Search

Abstract Robust data privacy regulations hinder the exchange of healthcare data among institutions, crucial for global insights and developing generalised clinical models. Federated learning (FL) is ideal for training global models using datasets from different institutions without compromising privacy. However, disparities in electronic healthcare records (EHRs) lead to inconsistencies in ML-ready data views, making FL challenging without extensive preprocessing and information loss. These differences arise from variations in services, care standards, and record-keeping practices. This paper addresses data view heterogeneity by introducing a knowledge abstraction and filtering-based FL framework that allows FL over heterogeneous data views without manual alignment or information loss. The knowledge abstraction and filtering mechanism maps raw input representations to a unified, semantically rich shared space for effective global model training. Experiments on three healthcare datasets demonstrate the framework’s effectiveness in overcoming data view heterogeneity and facilitating information sharing in a federated setup.

Abstract Objectives This study aims to enhance supervised human activity recognition based on spatiotemporal graph convolutional neural networks by addressing two key challenges: (1) extracting local spatial feature information from implicit joint connections that is unobtainable through standard graph convolutions on natural joint connections alone. (2) Capturing long-range temporal dependencies that extend beyond the limited temporal receptive fields of conventional temporal convolutions. Methods To achieve these objectives, we propose three novel modules integrated into the spatiotemporal graph convolutional framework: (1) a connectivity feature extraction module that employs attention to model implicit joint connections and extract their local spatial features. (2) A long-range frame difference feature extraction module that captures extensive temporal context by considering larger frame intervals. (3) A coordinate transformation module that enhances spatial representation by fusing Cartesian and spherical coordinate systems. Findings Evaluation across multiple datasets demonstrates that the proposed method achieves significant improvements over baseline networks, with the highest accuracy gains of 2.76 $$\%$$ % on the NTU-RGB+D 60 dataset (Cross-subject), 4.1 $$\%$$ % on NTU-RGB+D 120 (Cross-subject), and 4.3 $$\%$$ % on Kinetics (Top-1), outperforming current state-of-the-art algorithms. This paper delves into the realm of behavior recognition technology, a cornerstone of autonomous systems, and presents a novel approach that enhances the accuracy and precision of human activity recognition.

Abstract Deep neural networks have been integrated into the whole clinical decision procedure which can improve the efficiency of diagnosis and alleviate the heavy workload of physicians. Since most neural networks are supervised, their performance heavily depends on the volume and quality of available labels. However, few such labels exist for rare diseases (e.g., new pandemics). Here we report a medical multimodal large language model (Med-MLLM) for radiograph representation learning, which can learn broad medical knowledge (e.g., image understanding, text semantics, and clinical phenotypes) from unlabelled data. As a result, when encountering a rare disease, our Med-MLLM can be rapidly deployed and easily adapted to them with limited labels. Furthermore, our model supports medical data across visual modality (e.g., chest X-ray and CT) and textual modality (e.g., medical report and free-text clinical note); therefore, it can be used for clinical tasks that involve both visual and textual data. We demonstrate the effectiveness of our Med-MLLM by showing how it would perform using the COVID-19 pandemic “in replay”. In the retrospective setting, we test the model on the early COVID-19 datasets; and in the prospective setting, we test the model on the new variant COVID-19-Omicron. The experiments are conducted on 1) three kinds of input data; 2) three kinds of downstream tasks, including disease reporting, diagnosis, and prognosis; 3) five COVID-19 datasets; and 4) three different languages, including English, Chinese, and Spanish. All experiments show that our model can make accurate and robust COVID-19 decision-support with little labelled data.

Abstract Object detection holds a crucial role in medical diagnostics. Tasks like organ segmentation and malignancy diagnosis typically necessitate preliminary localization of corresponding anatomical structures. Precise positioning ensures that only pertinent regions require processing, leading to a potential reduction in computational and storage demands. Conventional image detection approaches necessitate numerous candidate boxes, resulting in redundant computations. Developing techniques capable of accurately detecting medical image objects without reliance on candidate boxes holds substantial practical significance. This paper introduces a 2D method for detecting medical image objects, which leverages multi-agent deep Q-network reinforcement learning and a multi-scale image representation. The method constructs a collaborative environment for multiple agents. These agents individually govern the upper-right corner and lower-left corner positions of the object detection frame, progressively converging toward the actual endpoint through iterative interactions. To expedite the detection process, a multi-scale image representation technique is employed. This method segments the process into three scales. Initially, within the coarse-scale space, the agent approximates the region containing the true endpoint, subsequently executing oscillatory movements. Progressively, it refines its approach within the fine-scale space, advancing toward the genuine endpoint with smaller iterative steps. The detection results demonstrate that collaborative detection among agents yields a 2.45 $$\%$$ % higher intersection over union compared to non-collaborative detection. Agents exhibit varying step sizes and fields of view in different scale spaces, leading to a reduction in detection time by 0.12 s compared to single-scale comparison. Experimental outcomes demonstrate the superiority of the medical image target detection method proposed in this study over prevailing mainstream detection algorithms.

Abstract Background The Neo-REGATTA study evaluated the effectiveness and safety of Docetaxel, oxaliplatin, and S-1 (DOS regimen) followed by radical resection vs. chemotherapy in advanced gastric adenocarcinoma patients with single non-curable factor. Methods This cohort study prospectively enrolled advanced gastric adenocarcinoma patients with single non-curable factor between November 2017 and June 2021. Patients without progression after four cycles of DOS were divided into resection group and chemotherapy group. The outcomes included overall survival (OS), progression-free survival (PFS) and safety. Effectiveness analysis was also performed by propensity score matching (PSM). Results A total of 73 patients were enrolled and 13 patients were withdrawn due to disease progression after 4 cycles of DOS. Afterwards, 35 and 25 participants were in the resection and chemotherapy groups, respectively. After a median follow-up time of 30.0 months, the median PFS and OS were 9.0 months, and 18.0 months for the chemotherapy group, but not reached in the resection group. After PSM, 19 matched participants were in each group, and the median PFS and OS were longer in resection group than that in chemotherapy group. The most common grade 3 or 4 adverse events both in the resection group and chemotherapy groups were neutropenia (5.7%, 8.0%) and leukopenia (5.7%, 8.0%). Conclusions Radical resection might provide survival benefit compared with continuous chemotherapy alone in advanced gastric adenocarcinoma patients who had a disease control after DOS, with a good safety profile. Trial registration The study protocol was registered on ClinicalTrial.gov (NCT03001726, 23/12/2016).

Aurora A plays a critical role in G2/M transition and mitosis, making it an attractive target for cancer treatment. Aurora A inhibitors showed remarkable antitumor effects in preclinical studies, but unsatisfactory outcomes in clinical trials have greatly limited their development. In this study, the Aurora A inhibitor alisertib upregulated programmed death ligand 1 (PD-L1) expression in a panel of tumor cells both in vitro and in vivo. Upregulation of the checkpoint protein PD-L1 reduced antitumor immunity in immune-competent mice, paradoxically inhibiting the antitumor effects of alisertib. Mechanistically, Aurora A directly bound to and phosphorylated cyclic GMP-AMP synthase (cGAS), suppressing PD-L1 expression in tumor cells. Aurora A inhibition by alisertib activated the cGAS/stimulator of IFN genes (STING)/NF-κB pathway and promoted PD-L1 expression. Combining alisertib with anti–PD-L1 antibody improved antitumor immunity and enhanced the antitumor effects of alisertib in immune-competent mice. Our results, which reveal the immunomodulatory functions of Aurora A inhibitors and provide a plausible explanation for the poor clinical outcomes with their use, offer a potential approach to improve the antitumor efficacy of these inhibitors.

Repetitive electrical nerve stimulation can induce a long-lasting perturbation of the axon's membrane potential, resulting in unstable stimulus-response relationships. Despite being observed in electrophysiology, the precise mechanism underlying electrical stimulation-dependent (ES-dependent) instability is still an open question. This study proposes a model to reveal a facet of this problem: how threshold fluctuation affects electrical nerve stimulations. This study proposes a new method based on a Circuit-Probability theory (C-P theory) to reveal the interlinkages between the subthreshold oscillation induced by neurons' resonance and ES-dependent instability of neural response. Supported by in-vivo studies, this new model predicts several key characteristics of ES-dependent instability and proposes a stimulation method to minimize the instability. This model provides a powerful tool to improve our understanding of the interaction between the external electric field and the complexity of the biophysical characteristics of axons.

In the realm of modern medicine, medical imaging stands as an irreplaceable pillar for accurate diagnostics. The significance of precise segmentation in medical images cannot be overstated, especially considering the variability introduced by different practitioners. With the escalating volume of medical imaging data, the demand for automated and efficient segmentation methods has become imperative. This study introduces an innovative approach to heart image segmentation, embedding a multi-scale feature and attention mechanism within an inverted pyramid framework. Recognizing the intricacies of extracting contextual information from low-resolution medical images, our method adopts an inverted pyramid architecture. Through training with multi-scale images and integrating prediction outcomes, we enhance the network’s contextual understanding. Acknowledging the consistent patterns in the relative positions of organs, we introduce an attention module enriched with positional encoding information. This module empowers the network to capture essential positional cues, thereby elevating segmentation accuracy. Our research resides at the intersection of medical imaging and sensor technology, emphasizing the foundational role of sensors in medical image analysis. The integration of sensor-generated data showcases the symbiotic relationship between sensor technology and advanced machine learning techniques. Evaluation on two heart datasets substantiates the superior performance of our approach. Metrics such as the Dice coefficient, Jaccard coefficient, recall, and F-measure demonstrate the method’s efficacy compared to state-of-the-art techniques. In conclusion, our proposed heart image segmentation method addresses the challenges posed by diverse medical images, offering a promising solution for efficiently processing 2D/3D sensor data in contemporary medical imaging.

In the era of artificial intelligence, the significance of images and videos as intuitive conveyors of information cannot be overstated. Computer vision techniques rooted in deep learning have revolutionized our ability to autonomously and accurately identify objects within visual media, making them a focal point of contemporary research. This study addresses the pivotal role of image object detection, particularly in the contexts of autonomous driving and security surveillance, by presenting an in-depth exploration of this field with a focus on enhancing the feature pyramid network. One of the key challenges in existing object detection methodologies lies in mitigating information loss caused by multi-scale feature fusion. To tackle this issue, we propose the enhanced feature pyramid, which adeptly amalgamates features extracted across different scales. This strategic enhancement effectively curbs information attrition across various layers, thereby strengthening the feature extraction capabilities of the foundational network. Furthermore, we confront the issue of excessive classification loss in image object detection tasks by introducing the gradient density loss function, designed to mitigate classification discrepancies. Empirical results unequivocally demonstrate the efficacy of our approach in enhancing the detection of multi-scale objects within images. When evaluated across benchmark datasets, including MS COCO 2017, MS COCO 2014, Pascal VOC 2007, and Pascal VOC 2012, our method achieves impressive average precision scores of 39.4%, 42.0%, 51.5%, and 49.9%, respectively. This performance clearly outperforms alternative state-of-the-art methods in the field. This research not only contributes to the evolving landscape of computer vision and object detection but also has practical implications for a wide range of applications, aligning with the transformative trends in the automotive industry and security technologies.

ABSTRACT: Aims: A high proportion of all patients with tuberculosis (TB) present with extrapulmonary TB (EPTB), including concurrent EPTB involving more than one extrapulmonary lesion site. However, previous reports only characterized lesions of single-site EPTB cases. This study aimed to investigate epidemiological characteristics and association rules of concurrent EPTB cases in China. Methods: An observational multi-centre study of 208,214 patients with EPTB lesions was undertaken in China from January 2011 to December 2017. Multi-variable logistic regression analysis was used to identify associations between gender and concurrent EPTB, and age and concurrent EPTB. Association rules were analysed for significance using the Apriori algorithm. Results: The most common EPTB lesion was tuberculous pleurisy (49.8%), followed by bronchial TB (14.8%) and tuberculous meningitis (7.6%). The most common type of concurrent EPTB was tuberculous pleurisy concurrent with tuberculous peritonitis (1.80%). In total, 22 association rules, including 20 strong association rules, were identified; among these, the highest confidence rates were found for tuberculous myelitis concurrent with tuberculous meningitis, and sacral TB concurrent with lumbar vertebral TB. The association rules of EPTB concurrent with other EPTB types were found to vary with gender and age. The confidence rate of tuberculous myelitis concurrent with tuberculous meningitis was higher in females (83.67%) than males, and was highest in patients aged 25–34 years (87.50%). Conclusions: Many types of concurrent EPTB were found. Greater awareness of concurrent EPTB disease characteristics is needed to ensure timely clinical diagnosis and treatment of this disease.

Toll‐like receptor 1 (TLR1) mediates the innate immune response to a variety of microbes through recognizing cell wall components (such as bacterial lipoproteins) in mammals. However, the detailed molecular mechanism of TLR1 involved in pathogen immunity in the representative hybrid yellow catfish (Pelteobagrus fulvidraco ♀ × P. vachelli ♂) has not been well studied. In the present study, we identified the TLR1 gene from the hybrid yellow catfish, and further comparative synteny data from multiple species confirmed that the TLR1 gene is highly conserved in teleosts. Phylogenetic analysis revealed distinguishable TLR1s in diverse taxa, suggesting consistence in evolution of the TLR1 proteins with various species. Structural prediction indicated that the three-dimensional structures of TLR1 proteins are relatively conserved among different taxa. Positive selection analysis showed that purifying selection dominated the evolutionary process of TLR1s and TLR1-TIR domain in both vertebrates and invertebrates. Expression pattern analysis based on the tissue distribution showed that TLR1 mainly transcribed in the gonad, gallbladder and kidney, and the mRNA levels of TLR1 in kidney were remarkably up-regulated after Aeromonas hydrophila stimulation, indicating that TLR1 participates in the inflammatory responses to exogenous pathogen infection in hybrid yellow catfish. Homologous sequence alignment and chromosomal location indicated that the TLR signaling pathway is very conserved in the hybrid yellow catfish. The expression patterns of TLR signaling pathway related genes (TLR1- TLR2 - MyD88 - FADD - Caspase 8) were consistent after pathogen stimulation, revealing that the TLR signaling pathway is triggered and activated after A. hydrophila infection. Our findings will lay a solid foundation for better understanding the immune roles of TLR1 in teleosts, as well as provide basic data for developing strategies to control disease outbreak in hybrid yellow catfish.

Background. Tuberculosis (TB), a multisystemic disease with protean presentation, remains a major global health problem. Although concurrent pulmonary tuberculosis (PTB) and extrapulmonary tuberculosis (EPTB) cases are commonly observed clinically, knowledge regarding concurrent PTB-EPTB is limited. Here, a large-scale multicenter observational study conducted in China aimed to study the epidemiology of concurrent PTB-EPTB cases by diagnostically defining TB types and then implementing association rules analysis. Methods. The retrospective study was conducted at 21 hospitals in 15 provinces in China and included all inpatients with confirmed TB diagnoses admitted from Jan 2011 to Dec 2017. Association rules analysis was conducted for cases with concurrent PTB and various types of EPTB using the Apriori algorithm. Results. Evaluation of 438,979TB inpatients indicated PTB was the most commonly diagnosed (82.05%) followed by tuberculous pleurisy (23.62%). Concurrent PTB-EPTB was found in 129,422 cases (29.48%) of which tuberculous pleurisy was the most common concurrent EPTB type observed. The multivariable logistic regression models demonstrated that odds ratios of concurrent PTB-EPTB cases varied by gender and age group. For PTB cases with concurrent EPTB, the strongest association was found between PTB and concurrent bronchial tuberculosis (lift = 1.09). For EPTB cases with concurrent PTB, the strongest association was found between pharyngeal/laryngeal tuberculosis and concurrent PTB (lift = 1.11). Confidence and lift values of concurrent PTB-EPTB cases varied with gender and age. Conclusions. Numerous concurrent PTB-EPTB case types were observed, with confidence and lift values varying with gender and age. Clinicians should screen for concurrent PTB-EPTB in order to improve treatment outcomes.

Prefabricated buildings’ quality, safety, and construction efficiency are closely related to the connections between various prefabricated components. This study optimizes the design of corrugated pipe-restrained grout anchor lap connections, further simplifying the construction methodology while ensuring reliable connection performance. The objective is to achieve cost savings and accelerate construction progress. A pull-out test of this new type of joint is conducted, generating two sets of 17 specimens each. The research examines the effects of grouting material strength, grout anchor steel bar anchorage length, and U-shaped stirrup spacing on the connection performance of corrugated pipe grout anchor connections. The results indicated that the specimens primarily experience two types of failure: anchorage failure with steel bar pull-out and steel bar tensile failure without anchorage failure. The strength of the grouting material and the anchorage length of the grout anchor steel bar positively correlate with bond anchorage performance. U-shaped stirrups contribute to restraint and enhance the bonding force between the corrugated pipe, grouting materials, and the grout anchor steel bar to a certain extent.

Visual saliency refers to the human’s ability to quickly focus on important parts of their visual field, which is a crucial aspect of image processing, particularly in fields like medical imaging and robotics. Understanding and simulating this mechanism is crucial for solving complex visual problems. In this paper, we propose a salient object detection method based on boundary enhancement, which is applicable to both 2D and 3D sensors data. To address the problem of large-scale variation of salient objects, our method introduces a multi-level feature aggregation module that enhances the expressive ability of fixed-resolution features by utilizing adjacent features to complement each other. Additionally, we propose a multi-scale information extraction module to capture local contextual information at different scales for back-propagated level-by-level features, which allows for better measurement of the composition of the feature map after back-fusion. To tackle the low confidence issue of boundary pixels, we also introduce a boundary extraction module to extract the boundary information of salient regions. This information is then fused with salient target information to further refine the saliency prediction results. During the training process, our method uses a mixed loss function to constrain the model training from two levels: pixels and images. The experimental results demonstrate that our salient target detection method based on boundary enhancement shows good detection effects on targets of different scales, multi-targets, linear targets, and targets in complex scenes. We compare our method with the best method in four conventional datasets and achieve an average improvement of 6.2% on the mean absolute error (MAE) indicators. Overall, our approach shows promise for improving the accuracy and efficiency of salient object detection in a variety of settings, including those involving 2D/3D semantic analysis and reconstruction/inpainting of image/video/point cloud data.

Abstract Objective As a critical feature of cancers, stemness is acknowledged as a contributor to the development of drug resistance in gastric cancer (GC). LncRNAs have been revealed to participate in this process. In this study, we tried to develop a stemness-related lncRNA pair signature as guidance for clinical decisions. Methods The analysis was initiated by collecting stemness-related lncRNAs in TCGA cohort. The differentially expressed stemness-related lncRNAs between normal and tumor tissues in GC patients from TCGA datasets were further collected to establish the signature based on Lasso and Cox regression analyses. The predictive efficacy of the signature for chemotherapy and immunotherapy was also tested. The practicality of this signature was also validated by Zhongshan cohort. Results A 13-DEsrlncRNA pair-based signature was established. The cutoff point acquired by the AIC algorithm divided the TCGA cohort into high and low risk groups. We found that the low-risk group presented with better survival (Kaplan-Meier analysis, p

A new subcategory, grade 3 neuroendocrine tumors, is incorporated into the grading system of pancreatic neuroendocrine neoplasms in the 2017 WHO c lassification in order to differentiate grade 3 neuroendocrine tumors from neuroendocri ne carcinomas. The 2019 WHO classification extends the concept of grade 3 neuroendo crine tumors to gastrointestinal high-grade neuroendocrine neoplasms. However, there is still limited study focusing on the gastric grade 3 neuroendocrine tumors and gastric neuroendocrine carcinomas. We retrospectively enrolled 151 gastric high-grade neuroendocrine neoplasms patients, who underwent radical resection from January 2007 to December 2015. Clinicopathologic and prognostic features were studied. The Sur veillance, Epidemiology, and End Results (SEER) database was used to verify the prognostic determinants found in the Zhongshan cohort. Neuroendocrine carcinomas showed a higher Ki67 index and higher mitotic count than grade 3 neuroendocrine tumors. We identified 109 (72.2%) patients with neuroendocrine carcinomas, 12 (7.9%) patients with grade 3 neur oendocrine tumors, and 30 (19.9%) patients with mixed neuroendocrine-non-neuroendocrine neoplasms. Although neuroendocrine carcinomas demonstrated higher Ki67 index (P = 0.004) and mitoses (P = 0.001) than grade 3 neuroendocrine tumors, their prognosis afte r radical resection did not demonstrate significant differences (P = 0.709). Tumor size, perineural invasion, and TNM stage were independent prognostic factors of gastric high-g rade neuroendocrine neoplasms.

Abstract L-glutamate is an important component of protein. It can prevent gastrointestinal damage caused by NSAIDs. We constructed two-phase enteric-coated granules of aspirin and L-glutamate compound by extrusion spheronization method and fluidized bed coating. The subliminal effective dose of L-glutamate is 100 mg/kg tested by model of gastric ulcer of rats induced by aspirin and drug administration. HPLC-UV and UV–Vis methods were adopted to determine content and cumulative release of aspirin and L-glutamate as quality analysis method indexes. The prescription and process optimization were carried out with yield, sphericity and dissolution. The two-phase compound granules have good sphericity of 0.93 ± 0.05 (aspirin pellets) and 0.94 ± 0.02 (L-glutamate pellets), content of salicylic acid (0.24 ± 0.03)%, dissolution of aspirin (2.36 ± 0.11)%. Quality evaluation and preliminary stability meet the commercial requirements. The stored environment of compound preparation should be sealed in a cool and dark place.

The physical principle of myelin development is obtained from our previous study by explaining Peter’s quadrant mystery: an externally applied negative and positive E-field can promote and inhibit the growth of the inner tongue of the myelin sheath, respectively. In this study, this principle is considered as a fundamental hypothesis, named Hypothesis-E, to explain more phenomena about myelin development systematically. Specifically, the g-ratio and the fate of the Schwann cell’s differentiation are explained in terms of the E-field. Moreover, an experiment is proposed to validate this theory.

Background: Although low molecular weight heparin (LMWH) is recommended to prevent venous thromboembolism (VTE) in patients with gastric cancer, it's difficult for surgeons to choose appropriate time to start anti-coagulation because of the risk of surgical field bleeding after radical gastrectomy. We compared the risk of VTE and bleeding between patients receiving LMWH within and beyond 48 ​h after surgery. Methods: The medical records of consecutive cases receiving radical gastrectomy from November 1st, 2017 to October 31st, 2018 in Zhongshan Hospital Gastric Cancer Center were carefully reviewed. Patients receiving LMWH within and beyond 48 ​h after surgery were regarded as Early Group (EG) and Delayed Group (DG), respectively. Incidence of VTE and bleeding complications were compared and risk factors of bleeding were evaluated. Results: Six hundred and sixty-five cases were enrolled, including 465 in EG and 200 in DG. No significant differences of clinicopathological or operative features were observed except for fewer combined resection (DG: 5.0%, EG: 1.9%, P ​= ​0.030) and shorter surgery duration (DG: 194min, EG: 168min, P ​

In the development of oligodendrocytes in the central nervous systems, the inner and outer tongue of the myelin sheath tend to be located within the same quadrant, which was named as Peters quadrant mystery. In this study, we conduct in silico investigations to explore the possible mechanisms underlying the Peters quadrant mystery. A biophysically detailed model of oligodendrocytes was used to simulate the effect of the actional potential-induced electric field across the myelin sheath. Our simulation suggests that the paranodal channel connecting the inner and outer tongue forms a low impedance route, inducing two high-current zones at the area around the inner and outer tongue. When the inner tongue and outer tongue are located within the same quadrant, the interaction of these two high-current-zones will induce a maximum amplitude and a polarity reverse of the voltage upon the inner tongue, resulting in the same quadrant phenomenon. This model indicates that the growth of myelin follows a simple principle: an external negative or positive E-field can promote or inhibit the growth of the inner tongue, respectively.

Background Gastric cancer is the third leading cause of cancer-related mortality worldwide. Most tripartite motif (TRIM) family proteins are known as E3 ubiquitin ligases and considerable previous research has revealed the involvement of TRIM proteins in carcinogenesis. TRIM15 is a protein from the TRIM family and the aim of this study is to investigate the role of TRIM15 in gastric cancer. Methods We conducted immunohistochemical staining to examine TRIM15 expression using samples from Zhongshan Hospital of Fudan University. We also conducted transwell assay as well as western blot by using gastric cancer cells. Results The expression of TRIM15 in gastric cancer tissues was higher than normal tissues. Present data demonstrated that high TRIM15 staining intensity had a positive relation to tumor invasion depth (P = 0.007), lymph node metastasis (P = 0.013), distant metastasis (P = 0.031), the tumor-node-metastasis (TNM) staging system (P = 0.002) and shorter overall survival (OS) in gastric cancer patients (P

The fish gut microbiome plays an important role in nutrition absorption and energy metabolism. Studying the gut microbes of cold-water fish is important to understand the dietary adaptation strategies in extreme environments. In this study, the gut samples of Schizothorax wangchiachii (SW, herbivorous), Schizothorax kozlovi (SK, omnivorous), and Percocypris pingi (PP, carnivorous) in the upper Yangtze River were collected, and we sequenced 16S rRNA amplicon to study the potential relationship between gut microbes and host species. The results showed that gut microbial composition and diversity were significantly different between the three cold-water fishes. These fishes had different key taxa in their gut microbes, including bacteria involved in the breakdown of food (e.g., Cetobacterium, Aeromonas, and Clostridium sensu stricto 10). The highest alpha diversity indices (e.g., Chao 1 index) were identified in the herbivore (SW), followed by the carnivore (PP), and the lowest in the omnivore (SK). Non-metric multidimensional scaling (NMDS) results revealed that the gut microbial community of these species was different between host species. The neutral community model (NCM) showed that the microbial community structure of SW was shaped by stochastic processes, and the highest species dispersal was found in SW, followed by PP, and the lowest in SK. The results of niche breadth agreed with these findings. Our results demonstrated that host species influenced the gut microbiome composition, diversity, and microbial community assembly processes of the three cold-water fishes. These findings implied that the variation of gut microbiome composition and function plays a key role in digesting and absorbing nutrients from different foods in cold-water fish.

Different subtypes of gastric cancer differentially respond to immune checkpoint inhibitors (ICI). This study aimed to investigate whether the Estimation of STromal and Immune cells in Malignant Tumor tissues using Expression data (ESTIMATE) algorithm is related to the classification and prognosis of gastric cancer and to establish an ESTIMATE-based gene signature to predict the prognosis for patients. The immune/stromal scores of 388 gastric cancer patients from TCGA were used in this analysis. The upregulated differentially expressed genes (DEGs) in patients with high stromal/immune scores were identified. The immune-related hub DEGs were selected based on protein-protein interaction (PPI) analysis. The prognostic values of the hub DEGs were evaluated in the TCGA dataset and validated in the GSE15460 dataset using the Kaplan-Meier curves. A prognostic signature was built using the hub DEGs by Cox proportional hazards model, and the accuracy was assessed using receiver operating characteristic (ROC) analysis. Different subtypes of gastric cancer had significantly different immune/stromal scores. High stromal scores but not immune scores were significantly associated with short overall survivals of TCGA patients. Nine hub DEGs were identified in PPI analysisThe expression of these hub DEG negatively correlated with the overall survival in the TCGA cohort, which was validated in the GSE15460 cohort. A 9-gene prognostic signature was constructed. The risk factor of patients was calculated by this signature. High-risk patients had significantly shorter overall survival than low-risk patients. ROC analysis showed that the prognostic model accurately identified high-risk individuals within different time frames. We established an effective 9-gene-based risk signature to predict the prognosis of gastric cancer patients, providing guidance for prognostic stratification.

The immune microenvironment plays a critical role in tumor biology. As a critical feature of cancers, stemness is acknowledged as a contributor to the development of drug resistance in gastric cancers (GCs). Long non-coding RNAs (lncRNAs) have been revealed to participate in this process. In this study, we aimed to develop a stemness-related lncRNA signature (SRLncSig) with guiding significance for immunotherapy. Three cohorts (TCGA, Zhongshan, and IMvigor210) were enrolled for analysis. A list of stemness-related lncRNAs (SRlncRNAs) was collected by co-expression strategy under the threshold of coefficient value >0.35 and p-value < 0.05. Cox and Lasso regression analysis was further applied to find out the SRlncRNAs with prognosis-predictive value to establish the SRLncSig in the TCGA cohort. IPS and TIDE algorithms were further applied to predict the efficacy of SRLncSig in TCGA and Zhongshan cohorts. IMvigor210 was composed of patients with clinical outcomes of immunotherapy. The results indicated that SRLncSig not only was confirmed as an independent risk factor for GCs but also identified as a robust indicator for immunotherapy. The patient with a lower SRLncSig score was more likely to benefit from immunotherapy, and the results were highly consistent in three cohorts. In conclusion, our study not only could clarify the correlations between stemness and immunotherapy in GC patients but also provided a model to guide the applications of immunotherapy in clinical practice.

As a new generation computed tomography (CT) technology, spectral CT has great potential in many aspects, especially in the identification and decomposition of materials. To achieve higher accuracy of materials decomposition, we propose a multi-constraint based nonlocal total variation (NLTV) method, named as MCNLTV. Because image-domain based material decomposition belongs to the two-step material decomposition method, the Filter Back-Projection (FBP) algorithm or SART algorithm is used to reconstruct spectral CT images in the first step. Then the material attenuation coefficient matrix is obtained from the reconstruction results. In the second step, MCNLTV regularization is utilized to obtain the material decomposition image. Both simulation experiments and real data experiments are carried out. Experiment results show that the proposed method can obtain higher accuracy of material decomposition than traditional total variation based material decomposition (TVMD), ROF-LLT regularization and direct inverse transformation (DI) for spectral CT.

Abstract Background Systematic interrogation of single-nucleotide variants (SNVs) is one of the most promising approaches to delineate the cellular heterogeneity and phylogenetic relationships at the single-cell level. While SNV detection from abundant single-cell RNA sequencing (scRNA-seq) data is applicable and cost-effective in identifying expressed variants, inferring sub-clones, and deciphering genotype-phenotype linkages, there is a lack of computational methods specifically developed for SNV calling in scRNA-seq. Although variant callers for bulk RNA-seq have been sporadically used in scRNA-seq, the performances of different tools have not been assessed. Results Here, we perform a systematic comparison of seven tools including SAMtools, the GATK pipeline, CTAT, FreeBayes, MuTect2, Strelka2, and VarScan2, using both simulation and scRNA-seq datasets, and identify multiple elements influencing their performance. While the specificities are generally high, with sensitivities exceeding 90% for most tools when calling homozygous SNVs in high-confident coding regions with sufficient read depths, such sensitivities dramatically decrease when calling SNVs with low read depths, low variant allele frequencies, or in specific genomic contexts. SAMtools shows the highest sensitivity in most cases especially with low supporting reads, despite the relatively low specificity in introns or high-identity regions. Strelka2 shows consistently good performance when sufficient supporting reads are provided, while FreeBayes shows good performance in the cases of high variant allele frequencies. Conclusions We recommend SAMtools, Strelka2, FreeBayes, or CTAT, depending on the specific conditions of usage. Our study provides the first benchmarking to evaluate the performances of different SNV detection tools for scRNA-seq data.

The immune microenvironment plays a critical role in tumor biology. The molecular profiles of immune components and related genes are of tremendous value for the study of primary resistance to immune checkpoint blockers (ICBs) for gastric cancer (GC) and serve as prognostic biomarkers to predict GC survival. Recent studies have revealed that tumor immune cell infiltration (ICI) is an indicator of the survival and responsiveness to chemotherapy in GC patients. Here, we describe the immune cell landscape based on the ESTIMATE and CIBERSORT algorithms to help separate GC into 3 ICI clusters using the unsupervised clustering method. Further in-depth analyses, such as differential expression gene (DEG) analysis and principal component analysis (PCA), help to establish an ICI scoring system. A low ICI score is characterized by an increased tumor mutation burden (TMB). The combination of the ICI score and TMB score better predicts the survival of GC patients. Analyses based on public and our own database revealed that the ICI scoring system could also help predict the survival and chemotherapy responsiveness of GC patients. The present study demonstrated that the ICI score may be an effective prognostic biomarker and predictive indicator for chemotherapy and immunotherapy.

Cryopreservation prolongs the storage time of cells and plays an important role in modern biology, agriculture, plant science and medicine. During cryopreservation, cells may suffer many damages, such as osmotic dehydration, large ice puncture and oxidative damages from reactive oxygen species (ROS). Classic cryoprotectants (CPAs) are failing to dispose of ROS, while antioxidants can turn ROS into harmless materials and regulate oxidative stress. The combination of antioxidants and CPAs can improve the efficiency of cryopreservation while negative results may occur by misuse of antioxidants. This paper discussed the feasibility of antioxidants in cryopreservation.

Anomaly detection is the foundation of intelligent operation and maintenance (O&M), and detection objects are evaluated by key performance indicators (KPIs). For almost all computer O&M systems, KPIs are usually the machine-level operating data. Moreover, these high-frequency KPIs show a non-Gaussian distribution and are hard to model, i.e., they are intricate KPI profiles. However, existing anomaly detection techniques are incapable of adapting to intricate KPI profiles. In order to enhance the performance under intricate KPI profiles, this study presents a seasonal adaptive KPI anomaly detection algorithm ASAD (Adaptive Seasonality Anomaly Detection). We also propose a new eBeats clustering algorithm and calendar-based correlation method to further reduce the detection time and error. Through experimental tests, our ASAD algorithm has the best overall performance compared to other KPI anomaly detection methods.

The novel coronavirus SARS-CoV-2 and the disease caused by it, COVID-19, have spread to virtually all countries worldwide within just a few months. The economic and sanitary impact has been enormous. In March 2020, the World Health Organization declared COVID-19 a pandemic. How to effectively prevent and control SARS-CoV-2 transmission while providing care to surgical patients during the pandemic is a crucial topic. In order to minimize the risk of cross-infection between patients and physicians, many hospitals have taken measures to limit outpatient services, elective hospitalizations, and the number of operations. Based on the prevention and control measures stipulated by major medical institutions in China, this overview provides recommendations for surgeons from three aspects: outpatient treatment, ward management and perioperative protection. Telemedicine should be encouraged as a means of social distancing. Outpatient examination should be selected. Reasonable spatial arrangement and effective environmental disinfection are important for ward management. Patient selection for surgery and timing of operations should be carefully discussed within multi-disciplinary teams. Appropriate personal protective equipment should be worn adapted to the situational risk.On December 31, 2019, China reported to the WHO Country Office a pneumonia of unknown cause detected in Wuhan [1,3]. Subsequently, the disease later named COVID-19 affected a substantial proportion of the population in Wuhan and spread to other areas of China. Relying on a nationwide shutdown and mandatory quarantine, China has effectively curtailed the domestic outbreak. However, due to the high transmissibility of SARS-Cov-2 and the mobility of people, COVID-19 spread to the rest of the world. Many hospitals worldwide were faced with confirmed and suspected SARS-Cov-2 infections, putting a huge strain on the safety of patients and employees. Consequently, surgical patients who seek medical care during the COVID-19 pandemic present significant challenges. This paper summarizes medical care and infection prevention and control in general surgery patients during the COVID-19, pandemic in the light of the current situation in China. It provides reference for surgeons and decision makers in health care in other countries suffering from the COVID-19 pandemic.

Abstract Background The safety of laparoscopic total gastrectomy (LTG) for the treatment of gastric cancer remains lack of clinical evidence. The Chinese Laparoscopic Gastrointestinal Surgery Study (CLASS) Group recently launched a multicenter randomized clinical trial (CLASS02–01) to compare the safety of LTG for clinical stage I gastric cancer with the conventional open total gastrectomy (OTG). Methods This CLASS02–01 trial is a prospective, multicenter, randomized, controlled, open, and non-inferiority trial. Two hundred patients who met the inclusion criteria and did not accord with the exclusion criteria will be randomly divided into LTG group (n = 100) and OTG group (n = 100). The primary purpose of this study is to evaluate the early operative morbidity and mortality of LTG compared with OTG for clinical stage I gastric adenocarcinoma. The second purpose is to evaluate the recovery course and compare the postoperative hospital stay of the patients enrolled in this study. Discussion This CLASS02–01 trial is the first prospective randomized two-arm controlled study to determine the safety of LTG compared with OTG. Through this trial, we hope to show that experienced surgeons can safely perform LTG with lymphadenectomy for gastric cancer. Trial registration ClinicalTrials.gov ID: NCT03007550. December 30, 2016.

Milk has two main components that have high nutritional value—milk protein (casein and whey protein), and lactose. These components are extensively used in various areas, especially in food, i.e., as sweeteners, stabilizers, functional food ingredients, nutritional fortifiers, etc. Non-enzymatic browning refers to a series of chemical reactions between sugars and proteins that make food more appetizing. Non-enzymatic browning reactions include degradation of ascorbic acid, lipid peroxidation, caramel reaction, and the Maillard reaction (MR). The MR, as one of the four non-enzymatic browning reactions, has been well studied and utilized in food fields. Milk protein and lactose, as two main components of milk, have high chemical activities; they are used as reactants to participate in the MR, generating Maillard reaction products (MRPs). The MR involves a condensation reaction between carbonyl groups of various sugars and amino groups of amino acids/proteins. These MRPs have different applications in various areas, including food flavor, food oxidation resistance, drug carriers, etc. This work presents the positive and negative effects of the MR, based on the two main components of milk, used in food and medicine, as well as avoidance approaches to prevent the occurrence of negative effects.

Abstract Background A new staging system recently proposed by the IGCA has demonstrated a better capacity of stratifying different prognoses for gastric cancer than the 7th edition AJCC staging system (AJCC7). The aim of this study was to evaluate the efficacy of the IGCA system in Chinese patients. Methods Medical records of patients with gastric cancer who received curative surgery in our center from January 2003 to December 2011 were reviewed retrospectively. All the lesions were staged according to both AJCC7 and IGCA staging systems. Overall survival (OS) of the patients was used as the observation endpoint. Results One thousand five hundred twenty-six cases were included in this study. By comparing the AJCC7 system with the IGCA systems, 395 cases were stratified into different stages, most of which were in stage III. The IGCA system could better stratify stage IIIB and IIIC patients (5-year OS, 38.1% vs. 29.0%; P = 0.005) than the AJCC7 system (5-year OS, 38.2% vs. 35.9%; P = 0.148). T3N3bM0, T4aN2M0 and T4aN3bM0 made up 97.5% (385/395) of the stage shift. T3N3bM0, which was stratified to stage IIIB in the AJCC7 system, showed a significant poorer prognosis than T4aN2M0 and T4aN3aM0, which were staged to IIIB and IIIC in the same system. The improper staging was revised in the IGCA staging system. Conclusions The IGCA staging system can stratify stage III gastric cancer patients more properly than the AJCC7 system.

Tuberculosis (TB) remains a serious global public health problem in the present. TB also affects other sites (extrapulmonary tuberculosis, EPTB), and accounts for a significant proportion of tuberculosis cases worldwide. In order to comprehensively understand epidemiology of EBTB in China, and improve early diagnosis and treatment, we conducted a large-scale multi-center observational study to assess the demographic data and the prevalence of common EPTB inpatients, and further evaluate the prevalence of EPTB concurrent with Pulmonary tuberculosis (PTB) and the associations between multiple EPTB types and gender-age group in China. All consecutive age≥15yr inpatients with a confirmed diagnosis of EPTB during the period from January 2011 to December 2017 were included in the study. The descriptive statistical analysis included median and quartile measurements for continuous variables, and frequencies and proportions with 95% confidence intervals (CIs) for categorical variables. Multinomial logistic regression analysis was used to compare the association of multiple EPTB types between age group and gender. The results showed that the proportion of 15-24 years and 25-34 years in EPTB inpatients were the most and the ratio of male: female was 1.51. Approximately 70% of EPTB inpatients were concurrent with PTB or other types of EPTB. The most common of EPTB was tuberculous pleurisy (50.15%), followed by bronchial tuberculosis (14.96%), tuberculous lymphadenitis of the neck (7.24%), tuberculous meningitis (7.23%), etc. It was found that many EPTB inpatients concurrent with PTB. The highest prevalence of EPTB concurrent with PTB was pharyngeal/laryngeal tuberculosis (91.31%), followed by bronchial tuberculosis (89.52%), tuberculosis of hilar lymph nodes (79.52%), tuberculosis of mediastinal lymph nodes (79.13%), intestinal tuberculosis (72.04%), tuberculous pleurisy (65.31%) and tuberculous meningitis (62.64%), etc. The results from EPTB concurrent with PTB suggested that females EPTB inpatients were less likely to be at higher risk of concurrent PTB (aOR = 0.819, 95%CI:0.803-0.835) after adjusted by age. As age increasing, the trend risk of concurrent PTB decreased (aOR = 0.994, 95%CI: 0.989-0.999) after adjusted by gender. Our study demonstrated that the common EPTB were tuberculous pleurisy, bronchial tuberculosis, tuberculous lymphadenitis of the neck, tuberculous meningitis, etc. A majority of patients with pharyngeal/laryngeal tuberculosis, bronchial tuberculosis, tuberculosis of hilar/mediastinal lymph nodes, intestinal tuberculosis, tuberculous pleurisy, tuberculous meningitis, etc. were concurrent with PTB. Female EPTB inpatients were less likely to be at higher risk of concurrent PTB, and as age increasing, the trend risk of concurrent PTB decreased. The clinicians should be alert to the presence of concurrent tuberculosis in EPTB, and all suspected cases of EPTB should be assessed for concomitant PTB to determine whether the case is infectious and to help for early diagnosis and treatment.

In December 2019, the novel coronavirus pneumonia (COVID-19) occurred in Wuhan, Hubei Province, China. The epidemic quickly broke out and spread throughout the country. Now it becomes a pandemic that affects the whole world. In this study, three models were used to fit and predict the epidemic situation in China: a modified SEIRD (Susceptible-Exposed-Infected-Recovered-Dead) dynamic model, a neural network method LSTM (Long Short-Term Memory), and a GWR (Geographically Weighted Regression) model reflecting spatial heterogeneity. Overall, all the three models performed well with great accuracy. The dynamic SEIRD prediction APE (absolute percent error) of China had been ≤ 1.0% since Mid-February. The LSTM model showed comparable accuracy. The GWR model took into account the influence of geographical differences, with R2 = 99.98% in fitting and 97.95% in prediction. Wilcoxon test showed that none of the three models outperformed the other two at the significance level of 0.05. The parametric analysis of the infectious rate and recovery rate demonstrated that China's national policies had effectively slowed down the spread of the epidemic. Furthermore, the models in this study provided a wide range of implications for other countries to predict the short-term and long-term trend of COVID-19, and to evaluate the intensity and effect of their interventions.

Background: Duodenal stump leakage (DSL) is a serious surgical complication after radical gastrectomy with Roux-en-Y or BillrothII reconstruction. This study was designed to evaluate the effectiveness of laparoscopic single purse-string suture for reinforcement of duodenal stump.Methods: A total of 183 patients harboring gastric adenocarcinoma following laparoscopic radical gastrectomy with Roux-en-Y or BillrothIIreconstruction and single purse-string suture for reinforcement of duodenal stump were retrospectively enrolled from Zhongshan Hospital of Fudan University (Shanghai, China) between January 2014 and December 2016. Operative variables and short-term complications were documented and analyzed. Clavien-Dindo classification system was used to identify surgical complications.Results: Among 183 patients, 108 (59.02%) patients received distal gastrectomy and 75 (40.98%) received total gastrectomy. 88 (48.09%) patients underwent Roux-en-Y reconstruction and 95 (51.91%) patients underwent Billroth-II reconstruction. The mean time of laparoscopic single purse-string suture was 5.01 ± 1.33 min (range from 3.6 to 10.2 min). Postoperative early complication occurred in 26 cases of the patients. There were 4 cases of system-related complications (2.19%), including 3 cases of pulmonary infection (1.64%) and 1 cases of cardiovascular event (0.55%); and 22 cases of surgery-related complications (12.02%), including 6 cases of intra-abdominal infection (3.28%), 4 cases of pancreatic leakage (2.19%), 4 cases of wound complications (2.19%), 3 cases of gastroparesis (1.64%), 2 cases of intra-abdominal bleeding (1.09%), 2 cases of ileus (1.09%), 1 cases of lymphatic leakage (0.55%), and no duodenal stump leakage.Conclusion: Reinforcement on duodenal stump using laparoscopic single purse-string suture during laparoscopic radical gastrectomy is simple and effective and could avoid the incidence of duodenal stump leakage to some extent.

Maintenance of dental health has attracted attention of researchers at present. Various materials have been constructed and applied for curing different dental diseases, although limitation of biocompatibility and safety is still a big challenge. To overcome these limitations, nanomaterials with unique properties are incorporated into various dental treatment materials used in dental applications, including endodontic treatment, periodontal treatment, implant treatment, oral surgery, and restorative treatment, etc. Especially, reactive oxygen species-based nanomaterials equipped with nanoscale properties and reactive oxygen activities can be used as sterilization agents in dentistry, along with being used as good fillers in the dental field. This review concludes the common reactive oxygen species (ROS) nanomaterials and reviews the utilization of ROS in dentistry, highlighting the potential application and safety in clinical treatment. The future prospect will also be proposed to conduct the clinic dental cure.

Due to the lack of an effective method for the treatment of fibrosis, there are numerous patients suffering from the effects of fibrosis. Severe fibrosis can cause dysfunction of relevant organs characterized by excessive deposition of extracellular matrix components. Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) is a long non-coding RNA that is widely expressed and highly conserved in human tissues. It can regulate gene expression at various molecular levels, involved in the fibrosis of the liver, heart, lung, and kidney. In this review, we first described the pathogenesis by which MALAT1 promotes fibrosis. Furthermore, we summarized current studies of MALAT1 in the fibrosis of various organs. Hope this review will contribute to a better understanding of the molecular mechanism of fibrosis and the potential of MALAT1 as a novel therapeutic target for fibrosis.

For a micro-Computed Tomography (Micro-CT) system, the microfocus X-ray tube is an essential component because the spatial resolution of CT images, in theory, is mainly determined by the size and stability of the X-ray focal spot of the microfocus X-ray tube. However, many factors, including voltage fluctuations, mechanical vibrations, and temperature changes, can cause the size and the stability of the X-ray focal spot to degrade. A new microfocus X-ray tube based on a point-like micro-target in which the X-ray target is irradiated with an unfocused electron beam was investigated. EGS4 Monte Carlo simulation code was employed for the calculation of the X-ray intensity produced from the point-like micro-target and the substrate. The effects of several arrangements of the target material, target and beam size were studied. The simulation results demonstrated that if the intensity of X-rays generated at the point-like target is greater than half of the X-ray intensity produced on the substrate, the X-ray focal spot is determined in part by the point-like target rather than by the electron beam in the conventional X-ray tube. In theory, since it is able to reduce those unfavorable effects such as the electron beam trajectory swinging and the beam size changing for the microfocus X-ray tube, it could alleviate CT image artifacts caused by the X-ray focal spot shift and size change.

The objective of this paper is to extract concave and convex feature regions via segmenting surface mesh of a mechanical part whose surface geometry exhibits drastic variations and concave-convex features are equally important when modeling. Referring to the original approach based on the minima rule (MR) in cognitive science, we have created a revised minima rule (RMR) and presented an improved approach based on RMR in the paper. Using the logarithmic function in terms of the minimum curvatures that are normalized by the expectation and the standard deviation on the vertices of the mesh, we determined the solution formulas for the feature vertices according to RMR. Because only a small range of the threshold parameters was selected from in the determined formulas, an iterative process was implemented to realize the automatic selection of thresholds. Finally according to the obtained feature vertices, the feature edges and facets were obtained by growing neighbors. The improved approach overcomes the inherent inadequacies of the original approach for our objective in the paper, realizes full automation without setting parameters, and obtains better results compared with the latest conventional approaches. We demonstrated the feasibility and superiority of our approach by performing certain experimental comparisons.

A pre-patient attenuator ("bowtie filter" or "bowtie") is used to modulate an incoming x-ray beam as a function of the angle of the x-ray with respect to a patient to balance the photon flux on a detector array. While the current dynamic bowtie design is focused on fan-beam geometry, in this study we propose a methodology for dynamic bowtie design in multi-slice/cone-beam geometry. The proposed 3D dynamic bowtie is an extension of the 2D prior art. The 3D bowtie consists of a highly attenuating bowtie (HB) filled in with heavy liquid and a weakly attenuating bowtie (WB) immersed in the liquid of the HB. The HB targets a balanced flux distribution on a detector array when no object is in the field of view (FOV). The WB compensates for an object in the FOV, and hence is a scaled-down version of the object. The WB is rotated and translated in synchrony with the source rotation and patient translation so that the overall flux balance is maintained on the detector array. First, the mathematical models of different scanning modes are established for an elliptical water phantom. Then, a numerical simulation study is performed to compare the performance of the scanning modes in the cases of the water phantom and a patient cross-section without any bowtie and with a dynamic bowtie. The dynamic bowtie can equalize the numbers of detected photons in the case of the water phantom. In practical cases, the dynamic bowtie can effectively reduce the dynamic range of detected signals inside the FOV. Furthermore, the WB can be individualized using a 3D printing technique as the gold standard. We have extended the dynamic bowtie concept from 2D to 3D by using highly attenuating liquid and moving a scale-reduced negative copy of an object being scanned. Our methodology can be applied to reduce radiation dose and facilitate photon-counting detection.

Aberrant chemokine (C-X-C motif) receptor CXCR4 expressions in malignant tissues have been reported, but its role in gastric cancer prognosis remains unknown. Our studies were designed to investigate the expression and prognostic significance of CXCR4 in patients with gastric cancer. CXCR4 expression was retrospectively analyzed by immunohistochemistry in 97 patients with gastric adenocarcinoma from China. Results were assessed for association with clinical features and overall survival by using Kaplan-Meier analysis. Prognostic values of CXCR4 expression and clinical outcomes were evaluated by Cox regression analysis. A molecular prognostic stratification scheme incorporating CXCR4 expression was determined by using receiver operating characteristic (ROC) analysis. The results show that CXCR4 predominantly localized in the cell membranes and cytoplasm. The protein level of CXCR4 was upregulation in gastric cancer tissues and upregulated expression of CXCR4 was only significantly associated with Lauren classification (P