Research Insights About Covid-19
We attempt to provide selected highlights in recent research findings
Last Update on 1 February 2021
B. Science and Engineering
January 29, 2021 (Annual Reviews in Control)
The Ockham’s razor applied to COVID-19 model fitting French data
Mirko Fiacchini & Mazen Alamir.
This paper presents a data-based model for fitting the pandemic evolution in France. The time series on the 13 regions of France has been considered for fitting and validating the model. A simple, two-dimensional model with only two parameters demonstrated to be able to reproduce the time series concerning the number of daily death, the hospitalizations, intensive care and emergency accesses, and the daily number of positive tests. These results might contribute to stimulating a debate on the suitability of much more complex models for reproducing and forecasting the pandemic evolution.
January 28, 2021 (Infection, Genetics and Evolution)
Machine learning predictive model for severe COVID-19
Jianhong Kang, Ting Chen, Honghe Luo et al.
Kang et al developed the predictive model for severe patients of COVID-19 based on the clinical date from the Tumor Center of Union Hospital affiliated with Tongji Medical College, China. They processed the results in the 5 steps. In feature selection, ALB showed a strong negative correlation (r = 0.771, P < 0.001) whereas GLB (r = 0.661, P < 0.001) and BUN (r = 0.714, P < 0.001) showed a strong positive correlation with severity of COVID-19. TensorFlow was then applied to develop a neural network model. According to them, the model achieved good prediction performance, with an area under the curve value of 0.95.
January 28, 2020 (Physica Medica)
A patient-specific approach for quantitative and automatic analysis of computed tomography images in lung disease: application to COVID-19 patients
L. Berta, C. De Mattia, F. Rizzetto et al.
The researchers proposed a patient-independent model for the estimation of the well-aerated volume of lungs in CT images. They then applied a Gaussian fit to the lower CT histogram data to provide the estimation of the well-aerated lung volume. Independence from CT reconstruction parameters and the respiratory cycle was analysed using healthy lung CT images and 4DCT acquisitions. The Gaussian metrics and first-order radiomic features calculated for a third cohort were compared with healthy lungs. Each lung was further segmented and a new biomarker derived from Gaussian fit parameter was proposed to represent the local density changes. They found that healthy subjects were significantly different from COVID-19 patients for all the metrics calculated. Graphical representation of the local biomarker provides spatial and quantitative information in a single 2D picture. They concluded that this model was able to consider the inter-and intra-subject variability and defined a local biomarker to quantify the severity of the disease.
January 14, 2021 (Chaos, Solitons & Fractals)
A SIR-type model describing the successive waves of COVID-19
Gustavo A. Munoz-Fernandez, Jesus M. Seoane, Juan B. Seoane-Sepulveda.
The authors show that the official data released by several countries (Italy, Spain and The USA) regarding the expansion of COVID-19 are compatible with a non-autonomous SIR type model with vital dynamics and non-constant population, calibrated according to exponentially decaying infection and death rates. They then construct a model whose outcomes for most relevant parameters, such as the number of active cases, cumulative deaths, daily new deaths and daily new cases fit available real data about the first and successive waves of COVID-19.
January 14, 2021 (European Journal of Radiology)
Multi-classifier-based identification of COVID-19 from chest computed tomography using generalizable and interpretable radiomics features
Lu Wang, Brendan Kelly, Edward H. Lee et al.
The researchers aimed to identify radiomic features that were significantly associated with the classification of COVID-19 pneumonia using multiple classifiers. 10 positive and 108 negative patients were retrospectively recruited from three hospitals. They used four classifiers (linear classifier, k-nearest neighbour, least absolute shrinkage and selection operator [LASSO], and random forest) to differentiate positive and negative patients. Using the Pyradiomics technique, 120 radiomic features were extracted from each image. The key radiomic features screened by different classifiers lead to significant differences in classification accuracy. The LASSO achieved the best performance on the external validation dataset and attained good agreement (Kappa score: 0.89) with radiologists (sensitivity: 75.6%, specificity: 78.2%, and AUC: 0.81).
January 12, 2021 (Journal of Infection and Public Health)
Modelization of Covid-19 Pandemic Spreading: A Machine Learning Forecasting with Relaxation Scenarios of Countermeasures
Moulay A. Lmater, Mohamed Eddabbah, Tariq Elmoussaoui et al.
The researchers performed a mathematical model according to a SIDR model for infectious diseases. Epidemiological data from Belgium, Morocco, Netherlands and Russia, are used to validate this model. They validated a new way of data aggregation, modelling and interpretation to predict the spread of Covid-19, evaluate the efficiency of countermeasures and suggest potential scenarios.
January 12, 2021 (Proceedings of the National Academy of Sciences)
Structure of SARS-CoV-2 ORF8, a rapidly evolving immune evasion protein
Thomas G. Flower, Cosmo Z. Buffalo, Richard M. Hooy et al.
This highly technical paper reports on the structure of the ORF8 which is a rapidly evolving accessory protein that has been proposed to interfere with immune responses. The structure reveals a ∼60-residue core similar to SARS-CoV-2 ORF7a, with the addition of two dimerization interfaces unique to SARS-CoV-2 ORF8. The presence of these interfaces shows how SARS-CoV-2 ORF8 can form unique large-scale assemblies not possible for SARS-CoV, potentially mediating unique immune suppression and evasion activities.
January 11, 2021 (Safety Science)
Diagnostic model for the society safety under COVID-19 pandemic conditions
Costas A. Varotsos, Vladimir F. Krapivin & Yong Xue
The authors developed an information-modeling method for assessing and predicting the consequences of the pandemic. They performed a detailed analysis of official statistical information. The method is based on the algorithm of multi-channel big data processing. COVID-19 data are analyzed using an instability indicator and a system of differential equations describing four groups of people: susceptible, infected, recovered and dead. Stochastic processes induced by COVID-19 are assessed with the instability indicator showing the level of stability of official data and the reduction of the level of uncertainty. Their findings show that COVID-19 divides the global population into three groups according to the relationship between Gross Domestic Product and the number of infected people.
January 4, 2021 (Communications Biology)
Fast automated detection of COVID-19 from medical images using convolutional neural networks
Shuang Liang, Huixiang Lui, Yu Gu et al.
The researchers use pseudo-colouring methods and a platform for annotating X-ray and computed tomography images to train the convolutional neural network, which achieves a performance similar to that of experts and provides high scores for multiple statistical indices (F1 scores > 96.72% and specificity >99.33%. Heatmaps are used to visualize the salient features extracted by the neural network. The proposed method represents a potential computer-aided diagnosis method for COVID-19 in clinical practice.