45 months (range from 40 to 93 months). They were diagnosed with progressive familial intrahepatic cholestasis, propionic academia, ornithine transcarbamylase and biliary atresia, respectively. Four donors were all mothers with a median age of 32.5 years (31–38 years). Computer tomography angiography indicated posterior right branches branched off separately from main portal veins (type III variation). Three of these donor livers had 1 orifice of right hepatic veins (RHV). In the remaining 1 donor liver, the RHV showed 3 orifices and an outflow patch plastic was performed. Inferior right hepatic veins weren’t found in four donor grafts. The median graft weight was 397.5 g (352–461 g) and the median graft-to-recipient weight ratio was 2.38% (1.44–2.80%). Results Postoperative complications occurred in neither donors nor recipients. Within the median follow-up duration of 29 months (14–64 months), four children are all alive with normal liver function. Conclusion In summary, for older children weighed more than 15 kg with donors’ variation of type III portal veins, the use of RPS grafts could be a feasible and favorable option.

This paper addresses the problem of temporal sentence grounding (TSG), which aims to identify the temporal boundary of a specific segment from an untrimmed video by a sentence query. Previous works either compare pre-defined candidate segments with the query and select the best one by ranking, or directly regress the boundary timestamps of the target segment. In this paper, we propose a novel localization framework that scores all pairs of start and end indices within the video simultaneously with a biaffine mechanism. In particular, we present a Context-aware Biaffine Localizing Network (CBLN) which incorporates both local and global contexts into features of each start/end position for biaffine-based localization. The local contexts from the adjacent frames help distinguish the visually similar appearance, and the global contexts from the entire video contribute to reasoning the temporal relation. Besides, we also develop a multi-modal self-attention module to provide fine-grained query-guided video representation for this biaffine strategy. Extensive experiments show that our CBLN significantly outperforms state-of-the-arts on three public datasets (ActivityNet Captions, TACoS, and Charades-STA), demonstrating the effectiveness of the proposed localization framework., Comment: Accepted by CVPR 2021

This paper aims for the language-based product image retrieval task. The majority of previous works have made significant progress by designing network structure, similarity measurement, and loss function. However, they typically perform vision-text matching at certain granularity regardless of the intrinsic multiple granularities of images. In this paper, we focus on the cross-modal similarity measurement, and propose a novel Hierarchical Similarity Learning (HSL) network. HSL first learns multi-level representations of input data by stacked encoders, and object-granularity similarity and image-granularity similarity are computed at each level. All the similarities are combined as the final hierarchical cross-modal similarity. Experiments on a large-scale product retrieval dataset demonstrate the effectiveness of our proposed method. Code and data are available at https://github.com/liufh1/hsl., Comment: Accepted by ICASSP 2021. Code and data will be available at https://github.com/liufh1/hsl

Temporal sentence localization in videos aims to ground the best matched segment in an untrimmed video according to a given sentence query. Previous works in this field mainly rely on attentional frameworks to align the temporal boundaries by a soft selection. Although they focus on the visual content relevant to the query, these single-step attention are insufficient to model complex video contents and restrict the higher-level reasoning demand for this task. In this paper, we propose a novel deep rectification-modulation network (RMN), transforming this task into a multi-step reasoning process by repeating rectification and modulation. In each rectification-modulation layer, unlike existing methods directly conducting the cross-modal interaction, we first devise a rectification module to correct implicit attention misalignment which focuses on the wrong position during the cross-interaction process. Then, a modulation module is developed to capture the frame-to-frame relation with the help of sentence information for better correlating and composing the video contents over time. With multiple such layers cascaded in depth, our RMN progressively refines video and query interactions, thus enabling a further precise localization. Experimental evaluations on three public datasets show that the proposed method achieves state-of-the-art performance. Extensive ablation studies are carried out for the comprehensive analysis of the proposed method.

Cross-domain sentiment classification aims to predict sentiment polarity on a target domain utilizing a classifier learned from a source domain. Most existing adversarial learning methods focus on aligning the global marginal distribution by fooling a domain discriminator, without taking category-specific decision boundaries into consideration, which can lead to the mismatch of category-level features. In this work, we propose an adversarial category alignment network (ACAN), which attempts to enhance category consistency between the source domain and the target domain. Specifically, we increase the discrepancy of two polarity classifiers to provide diverse views, locating ambiguous features near the decision boundaries. Then the generator learns to create better features away from the category boundaries by minimizing this discrepancy. Experimental results on benchmark datasets show that the proposed method can achieve state-of-the-art performance and produce more discriminative features.