Search

MotivationUsing DNA as the medium to store information has recently been recognized as a promising solution for long-term data storage. While several system prototypes have been demonstrated, the error characteristics in DNA data storage are discussed with limited content. Due to the data and process variations from experiment to experiment, the error variation and its effect on data recovery remain to be uncovered. To close the gap, we systematically investigate the storage channel, i.e., error characteristics in the storage process.ResultsWe first propose a new concept named sequence corruption to unify the error characteristics into the sequence level, easing the channel analysis. Then we derived the formulations of the data imperfection at the decoder including both sequence loss and sequence corruption, revealing the decoding demand and monitoring the data recovery. Furthermore, we extensively explored several data-dependent unevenness observed in the base error patterns and studied a few potential factors and their impacts on the data imperfection at the decoder both theoretically and experimentally. The results presented here introduce a more comprehensive channel model and offer a new angle towards the data recovery issue in DNA data storage by further elucidating the error characteristics of the storage process.Contactpoh.chuehloo@nus.edu.sg

Multigoal reinforcement learning (RL) extends the typical RL with goal-conditional value functions and policies. One efficient multigoal RL algorithm is the hindsight experience replay (HER). By treating a hindsight goal from failed experiences as the original goal, HER enables the agent to receive rewards frequently. However, a key assumption of HER is that the hindsight goals do not change the likelihood of the sampled transitions and trajectories used in training, which is not the fact according to our analysis. More specifically, we show that using hindsight goals changes such a likelihood and results in a biased learning objective for multigoal RL. We analyze the hindsight bias due to this use of hindsight goals and propose the bias-corrected HER (BHER), an efficient algorithm that corrects the hindsight bias in training. We further show that BHER outperforms several state-of-the-art multigoal RL approaches in challenging robotics tasks.

Introduction Early systemic and central nervous system viral replication and inflammation may impact brain integrity in people with HIV (PWH), leading to chronic cognitive symptoms not fully reversed by antiretroviral therapy (ART). This study examined associations between cognitive performance and markers of CNS injury associated with acute HIV infection and ART. Methods HIV-infected MSM and transgender women (average age: 27 and education: 13 years) enrolled within 100 days from estimated date of detectable infection [EDDI]. A cognitive performance (NP) protocol was administered at enrollment (before ART initiation) and every 24 weeks until week 192. An overall index of cognitive performance (NPZ) was created using local normative data. Blood (n=87) and cerebrospinal fluid (CSF; n=29) biomarkers of inflammation and neuronal injury were examined before ART initiation. Regression analyses assessed relationships between time since EDDI, pre-ART biomarkers, and NPZ. Results Adjusting for multiple comparisons, shorter time since EDDI was associated with higher pre-ART VL and multiple biomarkers in plasma and CSF. NPZ scores were within the normative range at baseline (NPZ=.52) and at each follow-up visit, with a modest increase through week 192. Plasma or CSF biomarkers were not correlated with NP scores at baseline or after ART. Conclusions Biomarkers of CNS inflammation, immune activation and neuronal injury peak early and then decline during acute HIV infection, confirming and extending results of other studies. Neither plasma nor CSF biomarkers during acute infection corresponded to NP scores before or after sustained ART in this cohort with few psychosocial risk factors for cognitive impairment.

III-nitride quantum dot (QD) lasers have potential applications in visible light regime, yet one of its most important dynamic parameters, the linewidth enhancement factor (LEF), has been rarely studied. We studied the LEF of InGaN QD lasers with a comprehensive theory model and found that the LEF values of InGaN QD lasers can be smaller and more robust against inhomogeneous broadening than those of InGaAs QD lasers. In addition, the LEF improvement from quantum well to QD in III-nitride semiconductors is much larger than that in III-arsenide. It is believed that the large confinement energy in III-nitrides and atomic-like density of states of QDs contribute to these advantages of III-nitride QD lasers. Our results not only reveal the advantages of using III-nitride QD for high spectral purity, low-chirp laser applications, but also provide a general guideline for designing QD lasers with lower LEF values in various semiconductor materials.

Image based Geo-localization is estimating the location of a query image by matching it to a large amount of images in Geo-tagged database. This matching task is very challenging due to the vast differences in visual appearance or modality of image pairs on different platforms, e.g., one image from the RGB camera, the other from the light detection and ranging (LIDAR) sensor. The spatial layout of the scene can provide important clues and significantly reduce matching ambiguity. Therefore, we propose a novel deep network that embeds spatial configuration of the scenes into feature representation. Specifically, we design a spatial-scale attention (SSA) module to highlight the salience correspondence layout features at different scales. The encoded features not only represent the emergence of certain objects, but also reflect the relative locations of the objects. By this way, we learn more discriminative deep feature representations, leading to a higher recall. The experimental results on two standard cross-view benchmark datasets (CVUSA and CVACT) and a cross-modal dataset (GRAL) demonstrate that our method performs better than the state-of-the-art methods. Remarkably, the recall rate@top-1 improves from 27.6% in [1] to 40.5% on the GRAL dataset.

We present a vertically aligned multiwall carbon nanotube (VA-MWCNT) based humidity sensor with low cost and simple fabrication method compatible with mass production. The VA-MWCNT is well aligned and possesses an interconnect topography. Initially, the VA-MWCNT showed superhydrophobic characteristics, which are not suitable for humidity sensing. Therefore, to protect the topography of the material and make the surface more hydrophilic, a low cost and easy handling treatment method was adopted with O2 plasma surface functionalization. The fabricated device showed good linearity at the range of 40% RH–90% RH, low hysteresis ( H = 1.5%RH), fast response (3.3 s), high stability ( Cv Rr _max 0.6% ), demonstrating excellent performance in both static and dynamic measurements. The high performance and low-cost fabrication methods make the VA-MWCNT-based sensor a promising candidate for the application of humidity sensing. Moreover, benefiting from the simple fabrication process, it can be easily extended to resistive sensing fusion with the integration of the temperature sensor and flow sensor for more complex applications.

With the rapid development of e-Commerce and takeaway platforms, retailers have gradually developed multi-channel operations. However, limited empirical studies explored the effects of an online channel offered by takeaway platforms on the store performance. Does an online channel addition have a synergy effect or a cannibalization effect on store performance? We empirically investigate these effects by analyzing a large dataset including diverse samples collected from multiple retailers across various regions. The dataset includes 2115 stores across 25 retailers for 10 months that includes two types of retail formats and covers 16 provinces and 21 cities of China. We study the impacts of the newly introduced online channel on the incumbent offline channel and the overall store performance. The empirical results reveal that the online channel addition mainly has a synergy effect. Specifically, for the overall store, it has a positive effect on the sales and product variety, whereas it has a negative effect on the basket size. Surprisingly, an online channel addition also has a positive effect on the offline sales. Our study adds novel values to multi-channel retailing literature by empirically researching the cannibalization and synergy effect of a new type of online channel, offered by takeaway platforms. It can provide insights for retail enterprises who are interested in introducing O2O model.

This paper applies mechanism design to the supply chain enterprise’s pollution abatement problem with carbon tax. To maximize the government’s expected utility, an uncertain contract model is presented in the framework of principal-agent theory, where the government’s assessment of the supply chain enterprise’s carbon emission level is described as an uncertain variable. Afterwards, the equivalent model is provided to obtain the optimal contract for the uncertain pollution abatement problem. The results demonstrate that the supply chain enterprise’s optimal output decreases with the carbon emission level. Furthermore, the government’s optimal transfer payment decreases with the carbon emission level if the carbon tax is low. In contrast, if the carbon tax is high, the optimal transfer payment increases with the carbon emission level. In addition, an increase in the carbon emission level decreases the optimal utilities of both the government and the supply chain enterprise and also leads to the supply chain enterprise’s incremental marginal utility. Finally, we provide a numerical example, which illustrates the effectiveness and practicability of the proposed model.

There are different paradigms in educational technology. Under the background of big data era, data science, learning analysis and education have made great achievements. In the field of education under big data, all kinds of new paradigms are constantly emerging and have achieved very good results in actual education. In the era of education big data, how to fully tap the value of big data for online education practice, decision-making, evaluation and research, and how to avoid the risk of big data are important issues in the current education reform and development. This paper analyzes the application of the current scientific paradigm in education, constructs the construction paradigm of online education evaluation model, and puts forward a new education concept in order to promote the development of the new paradigm of big data online education technology research. Applying this paradigm, a series of educational evaluation models are constructed from the macro, miso and micro levels, which play a positive role in the research, decision-making, practice and evaluation of related fields.

With the development of the digital economy, the emerging marketing strategy of the e-commerce flash sales has been changing the traditional purchasing habits of customers. This imposes new decision-making challenges for companies involved in flash sales. It is important for companies to build the accurate product demand forecast analysis focusing on the characteristics of the flash sales and customer behaviors. In this paper, VIPS (Weipinhui, a Chinese e-commerce platform) is taken as a case study with the key focus on how sentiment factors in customer reviews affect product demand in flash sale platforms. The paper adopts two sentiment analysis methods based on emotional dictionaries. The method with a higher evaluation index is adopted to integrate the emotional factors into the autoregressive model for product demand and assessment. The experiments prove that the autoregressive model for integrating the sentiment factors demonstrates better forecasting performances than the models without sentiment factors. The experiments further confirm that when product demand for the previous two weeks and customer review sentiment factors in the previous week are taken into consideration, demand forecast effects are most accurate.

Objective: The goal of the present study was to determine the expression of yes-associated protein 1 (YAP1) in renal tissues of mice with lupus nephritis (LN) and elucidate its role in the progression of renal fibrosis. Methods: C57BL/6 mice and MRL/lpr mice were selected for experimental comparison. Mouse kidney tissues were removed and sectioned for hematoxylin and eosin staining, Masson’s trichome staining, Sirius staining, and immunohistochemistry. The mRNA and protein levels of YAP1 in mouse kidney tissues were detected, and the correlation between YAP1 and fibronectin (FN) mRNA levels was analyzed. Mouse renal epithelial cells were used for in vitro experiments. After transfection and stimulation, the cells were divided into 4 groups, namely the C57BL/6 serum group (group 1), the MRL/lpr serum group (group 2), the MRL/lpr serum + siRNA-negative control group (group 3), and the MRL/lpr serum + siRNA-YAP1 group (group 4). Epithelial-mesenchymal transition (EMT) markers in each group were detected by Western blotting and immunofluorescence staining. Serum creatinine, blood urea nitrogen, and urinary protein levels were detected and assessed for their correlation with YAP1 mRNA levels by Spearman’s analysis. Results: Compared to C57BL/6 mice, MRL/lpr mice exhibited obvious changes in fibrosis in renal tissues. In addition, YAP1 expression was significantly higher in the renal tissues of MRL/lpr mice than in those of C57BL/6 mice, and YAP1 mRNA levels were positively correlated with those of FN. YAP1 silencing in lupus serum-stimulated cells could effectively relieve serum-induced EMT. Finally, we observed that YAP1 mRNA levels in mouse kidney tissue were significantly and positively correlated with the degree of renal function injury. Conclusion: YAP1 expression in the kidney tissues of LN mice was higher than that observed in normal mice, indicating that YAP1 may play an important role in the occurrence and development of LN.

In recent years, research into 3D shape recognition in the field of multimedia and computer vision has attracted wide attention. With the rapid development of deep learning, various deep models have achieved state-of-the-art performance based on different representations. There are many modalities for representing a 3D model, such as point cloud, multiview, and panorama view. Deep learning models based on these different modalities have different concerns, and all of them have achieved high performance for 3D shape recognition. However, all of these methods ignore the multimodality information in conditions where the same 3D model is represented by different modalities. Thus, we can obtain a better descriptor by guiding the training to consider these multiple representations. In this article, we propose MMFN, a novel multimodal fusion network for 3D shape recognition that employs correlations between the different modalities to generate a fused descriptor, which is more robust. In particular, we design two novel loss functions to help the model learn the correlation information during training. The first is correlation loss, which focuses on the correlations among different descriptors generated from different structures. This approach reduces the training time and improves the robustness of the fused descriptor of the 3D model. The second is instance loss, which preserves the independence of each modality and utilizes feature differentiation to guide model learning during the training process. More specifically, we use the weighted fusion method, which applies statistical methods to obtain robust descriptors that maximize the advantages of the information from the different modalities. We evaluated the proposed method on the ModelNet40 and ShapeNetCore55 datasets for 3D shape classification and retrieval tasks. The experimental results and comparisons with state-of-the-art methods demonstrate the superiority of our approach.

To determine the efficacy and safety of Endostar combined with gemcitabine and cisplatin in the treatment of NSCLC, and provide evidence-based reference for clinical drug use. Retrieved from Cochrane Library, PubMed, Embase, ClinicalTrials, CNKI, Wanfang and VIP database, randomized controlled trials (RCT) were collected. Pooled standard mean differences (SMD) and 95% confidence intervals (CI) were analyzed using R software with a random-effects model. Data from a total of 27 RCTs were included, involving 1646 patients were analyzed. Results of meta-analysis showed that response rate and clinical benefit rate of trial group were significantly higher than those of control group. There was no statistical significance in the incidence of leucopenia, thrombocytopenia and gastrointestinal reaction between 2 groups. The results showed that Endostar combined with gemcitabine and cisplatin may generally improve therapeutic efficacy of NSCLC patients, without increasing the incidence of adverse reactions.

Emotion recognition has many potential applications in the real world. Among the many emotion recognition methods, electroencephalogram (EEG) shows advantage in reliability and accuracy. However, the individual differences of EEG limit the generalization of emotion classifiers across subjects. Moreover, due to the nonstationary characteristic of EEG, the signals of one subject change over time, which is a challenge to acquire models that could work across sessions. In this article, we propose a novel domain adaptation method to generalize the emotion recognition models across subjects and sessions. We use neural networks to implement the emotion recognition models, which are optimized by minimizing the classification error on the source while making the source and the target similar in their latent representations. Considering the functional differences of the network layers, we use adversarial training to adapt the marginal distributions in the early layers and perform association reinforcement to adapt the conditional distributions in the last layers. In this way, we approximately adapt the joint distributions by simultaneously adapting marginal distributions and conditional distributions. The method is compared with multiple representatives and recent domain adaptation algorithms on benchmark SEED and DEAP for recognizing three and four affective states, respectively. The experimental results show that the proposed method reaches and outperforms the state of the arts.

BackgroundSevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an emerging coronavirus that has resulted in more than 2 000 000 laboratory-confirmed cases including over 145 000 deaths. Although SARS-CoV-2 and SARS-CoV share a number of common clinical manifestations, SARS-CoV-2 appears to be highly efficient in person-to-person transmission and frequently causes asymptomatic or presymptomatic infections. However, the underlying mechanisms that confer these viral characteristics of high transmissibility and asymptomatic infection remain incompletely understood.MethodsWe comprehensively investigated the replication, cell tropism, and immune activation profile of SARS-CoV-2 infection in human lung tissues with SARS-CoV included as a comparison.ResultsSARS-CoV-2 infected and replicated in human lung tissues more efficiently than SARS-CoV. Within the 48-hour interval, SARS-CoV-2 generated 3.20-fold more infectious virus particles than did SARS-CoV from the infected lung tissues (P < .024). SARS-CoV-2 and SARS-CoV were similar in cell tropism, with both targeting types I and II pneumocytes and alveolar macrophages. Importantly, despite the more efficient virus replication, SARS-CoV-2 did not significantly induce types I, II, or III interferons in the infected human lung tissues. In addition, while SARS-CoV infection upregulated the expression of 11 out of 13 (84.62%) representative proinflammatory cytokines/chemokines, SARS-CoV-2 infection only upregulated 5 of these 13 (38.46%) key inflammatory mediators despite replicating more efficiently.ConclusionsOur study provides the first quantitative data on the comparative replication capacity and immune activation profile of SARS-CoV-2 and SARS-CoV infection in human lung tissues. Our results provide important insights into the pathogenesis, high transmissibility, and asymptomatic infection of SARS-CoV-2.

Cross-media 3D model recognition is an important and challenging task in computer vision, which can be utilized in many applications such as landmark detection, image set classification, etc. In recent years, with the development of deep learning, many approaches have been proposed to handle the 3D model recognition problem. However, all of these methods focus on the structure information representation and the multi-view information fusion, and ignore the spatial and temporal information. So that it is not suitable for the cross-media 3D model recognition. In this paper, we utilize the sequence views to represent each 3D model and propose a novel Multi-view Convolutional LSTM Network (MVCLN), which utilizes the LSTM structure to extract temporal information and applies the convolutional operation to extract spatial information. More especially, the spatial and temporal information both are considered during the training process, which can effectively utilize the differences between the view's spatial information to improve the final performance. Meanwhile, we also introduce the classic attention model to define the weight of each view, which can reduce the redundant information of view's spatial information in the information fusion step. We evaluate the proposed method on the ModelNet40 for 3D model classification and retrieval task. We also construct a dataset utilizing the overlap categories of MV-RED, ShapenetCore and ModelNet to demonstrate the effectiveness of our approach for the cross-media 3D model recognition. Experimental results and comparisons with the state-of-the-art methods demonstrate that our framework can achieve superior performance.

Sparse principal component analysis (PCA) is used to obtain stable and interpretable principal components (PCs) from high-dimensional data. A robust sparse PCA method is proposed to handle potentia...

In population-based health research, the so-called population attributable fraction is an important quantity that calculates the percentage of excess risk of morbidity and mortality associated with modifiable risk factors for a given population. While the concept of “risk” is usually measured by event probabilities, in practice it may be of a more direct interest to know the excess life expectancy associated with the modifiable risk factors instead, particularly when mortality is of the ultimate concern. In this paper, we thus propose to study a novel quantity, termed “attributable life expectancy,” to measure the population attributable fraction of life expectancy. We further develop a model-based approach for the attributable life expectancy under the Oakes–Dasu proportional mean residual life model, and establish its asymptotic properties for inferences. Numerical studies that include Monte-Carlo simulations and an actual analysis of the mortality associated with smoking cessation in an Asia Cohort Consortium are conducted to evaluate the performance of our proposed method.

There is an urgent need for animal models of COVID-19 to study immunopathogenesis and test therapeutic intervenes. In this study we showed that NSG mice engrafted with human lung (HL) tissue (NSG-L mice) could be infected efficiently by SARS-CoV-2, and that live virus capable of infecting Vero cells was found in the HL grafts and multiple organs from infected NSG-L mice. RNA-seq examination identified a series of differentially expressed genes, which are enriched in viral defense responses, chemotaxis, interferon stimulation, and pulmonary fibrosis between HL grafts from infected and control NSG-L mice. Furthermore, when infecting humanized mice with human immune system (HIS) and autologous HL grafts (HISL mice), the mice had bodyweight loss and hemorrhage and immune cell infiltration in HL grafts, which were not observed in immunodeficient NSG-L mice, indicating the development of anti-viral immune responses in these mice. In support of this possibility, the infected HISL mice showed bodyweight recovery and lack of detectable live virus at the later time. These results demonstrate that NSG-L and HISL mice are susceptible to SARS-CoV-2 infection, offering a useful in vivo model for studying SARS-CoV-2 infection and the associated immune response and immunopathology, and testing anti-SARS-CoV-2 therapies.

Monolayer (ML)-scale GaN/AlN multiple quantum well (MQW) structures for electron-beam-pumped ultraviolet (UV) emitters are grown on c-sapphire substrates by using plasma-assisted molecular beam epitaxy under controllable metal-rich conditions, which provides the spiral growth of densely packed atomically smooth hillocks without metal droplets. These structures have ML-stepped terrace-like surface topology in the entire QW thickness range from 0.75–7 ML and absence of stress at the well thickness below 2 ML. Satisfactory quantum confinement and mitigating the quantum-confined Stark effect in the stress-free MQW structures enable one to achieve the relatively bright UV cathodoluminescence with a narrow-line (~15 nm) in the sub-250-nm spectral range. The structures with many QWs (up to 400) exhibit the output optical power of ~1 W at 240 nm, when pumped by a standard thermionic-cathode (LaB6) electron gun at an electron energy of 20 keV and a current of 65 mA. This power is increased up to 11.8 W at an average excitation energy of 5 µJ per pulse, generated by the electron gun with a ferroelectric plasma cathode at an electron-beam energy of 12.5 keV and a current of 450 mA.

Machine learning has allowed many systems that we interact with to improve performance and personalize. Recommender systems in particular are one of the largest users of machine learning in production environments that have improved performance of real-world systems. Learning in these interactive systems requires models that combine very diverse signals, including the logs of the interactive system (indicating if the intervention succeeded or failed) augmented with other data sources including: collaborative filtering, text, and image data. Bayesian inference is a compelling method to combine these diverse signals in a principled manner, but deployment of systems based on Bayesian principles remain challenging. The reward signal in the system logs is often uneven. Accurate estimation of reward is possible for exploiting actions, but often poor for other actions (exploration). Non-Bayesian methods such as inverse propensity score methods, the reinforce algorithm, and other heuristic-based approaches currently dominate practice. These commonly-used heuristics are often ineffective at leveraging diverse data. In contrast, Bayesian methods offer a principled, robust framework for learning from uneven signals and combining different types of information. Drawing upon the bandit and reinforcement learning community, in this workshop we will explore innovations in Bayesian inference for real world interactive systems, and consider advantages and limitations of the Bayesian approach.

The diagnosis and treatment of depression is very important since it brings a heavy burden to family and society. Because of the high sensitivity, relatively low cost, and convenient recording, electroencephalogram (EEG) has become an important tool for monitoring brain activity and is gradually being used in the auxiliary diagnosis of mental diseases. EEG signals are typically non-linear and non-stationary. Therefore, they are suitable to be dealt with by time-frequency analysis technique. In this paper, we propose a strategy that combines the time-frequency analysis technique and temporal convolution network for depression recognition. Firstly, we use the method of intrinsic time-scale decomposition to decompose each EEG recording to several components. And secondly, some statistical indices are calculated from the instantaneous amplitudes and instantaneous frequencies of these components to form the feature vectors. Thirdly, an improved temporal convolution network (TCN) is used to detect the depression from normal controls. Temporal convolution network is not only suitable for the sequence model, but also retains the parallel computing characteristics of the convolutional neural network. To improve the model performance, we further modify the original softmax loss of TCN as L-softmax. Experiments show the effectiveness of the proposed model. Furthermore, we find that the depressive patients and normal controls shows different patterns through functional connectivity analysis. Our analysis results can be used as an auxiliary tool to help psychiatrists diagnose patients with depression.

Background:Military personnel are widely exposed to risk factors for heat-related illnesses. Knowledge, attitudes, and practices (KAP) are three of the most important means by which to prevent such illnesses, but there has been a lack of investigations into and correlation analyses of KAP. This study aimed to explore the heat-related KAP of military personnel in China.Methods:We conducted a cross-sectional study (June 1-25, 2019). A total of 646 military personnel were recruited from two Chinese Navy troops in the tropical zone and one troop in the temperate zone. We collected data on demographic characteristics and KAP scores using questionnaires. Univariate analysis and Scheffe's method were used for data analyses.Results:The mean KAP scores were 10.37 (range = 3–13, standard deviation = 1.63) for knowledge (K-score), 7.76 (range = 0–16, SD = 2.65) for attitudes (A-score), and 3.80 (range = 1–6, SD = 1.12) for practices (P-score). There were noticeable differences in mean K-score according to age, military rank, and educational level (P< 0.05). Participants from the tropical zone had higher A-scores (P< 0.05) and higher P-scores (P< 0.001) than those from the temperate zone. Additionally, participants with relevant experience also had higher A-scores (P< 0.05) than those without such experience.Conclusions:Military personnel's awareness of preventive and first-aid measures against heat-related illnesses need to be strengthened. It will be very important to develop educational programmes and enrich systematic educational resources to raise this awareness.

Background Endometriosis is a chronic hormonal inflammatory disease characterized by the presence of endometrial tissue outside the uterus. Endometriosis often causes infertility, which brings physical and mental pain to patients and their families. Methods We examined the functions of heat shock factor 1 (HSF1) in endometriosis development through cell count assay, cell-scratch assay and clone formation experiments. We used quantitative real-time PCR (qRT-PCR) and Western blot (WB) to detect HSF1 expression. Glucose and lactate levels were determined using a glucose (GO) assay kit and a lactate assay kit. Furthermore, we used a HSF1 inhibitor-KRIBB11 to establish a mouse model of endometriosis. Results Our data demonstrated that HSF1 promoted endometriosis development. Interestingly, HSF1 enhanced glycolysis via up-regulating PFKFB3 expression in endometriosis cells, which was a key glycolysis enzyme. Consistently, the HSF1 inhibitor KRIBB11 could abrogate endometriosis progression in vivo and in vitro. Conclusions Findings indicate that HSF1 plays an important role in endometriosis development, which might become a new target for the treatment of endometriosis. Electronic supplementary material Supplementary data are available.

Background: The COVID-19 global pandemic has posed unprecedented challenges to health care systems all over the world. The speed of the viral spread results in a tsunami of patients, which begs for a reliable screening tool using readily available data to predict disease progression.Methods: Multicenter retrospective cohort study was performed to develop and validate a triage model. Patient demographic and non-laboratory clinical data were recorded. Using only the data from Zhongnan Hospital, step-wise multivariable logistic regression was performed, and a prognostic nomogram was constructed based on the independent variables identifies. The discrimination and calibration of the model were validated. External independent validation was performed to further address the utility of this model using data from Jinyintan Hospital.Results: A total of 716 confirmed COVID-19 cases from Zhongnan Hospital were included for model construction. Men, increased age, fever, hypertension, cardio-cerebrovascular disease, dyspnea, cough, and myalgia are independent risk factors for disease progression. External independent validation was carried out in a cohort with 201 cases from Jinyintan Hospital. The area under the curve (AUC) was 0.787 (95% confidence interval [CI]: 0.747–0.827) in the training group and 0.704 (95% CI: 0.632–0.777) in the validation group.Conclusions: We developed a novel triage model based on basic and clinical data. Our model could be used as a pragmatic screening aid to allow for cost efficient screening to be carried out such as over the phone, which may reduce disease propagation through limiting unnecessary contact. This may help allocation of limited medical resources.

Background: The relationship of the area of the aqueduct on quantification of the aqueductal stroke volume (SV) and max velocity need further investigation. Our aim was to assess the influence of the area of the aqueduct on quantification of the aqueductal SV and max velocity measured with phase contrast magnetic resonance imaging (PC-MRI) within the cerebral aqueduct at the level of the intercollicular sulcus.Materials and Methods: Nine healthy volunteers (mean age 29.6 yrs) were enrolled in the study and brain MRIs were performed on a 3.0T system. Quantitative analysis of aqueductal cerebrospinal fluid (CSF) flow was performed using manual regions of interest (ROI) placement. ROIs were separately drawn for each of 12 phases of the cardiac cycle, and changes in aqueduct size during the cardiac cycle were determined. Stroke volumes were calculated uses the first and ninth aqueductal ROIs and compared to each other. Max velocities at the 12 phases were also collected, and the relationship between the area and max velocity and the impact on SV were analyzed.Results: There was variation in the size of the aqueduct during the cardiac cycle, the first area（S1）was larger than the ninth（S9）. The first max velocity（Vmax1）was less than the ninth（Vmax9）. Additionally, there was a significant different between the stroke volume calculated using the first aqueductal ROI (SV1) and the ninth（SV9）. Conclusions: There is variation in the size of the cerebral aqueduct which is used to calculate stroke volume and other CSF flow parameters during the cardiac cycle. The maximum velocity may be inversely proportional to the area of ​​the aqueduct. In order to establish reliable reference values for CSF flow parameters in future studies, a variable ROI, to account for cardiac cycle variation, should be considered and incorporated.

Non-suicidal self-injury (NSSI) refers to any intentional, self-inflicted behavior that causes direct damage to body tissues (Kerr et al., 2010), and has emerged as a challenging public health issue worldwide, especially among adolescents. The most common presentations of NSSI include skin-cutting, severe scratching, and burning (Whitlock et al., 2006). In a community-based study conducted in England, 6% of 5506 adolescents reported self-harm in 2014, and 1320 adolescents (74% aged 15‒17 years) presented to hospitals due to non-fatal self-harm (Geulayov et al., 2018). In Australia, 8% of adolescents aged less than 20 years reported a history of self-harm (Hiscock et al., 2018). In addition, the prevalence of NSSI has significantly increased in recent years. It was reported by McManus et al. (2019) that the prevalence of self-reported lifetime NSSI increased from 2.4% in 2000 to 6.4% in 2014, with the largest increase in females aged 16–24 years (from 6.5% to 19.7%). Researchers have also proposed an "iceberg" model of self-harm, indicating that the phenomenon of NSSI is often concealed and that most individuals with NSSI do not seek help from others (McMahon et al., 2014; Geulayov et al., 2018).

Zebra crossing segmentation is an important part of blind navigation. In this paper, We proposed a method based on Dilated Convolution Neural Network for accurate recognition of zebra crossing. The Dilated Convolution Neural Network is used to automatically identify all the regions that meet the characteristics of zebra crossing. After introducing residual network and Dilated Convolution, the detection effect of the model is greatly enhanced. The improved model improves the accuracy and reliability of zebra crossing recognition, and reduces the false detection rate. Through sample training and data testing, the results show that the Dilated Convolution Neural Network can recognize zebra crossing with an accuracy of 91.24%.

AIM: To perform a Meta-analysis to explore the correlation between metabolic syndrome and intraocular pressure (IOP). METHODS: We searched PubMed and Embase in November 2017 for studies discussing the relationship between metabolic syndrome components and IOP in patients. Pearson correlation coefficients, odds ratios and standardized betas were extracted from inclusive studies. Heterogeneity and publication bias were checked. RESULTS: Of 295 articles, 10 met inclusion criteria and provided sufficient data for Meta-analysis. Results showed a significant positive relation between metabolic syndrome and IOP (Z=0.47, 95%CI: 0.15-0.79, P=0.005). The five components [waist circumference, hypertriglyceridemia, high blood pressure, high fasting glucose and low high density lipoprotein (HDL)-cholesterol] of metabolic syndrome all showed positive correlation with IOP except the low HDL-cholesterol which had no statistical significance. The pooled Z was 0.08 (95%CI: 0.04-0.12), 0.16 (95%CI: 0.11-0.21), 0.16 (95%CI: 0.10-0.22), 0.30 (95%CI: 0.20-0.40) and 0.12 (95%CI: 0.08-0.16), respectively. Begg's test and Egger's test showed no evidence of significant publication bias of this Meta-analysis. CONCLUSION: Our findings suggest that metabolic syndrome and its components are significantly associated with IOP, besides the HDL-cholesterol. This association may be used to control IOP by intervening the occurrence of metabolic syndrome.

Background: Infectious keratitis after refractive surgery is uncommon, and mycobacterium abscess has never been observed in corneal infections in China. Purpose: To propose an unusual presentation of two cases about Mycobacterium abscess infection after laser in situ keratomileusis (LASIK). Participants: Both cases 1 and 2 were females who exhibited a corneal ulceration after LASIK in the same hospital. Methods: The first patient in the study was a 28-year-old woman who underwent Epi-LASIK surgery in both eyes. She developed an infectious corneal ulcer in one eye after 50 days post-surgery. She was referred to our out-patient clinic 3 months post-surgery with corneal perforation, and was treated with a therapeutic penetrating keratoplasty. The second patient was a 27-year-old woman who developed infectious keratitis in one eye after 2 months after bilateral LASIK and was referred to us 3 months later. Results: Cultures in both the cases were identified as a Mycobacterium abscess, and case 1 was treated with penetrating keratoplasty eventually, while case 2 recovered after sensitive treatment. Conclusions: Infectious keratitis after refractive surgery is uncommon, and Mycobacterium abscess has never been observed in corneal infections. An accurate diagnosis and identification of the causative agent is very important.

We establish that stock liquidity is conducive to less corporate diversification. Two potential channels are identified: the financial constraint channel and the corporate governance channel. Specifically, the negative effect of liquidity on diversification is stronger among financially-constrained firms, since higher liquidity helps firms improve external capital markets and thus reduces the need to broaden the internal capital markets through diversification. Moreover, the effect of liquidity on diversification is strengthened among firms with severe information asymmetry, since enhanced price informativeness caused by increased liquidity promotes market monitoring and deters managers’ opportunistic diversification decisions. Meanwhile, we rule out the alternative explanation that liquidity discourages diversification by facilitating blockholders’ monitoring on managers’ decisions. Our results suggest that stock liquidity plays a positive role in corporate decision making.

In theory, discretionary pricing improves the company's performance by allowing managers to incorporate local information into their pricing decisions. However, in practice, managers are vulnerable to behavioral biases that can lead to inaccurate estimation of the customer's willingness to pay and price sensitivity, and precipitate poor decisions. This paper investigates the effectiveness of discretionary pricing and examines how information advantage and behavioral bias influence managers' pricing decisions. Collaborating with a pharmacy retailer, we analyze a quasi-field experiment that delegates pricing power to local pharmacy managers. Our analysis reveals that under discretionary pricing, managers incorporate local information into their pricing decisions. However, they are also subject to behavioral biases, which leads to aggressive, sub-optimal price adjustments. We find that managers increase prices of high-priced drugs by 8.6\%, resulting in significant sales and revenue losses. The effect is particularly prominent in stores located in low-competition areas. Moreover, we find that manager experience can alleviate such behavioral biases---more experienced managers from larger stores are not subject to these biases. Our findings highlight the importance of recognizing managers' behavioral biases in business decisions. We provide implications for pricing practices and offer solutions that allow companies to leverage local information while reducing the impact of human biases.

Accurate segmentation of cardiac structures can assist doctors to diagnose diseases, and to improve treatment planning, which is highly demanded in the clinical practice. However, the shortage of annotation and the variance of the data among different vendors and medical centers restrict the performance of advanced deep learning methods. In this work, we present a fully automatic method to segment cardiac structures including the left (LV) and right ventricle (RV) blood pools, as well as for the left ventricular myocardium (MYO) in MRI volumes. Specifically, we design a semi-supervised learning method to leverage unlabelled MRI sequence timeframes by label propagation. Then we exploit style transfer to reduce the variance among different centers and vendors for more robust cardiac image segmentation. We evaluate our method in the M&Ms challenge (https://www.ub.edu/mnms/), ranking 2nd place among 14 competitive teams.

Image classification is an important issue in pattern recognition, the high dimension features is a challenging task since only a few number of them are effective in classification. To improve the classification efficiency, it is necessary to reduce the dimensionality of image features before classification. This study provides a novel image classification application based on Bacterial Colony Optimization, which can decreases the computation burden and improves the classification's efficiency. Specifically, the elimination strategy in original algorithm is removed, and the communication, chemotaxis, migration, and reproduction strategies are kept. Additionally, the communication and chemotaxis step size of the Bacterial Colony Optimization are modified for feature selection in image classification. Several comparision experiments on two public image datasets are conducted to verify the effectiveness of the method. Experimental results prove that the method can greatly improve the classification accuracy and efficiency.

Accurate segmentation of brain tumors from magnetic resonance imaging (MRI) is clinically relevant in diagnoses, prognoses and surgery treatment, which requires multiple modalities to provide complementary morphological and physiopathologic information. However, missing modality commonly occurs due to image corruption, artifacts, different acquisition protocols or allergies to certain contrast agents in clinical practice. Though existing efforts demonstrate the possibility of a unified model for all missing situations, most of them perform poorly when more than one modality is missing. In this paper, we propose a novel Adversarial Co-training Network (ACN) to solve this issue, in which a series of independent yet related models are trained dedicated to each missing situation with significantly better results. Specifically, ACN adopts a novel co-training network, which enables a coupled learning process for both full modality and missing modality to supplement each other’s domain and feature representations, and more importantly, to recover the ‘missing’ information of absent modalities. Then, two unsupervised modules, i.e., entropy and knowledge adversarial learning modules are proposed to minimize the domain gap while enhancing prediction reliability and encouraging the alignment of latent representations, respectively. We also adapt modality-mutual information knowledge transfer learning to ACN to retain the rich mutual information among modalities. Extensive experiments on BraTS2018 dataset show that our proposed method significantly outperforms all state-of-the-art methods under any missing situation.

Manually segmenting the hepatic vessels from Computer Tomography (CT) is far more expertise-demanding and laborious than other structures due to the low-contrast and complex morphology of vessels, resulting in the extreme lack of high-quality labeled data. Without sufficient high-quality annotations, the usual data-driven learning-based approaches struggle with deficient training. On the other hand, directly introducing additional data with low-quality annotations may confuse the network, leading to undesirable performance degradation. To address this issue, we propose a novel mean-teacher-assisted confident learning framework to robustly exploit the noisy labeled data for the challenging hepatic vessel segmentation task. Specifically, with the adapted confident learning assisted by a third party, i.e., the weight-averaged teacher model, the noisy labels in the additional low-quality dataset can be transformed from ‘encumbrance’ to ‘treasure’ via progressive pixel-wise soft-correction, thus providing productive guidance. Extensive experiments using two public datasets demonstrate the superiority of the proposed framework as well as the effectiveness of each component.

Automatic brain tumor segmentation from multi-modality Magnetic Resonance Images (MRI) using deep learning methods plays an important role in assisting the diagnosis and treatment of brain tumor. However, previous methods mostly ignore the latent relationship among different modalities. In this work, we propose a novel end-to-end Modality-Pairing learning method for brain tumor segmentation. Paralleled branches are designed to exploit different modality features and a series of layer connections are utilized to capture complex relationships and abundant information among modalities. We also use a consistency loss to minimize the prediction variance between two branches. Besides, learning rate warmup strategy is adopted to solve the problem of the training instability and early over-fitting. Lastly, we use average ensemble of multiple models and some post-processing techniques to get final results. Our method is tested on the BraTS 2020 online testing dataset, obtaining promising segmentation performance, with average dice scores of 0.891, 0.842, 0.816 for the whole tumor, tumor core and enhancing tumor, respectively. We won the second place of the BraTS 2020 Challenge for the tumor segmentation task.

BackgroundEndometriosis is a chronic hormonal inflammatory disease characterized by the presence of endometrial tissue (glands and stroma) outside the uterus. Endometriosis seriously affects the physical health of women of childbearing age, often causes infertility, and affects the body and mind of patients and their families.MethodsWe examined the effect of HSF1 on endometriosis through cell count, scratch and clone formation experiments. We used real-time quantitative PCR and western blotting to detect the effect of HSF1 on mRNA and protein of endometriosis cells. Collect the cell culture medium and Glucose levels and lactate levels were determined using a glucose (GO) assay kit and a lactate assay kit. Furthermore, we established a mouse model of endometriosis, and the effect of HSF1 on endometriosis was observed by inhibiting HSF1 with KRIBB11 in the mice.ResultsHSF1 is highly expressed in endometriosis and plays an indispensable role in endometriosis development in both cell and animal studies. We found that HSF1 promotes endometriosis development and glucose consumption and lactic acid production. Further research showed that HSF1 functions in endometriosis by up-regulating PFKFB3, a key enzyme in glycolysis. And the HSF1 inhibitor KRIBB11 can abrogate all of the above experimental effects both in vivo and in vitro. ConclusionsOur study shows that HSF1 plays a significant role in the occurrence and development of endometriosis, which may become a new target for the treatment of endometriosis and provide a new idea for the clinical treatment of endometriosis.

Purpose: Accurate segmentation of lung and infection in COVID-19 CT scans plays an important role in the quantitative management of patients. Most of the existing studies are based on large and private annotated datasets that are impractical to obtain from a single institution, especially when radiologists are busy fighting the coronavirus disease. Furthermore, it is hard to compare current COVID-19 CT segmentation methods as they are developed on different datasets, trained in different settings, and evaluated with different metrics. Methods: To promote the development of data-efficient deep learning methods, in this paper, we built three benchmarks for lung and infection segmentation based on 70 annotated COVID-19 cases, which contain current active research areas, e.g., few-shot learning, domain generalization, and knowledge transfer. For a fair comparison among different segmentation methods, we also provide standard training, validation and testing splits, evaluation metrics and, the corresponding code. Results: Based on the state-of-the-art network, we provide more than 40 pre-trained baseline models, which not only serve as out-of-the-box segmentation tools but also save computational time for researchers who are interested in COVID-19 lung and infection segmentation. We achieve average Dice Similarity Coefficient (DSC) scores of 97.3\%, 97.7\%, and 67.3\% and average Normalized Surface Dice (NSD) scores of 90.6\%, 91.4\%, and 70.0\% for left lung, right lung, and infection, respectively. Conclusions: To the best of our knowledge, this work presents the first data-efficient learning benchmark for medical image segmentation and the largest number of pre-trained models up to now. All these resources are publicly available, and our work lays the foundation for promoting the development of deep learning methods for efficient COVID-19 CT segmentation with limited data., accepted for publication in Medical Physics

This paper proposes a novel design of an underactuated finger that integrating the advantages of two most frequently used grasping method: parallel-pinching method and self-adaptive method. The finger has an underactuated mechanism using gears, tension spring and flexible belt, GFPA finger. The designed gear transmission and flexible belt-tension spring in the fingers could let the finger grasping object in a merged movement of parallel pinching and self-adaptive, which not only has the advantages of parallel-pinching to grasp the shape of the plate, but also has the advantages of adaptive way to adapt to the shape of the object. Due to the special structure including gear mechanism and tension spring-flexible belt mechanism, the finger could firstly rotate the proximal phalanx while keep the distal phalanx in a parallel motion through the meshed series gear. The distal phalanx will rotate to envelope the object only after the proximal phalanx contact with the object. The GFPA finger has important application prospects in the field of humanoid, aeronautics and Astronautics, etc.

Background and Objective Brain tumors are among the most deadly cancers worldwide. Due to the development of deep convolutional neural networks, many brain tumor segmentation methods help clinicians diagnose and operate. However, most of these methods insufficiently use multi-scale features, reducing their ability to extract brain tumors' features and details. To assist clinicians in the accurate automatic segmentation of brain tumors, we built a new deep learning network to make full use of multi-scale features for improving the performance of brain tumor segmentation. Methods We propose a novel cross-level connected U-shaped network (CLCU-Net) to connect different scales' features for fully utilizing multi-scale features. Besides, we propose a generic attention module (Segmented Attention Module, SAM) on the connections of different scale features for selectively aggregating features, which provides a more efficient connection of different scale features. Moreover, we employ deep supervision and spatial pyramid pooling (SSP) to improve the method's performance further. Results We evaluated our method on the BRATS 2018 dataset by five indexes and achieved excellent performance with a Dice Score of 88.5%, a Precision of 91.98%, a Recall of 85.62%, a Params of 36.34M and Inference Time of 8.89ms for the whole tumor, which outperformed six state-of-the-art methods. Moreover, the performed analysis of different attention modules' heatmaps proved that the attention module proposed in this study was more suitable for segmentation tasks than the other existing popular attention modules. Conclusion Both the qualitative and quantitative experimental results indicate that our cross-level connected U-shaped network with selective feature aggregation attention module can achieve accurate brain tumor segmentation and is considered quite instrumental in clinical practice implementation.

There is a large body of literature linking anatomic and geometric characteristics of kidney tumors to perioperative and oncologic outcomes. Semantic segmentation of these tumors and their host kidneys is a promising tool for quantitatively characterizing these lesions, but its adoption is limited due to the manual effort required to produce high-quality 3D segmentations of these structures. Recently, methods based on deep learning have shown excellent results in automatic 3D segmentation, but they require large datasets for training, and there remains little consensus on which methods perform best. The 2019 Kidney and Kidney Tumor Segmentation challenge (KiTS19) was a competition held in conjunction with the 2019 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) which sought to address these issues and stimulate progress on this automatic segmentation problem. A training set of 210 cross sectional CT images with kidney tumors was publicly released with corresponding semantic segmentation masks. 106 teams from five continents used this data to develop automated systems to predict the true segmentation masks on a test set of 90 CT images for which the corresponding ground truth segmentations were kept private. These predictions were scored and ranked according to their average So rensen-Dice coefficient between the kidney and tumor across all 90 cases. The winning team achieved a Dice of 0.974 for kidney and 0.851 for tumor, approaching the inter-annotator performance on kidney (0.983) but falling short on tumor (0.923). This challenge has now entered an "open leaderboard" phase where it serves as a challenging benchmark in 3D semantic segmentation., 24 pages, 11 figures

The Red Sea, home to the second-longest coral reef system in the world, is a vital resource for the Kingdom of Saudi Arabia. The Red Sea provides 90% of the Kingdom’s potable water by desalinization, supporting tourism, shipping, aquaculture, and fishing industries, which together contribute about 10%–20% of the country’s GDP. All these activities, and those elsewhere in the Red Sea region, critically depend on oceanic and atmospheric conditions. At a time of mega-development projects along the Red Sea coast, and global warming, authorities are working on optimizing the harnessing of environmental resources, including renewable energy and rainwater harvesting. All these require high-resolution weather and climate information. Toward this end, we have undertaken a multipronged research and development activity in which we are developing an integrated data-driven regional coupled modeling system. The telescopically nested components include 5-km- to 600-m-resolution atmospheric models to address weather and climate challenges, 4-km- to 50-m-resolution ocean models with regional and coastal configurations to simulate and predict the general and mesoscale circulation, 4-km- to 100-m-resolution ecosystem models to simulate the biogeochemistry, and 1-km- to 50-m-resolution wave models. In addition, a complementary probabilistic transport modeling system predicts dispersion of contaminant plumes, oil spill, and marine ecosystem connectivity. Advanced ensemble data assimilation capabilities have also been implemented for accurate forecasting. Resulting achievements include significant advancement in our understanding of the regional circulation and its connection to the global climate, development, and validation of long-term Red Sea regional atmospheric–oceanic–wave reanalyses and forecasting capacities. These products are being extensively used by academia, government, and industry in various weather and marine studies and operations, environmental policies, renewable energy applications, impact assessment, flood forecasting, and more.

Fiber Bragg Grating (FBG) sensors need to be mounted at the outer surface of any metallic test piece to monitor its structural integrity through measuring physical quantities, such as strain, pressure, vibration, and temperature. High-temperature epoxies or ceramic epoxies are used to mount the said sensors to operate under high-temperature environment conditions. The unsteadiness of the outer surface of the metallic piece and non-uniformity of the applied epoxy over the sensing head, affect its operation and measurement accuracy in the long run. Here, in the first part of a twopart article, we have investigated the reliable functioning and spectrum evolution of the FBG sensors, mounted on a carbonsteel tube, for temperature up to 500° C. Polyimide coated femtosecond laser written FBG sensors are surface mounted on the said tube using a high-temperature adhesive. There is no peak splitting, and chirp phenomenon experienced by the sensors at the end of two thermal cycles. The second part of the two-part article focuses on the mounting and operation of gold-coated FBG sensors in the high-temperature operating environment.

Coronavirus disease 2019 (COVID-19) is a highly contagious virus spreading all around the world. Deep learning has been adopted as an effective technique to aid COVID-19 detection and segmentation from computed tomography (CT) images. The major challenge lies in the inadequate public COVID-19 datasets. Recently, transfer learning has become a widely used technique that leverages the knowledge gained while solving one problem and applying it to a different but related problem. However, it remains unclear whether various non-COVID19 lung lesions could contribute to segmenting COVID-19 infection areas and how to better conduct this transfer procedure. This paper provides a way to understand the transferability of non-COVID19 lung lesions. Based on a publicly available COVID-19 CT dataset and three public non-COVID19 datasets, we evaluate four transfer learning methods using 3D U-Net as a standard encoder-decoder method. The results reveal the benefits of transferring knowledge from non-COVID19 lung lesions, and learning from multiple lung lesion datasets can extract more general features, leading to accurate and robust pre-trained models. We further show the capability of the encoder to learn feature representations of lung lesions, which improves segmentation accuracy and facilitates training convergence. In addition, our proposed Hybrid-encoder learning method incorporates transferred lung lesion features from non-COVID19 datasets effectively and achieves significant improvement. These findings promote new insights into transfer learning for COVID-19 CT image segmentation, which can also be further generalized to other medical tasks., Comment: 11 pages

To design and develop artificial intelligence (AI) hydrocephalus (HYC) imaging diagnostic model using a transfer learning algorithm and evaluate its application in the diagnosis of HYC by non-contrast material-enhanced head computed tomographic (CT) images. A training and validation dataset of non-contrast material-enhanced head CT examinations that comprised of 1000 patients with HYC and 1000 normal people with no HYC accumulating to 28,500 images. Images were pre-processed, and the feature variables were labeled. The feature variables were extracted by the neural network for transfer learning. AI algorithm performance was tested on a separate dataset containing 250 examinations of HYC and 250 of normal. Resident, attending and consultant in the department of radiology were also tested with the test sets, their results were compared with the AI model. Final model performance for HYC showed 93.6% sensitivity (95% confidence interval: 77%, 97%) and 94.4% specificity (95% confidence interval: 79%, 98%), with area under the characteristic curve of 0.93. Accuracy rate of model, resident, attending, and consultant were 94.0%, 93.4%, 95.6%, and 97.0%. AI can effectively identify the characteristics of HYC from CT images of the brain and automatically analyze the images. In the future, AI can provide auxiliary diagnosis of image results and reduce the burden on junior doctors.

Nonalcoholic fatty liver disease (NAFLD) is the hepatic manifestation of metabolic syndrome as an arising health issue globally. Obesity and diabetes seem coherent with high NAFLD occurrence. Currently, its underlying mechanism of cause and progression remains to be elicited and effective therapy is unavailable. Huge investment on studying its pathogenesis and innovative therapy is underway. As mechanistic studies are more difficult or unethical to be performed in humans, animal model that recapitulates the human disease is urgently needed. Nonhuman primates (NHPs) with NAFLD can highly mimic the human disease. This study aimed to induce NAFLD by various diets (Table 1) in normoglycemic and diabetic NHPs with biomarker and histopathology assays to quantify NAFLD occurrence. Our data show that diabetic monkeys with high-fat-high-fructose (HFHF) diet developed NAFLD which could assemble to the human disease and be highly valuable for NAFLD research. However, low choline (LC), LC+10% fructose, LC+low methionine (LCM)+10% fructose diets for 31 weeks had no significant effects on liver function, except significant decrease in body weight and food intake in LCM+10%F group and some increase in liver histological scores in LC+10%F group (Table 1). In summary, normal diet with LC or LCM ± high fructose is hardly to induce NAFLD in normoglycemic NHPs, at least in 31 weeks. HFHF could induce or enhance NAFLD in diabetic NHPs. Disclosure J. Gao: Employee; Self; Crown Bioscience. D. Wang: Employee; Self; Crown Bioscience. Y. Ji: Employee; Self; Crown Bioscience. X. Zhang: Employee; Self; Crown Bioscience. C. Yang: None. J. Zhang: Employee; Self; Crown Bioscience. W. Ye: Employee; Self; Crown Bioscience. F. Shi: Employee; Self; Crown Bioscience. M. Jiang: Employee; Self; Crown Bioscience. Q. Jin: Employee; Self; Crown Bioscience. D. Zhang: None. X. Wang: None. W. Song: Employee; Self; Crown Bioscience. K. Chng: None. H. Jiang: Employee; Self; SYSE Bio-tec Co., Ltd. Y. Wang: None. Y. Xiao: Employee; Self; Crown Bioscience.