Spatio-temporal Graph Learning for Epidemic Prediction

The COVID-19 pandemic has posed great challenges to public health services, government agencies, and policymakers, raising huge social conflicts between public health and economic resilience. Policies such as reopening or closure of business activities are formulated based on scientific projections of infection risks obtained from infection dynamics models. Though most parameters in epidemic prediction service models can be set with domain knowledge of COVID-19, a key parameter, namely, human mobility, is often challenging to estimate due to complex spatio-temporal correlations and social contexts under escalating COVID-19 facilities. Moreover, how to integrate the various implicit features to accurately predict infectious cases is still an open issue. To address this challenge, we formulate the problem as a spatio-temporal network representation problem and propose STEP, a Spatio-Temporal Epidemic Prediction framework, to estimate pandemic infection risk of a city by integrating various real-world conditions (e.g., City Risk Index, climate, and medical conditions) into graph-structured data. We also employ a multi-head attention mechanism in representation learning to extract implicit features for a given city. Extensive experiments have been conducted upon the real-world dataset for 51 states (50 states and Washington, D.C.) of the USA. Experimental results show that STEP can yield more accurate pandemic infection risk estimation than baseline methods. Moreover, STEP outperforms other methods in both short-term and long-term prediction. [ABSTRACT FROM AUTHOR] Copyright of ACM Transactions on Intelligent Systems & Technology is the property of Association for Computing Machinery and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

CupMar: A deep learning model for personalized news recommendation based on contextual user-profile and multi-aspect article representation

In modern days, making recommendation for news articles poses a great challenge due to vast amount of online information. However, providing personalized recommendations from news articles, which are the sources of condense textual information is not a trivial task. A recommendation system needs to understand both the textual information of a news article, and the user contexts in terms of long-term and temporary preferences via the user's historic records. Unfortunately, many existing methods do not possess the capability to meet such need. In this work, we propose a neural deep news recommendation model called CupMar, that not only is able to learn the user-profile representation in different contexts, but also is able to leverage the multi-aspects properties of a news article to provide accurate, personalized news recommendations to users. The main components of our CupMar approach include the News Encoder and the User-Profile Encoder. Specifically, the News Encoder uses multiple properties such as news category, knowledge entity, title and body content with advanced neural network layers to derive informative news representation, while the User-Profile Encoder looks through a user's browsed news, infers both of her long-term and recent preference contexts to encode a user representation, and finds the most relevant candidate news for her. We evaluate our CupMar model with extensive experiments on the popular Microsoft News Dataset (MIND), and demonstrate the strong performance of our approach.

Beyond CNNs: Exploiting Further Inherent Symmetries in Medical Image Segmentation.

Automatic tumor or lesion segmentation is a crucial step in medical image analysis for computer-aided diagnosis. Although the existing methods based on convolutional neural networks (CNNs) have achieved the state-of-the-art performance, many challenges still remain in medical tumor segmentation. This is because, although the human visual system can detect symmetries in 2-D images effectively, regular CNNs can only exploit translation invariance, overlooking further inherent symmetries existing in medical images, such as rotations and reflections. To solve this problem, we propose a novel group equivariant segmentation framework by encoding those inherent symmetries for learning more precise representations. First, kernel-based equivariant operations are devised on each orientation, which allows it to effectively address the gaps of learning symmetries in existing approaches. Then, to keep segmentation networks globally equivariant, we design distinctive group layers with layer-wise symmetry constraints. Finally, based on our novel framework, extensive experiments conducted on real-world clinical data demonstrate that a group equivariant Res-UNet (called GER-UNet) outperforms its regular CNN-based counterpart and the state-of-the-art segmentation methods in the tasks of hepatic tumor segmentation, COVID-19 lung infection segmentation, and retinal vessel detection. More importantly, the newly built GER-UNet also shows potential in reducing the sample complexity and the redundancy of filters, upgrading current segmentation CNNs, and delineating organs on other medical imaging modalities.

Integrated Prediction Framework for Clinical Scores of Cognitive Functions in ESRD Patients.

The clinical scores are applied to determine the stage of cognitive function in patients with end-stage renal disease (ESRD). However, accurate clinical scores are hard to come by. This paper proposed an integrated prediction framework with GPLWLSV to predict clinical scores of cognitive functions in ESRD patients. GPLWLSV incorporated three parts, graph theoretic algorithm (GTA) and principal component analysis (PCA), whale optimization algorithm with Levy flight (LWOA), and least squares support vector regression machine (LSSVRM). GTA was adopted to extract features from the brain functional networks in ESRD patients, while PCA was used to select features. LSSVRM was built to explore the relationship between the selected features and the clinical scores of ESRD patients. Whale optimization algorithm (WOA) was introduced to select better parameters of the kernel function in LSSVRM; it aims to improve the exploration competence of LSSVRM. Levy flight was used to optimize the ability to jump out of local optima in WOA and improve the convergence of coefficient vectors in WOA, which lead to an increase in the generalization ability and convergence speed of WOA. The results validated that the prediction accuracy of GPLWLSV was higher than that of several comparable frameworks, such as GPSV, GPLSV, and GPWLSV. In particular, the average of root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) between the predicted scores and the actual scores of ESRD patients was 2.40, 2.06, and 9.83%, respectively. The proposed framework not only can predict the clinical scores more accurately but also can capture imaging markers associated with decline of cognitive function. It helps to understand the potential relationship between structural changes in the brain and cognitive function of ESRD patients.

Pulmonary and Renal Long COVID at Two-year Revisit (preprint)

More than 450 million individuals have recovered from COVID-19, but little is known about the host responses to long COVID. We performed proteomic and metabolomic analyses of 991 blood and urine specimens from 144 COVID-19 patients with comprehensive clinical data and up to 763 days of follow up. Our data showed that the lungs and kidneys are the most vulnerable organs in long COVID patients. Pulmonary and renal long COVID of one-year revisit can be predicted by a machine learning model based on clinical and multi-omics data collected during the first month from the disease onset with an ACC of 87.5%. Serum protein SFTPB and ATR were associated with pulmonary long COVID and might be potential therapeutic targets. Notably, our data show that all the patients with persistent pulmonary ground glass opacity or patchy opacity lesions developed into pulmonary fibrosis at two-year revisit. Together, this study depicts the longitudinal clinical and molecular landscape of COVID-19 with up to two-year follow-up and presents a method to predict pulmonary and renal long COVID.

Machine learning research towards combating COVID-19: Virus detection, spread prevention, and medical assistance.

COVID-19 was first discovered in December 2019 and has continued to rapidly spread across countries worldwide infecting thousands and millions of people. The virus is deadly, and people who are suffering from prior illnesses or are older than the age of 60 are at a higher risk of mortality. Medicine and Healthcare industries have surged towards finding a cure, and different policies have been amended to mitigate the spread of the virus. While Machine Learning (ML) methods have been widely used in other domains, there is now a high demand for ML-aided diagnosis systems for screening, tracking, predicting the spread of COVID-19 and finding a cure against it. In this paper, we present a journey of what role ML has played so far in combating the virus, mainly looking at it from a screening, forecasting, and vaccine perspective. We present a comprehensive survey of the ML algorithms and models that can be used on this expedition and aid with battling the virus.

No Statistically Apparent Difference in Antiviral Effectiveness Observed Among Ribavirin Plus Interferon-Alpha, Lopinavir/Ritonavir Plus Interferon-Alpha, and Ribavirin Plus Lopinavir/Ritonavir Plus Interferon-Alpha in Patients With Mild to Moderate Coronavirus Disease 2019: Results of a Randomized, Open-Labeled Prospective Study.

BACKGROUND: Currently, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has spread globally, causing an unprecedented pandemic. However, there is no specific antiviral therapy for coronavirus disease 2019 (COVID-19). We conducted a clinical trial to compare the effectiveness of three antiviral treatment regimens in patients with mild to moderate COVID-19. METHODS: This was a single-center, randomized, open-labeled, prospective clinical trial. Eligible patients with mild to moderate COVID-19 were randomized into three groups: ribavirin (RBV) plus interferon-α (IFN-α), lopinavir/ritonavir (LPV/r) plus IFN-α, and RBV plus LPV/r plus IFN-α at a 1:1:1 ratio. Each patient was invited to participate in a 28-d follow-up after initiation of an antiviral regimen. The outcomes include the difference in median interval to SARS-CoV-2 nucleic acid negativity, the proportion of patients with SARS-CoV-2 nucleic acid negativity at day 14, the mortality at day 28, the proportion of patients re-classified as severe cases, and adverse events during the study period. RESULTS: In total, we enrolled 101 patients in this study. Baseline clinical and laboratory characteristics of patients were comparable among the three groups. In the analysis of intention-to-treat data, the median interval from baseline to SARS-CoV-2 nucleic acid negativity was 12 d in the LPV/r+IFN-α-treated group, as compared with 13 and 15 d in the RBV+IFN-α-treated group and in the RBV+LPV/r+ IFN-α-treated group, respectively (p=0.23). The proportion of patients with SARS-CoV-2 nucleic acid negativity in the LPV/r+IFN-α-treated group (61.1%) was higher than the RBV+ IFN-α-treated group (51.5%) and the RBV+LPV/r+IFN-α-treated group (46.9%) at day 14; however, the difference between these groups was calculated to be statistically insignificant. The RBV+LPV/r+IFN-α-treated group developed a significantly higher incidence of gastrointestinal adverse events than the LPV/r+ IFN-α-treated group and the RBV+ IFN-α-treated group. CONCLUSIONS: Our results indicate that there are no significant differences among the three regimens in terms of antiviral effectiveness in patients with mild to moderate COVID-19. Furthermore, the combination of RBV and LPV/r is associated with a significant increase in gastrointestinal adverse events, suggesting that RBV and LPV/r should not be co-administered to COVID-19 patients simultaneously. CLINICAL TRIAL REGISTRATION: www.ClinicalTrials.gov, ID: ChiCTR2000029387. Registered on January 28, 2019.

Proteomic and Metabolomic Characterization of COVID-19 Patient Sera (preprint)

Severe COVID-19 patients account for most of the mortality of this disease. Early detection and effective treatment of severe patients remain major challenges. Here, we performed proteomic and metabolomic profiling of sera from 46 COVID-19 and 53 control individuals. We then trained a machine learning model using proteomic and metabolomic measurements from a training cohort of 18 non-severe and 13 severe patients. The model correctly classified severe patients with an accuracy of 93.5%, and was further validated using ten independent patients, seven of which were correctly classified. We identified molecular changes in the sera of COVID-19 patients implicating dysregulation of macrophage, platelet degranulation and complement system pathways, and massive metabolic suppression. This study shows that it is possible to predict progression to severe COVID-19 disease using serum protein and metabolite biomarkers. Our data also uncovered molecular pathophysiology of COVID-19 with potential for developing anti-viral therapies.Funding: This work is supported by grants from Westlake Special Program for COVID19 (2020), and Tencent foundation (2020), National Natural Science Foundation of China (81972492, 21904107, 81672086), Zhejiang Provincial Natural Science Foundation for Distinguished Young Scholars (LR19C050001), Hangzhou Agriculture and Society Advancement Program (20190101A04). Conflict of Interest: The research group of T.G. is partly supported by Tencent, Thermo Fisher Scientific, SCIEX and Pressure Biosciences Inc. C.Z., Z.K., Z.K. and S.Q. are employees of DIAN Diagnostics.

Proteomic and Metabolomic Characterization of COVID-19 Patient Sera (preprint)

Severe COVID-19 patients account for most of the mortality of this disease. Early detection and effective treatment of severe patients remain major challenges. Here, we performed proteomic and metabolomic profiling of sera from 46 COVID-19 and 53 control individuals. We then trained a machine learning model using proteomic and metabolomic measurements from a training cohort of 18 non-severe and 13 severe patients. The model correctly classified severe patients with an accuracy of 93.5%, and was further validated using ten independent patients, seven of which were correctly classified. We identified molecular changes in the sera of COVID-19 patients implicating dysregulation of macrophage, platelet degranulation and complement system pathways, and massive metabolic suppression. This study shows that it is possible to predict progression to severe COVID-19 disease using serum protein and metabolite biomarkers. Our data also uncovered molecular pathophysiology of COVID-19 with potential for developing anti-viral therapies.