A COVID-19 Detection Model Based on Convolutional Neural Network and Residual Learning

A model that can obtain rapid and accurate detection of coronavirus disease 2019 (COVID-19) plays a significant role in treating and preventing the spread of disease transmission. However, designing such a model that can balance the detection accuracy and weight parameters of memory well to deploy a mobile device is challenging. Taking this point into account, this paper fuses the convolutional neural network and residual learning operations to build a multi-class classification model, which improves COVID-19 pneumonia detection performance and keeps a trade-off between the weight parameters and accuracy. The convolutional neural network can extract the COVID-19 feature information by repeated convolutional operations. The residual learning operations alleviate the gradient problems caused by stacking convolutional layers and enhance the ability of feature extraction. The ability further enables the proposed model to acquire effective feature information at a low cost, which can make our model keep small weight parameters. Extensive validation and comparison with other models of COVID-19 pneumonia detection on the well-known COVIDx dataset show that (1) the sensitivity of COVID-19 pneumonia detection is improved from 88.2% (non-COVID-19) and 77.5% (COVID-19) to 95.3% (non-COVID-19) and 96.5% (COVID-19), respectively. The positive predictive value is also respectively increased from 72.8% (non-COVID-19) and 89.0% (COVID-19) to 88.8% (non-COVID-19) and 95.1% (COVID-19). (2) Compared with the weight parameters of the COVIDNet-small network, the value of the proposed model is 13 M, which is slightly higher than that (11.37 M) of the COVIDNet-small network. But, the corresponding accuracy is improved from 85.2% to 93.0%. The above results illustrate the proposed model can gain an efficient balance between accuracy and weight parameters. © 2023 Tech Science Press. All rights reserved.

A Deep Learning Network for Detecting COVID-19 From Chest X-Ray Images

Rapid and accurate detection of COVID-19 plays a significant role in treating and preventing the spread of disease transmission. To this end, we fuse the convolutional neural network and residual learning operation to build a multi-class classification model, which has a few parameters and is more conducive to be deployed on a mobile device. Extensive experiments show that our proposed model gains competitive performance. Compared with the COVIDNet-small network, the sensitivity of COVID-19 pneumonia detection is improved from 88.2% (non-COVID-19) and 77.5% (COVID-19) to 95.3% (non-COVID-19) and 96.5% (COVID-19). Alternatively, the Positive predictive value is increased from 72.8% (non-COVID-19) and 89.0% (COVID-19) to 88.8% (non-COVID-19) and 95.1 % (COVID-19). The accuracy is also improved from 85.2 % to 93.0 %, which is very close to the value (93.3 %) of the COVIDNet-large network. But, the weight parameters (13M) of the proposed model are slightly higher than that (11.37M) of the COVIDNet-small network, but only about one-third of that (37.85M) of the COVIDNet-large network. © 2022 IEEE.

The City Agent Model of COVID-19 based on GIS and Application: A Case Study of Guangzhou

Since December 2019, a new type of coronavirus pneumonia has occurred in Wuhan, Hubei. The strong spread ability of the new coronavirus has led to the rapidly emergence of new coronaviruses throughout the country and even all over the world. In order to portray the spread line of the new coronavirus within the city and then provide reasonable suggestions for the prevention and control of the urban epidemic, this article constructs a new coronavirus intelligent simulation model by combining complex network theory and GIS technology based on the behavior and social relationships of individuals in the city. Considering to the facts that it is necessary to strictly prevent the import of overseas cases to prevent the local epidemic from rebounding in cities with complex composition of population. This agent model takes the first entry point for overseas entry, Guangzhou city, as the research object to review the development of the epidemic. The attributes and rules of the model was determined by collecting statistical data from the literatures. Then the parameters were fitted by the Markov chain Monte Carlo method to achieve an accurate review of the epidemic situation in Guangzhou. The model is of high accuracy whose MAPE value have achieved 0.17. Meanwhile, this model also has good applicability which can simulate the impact of imported cases from abroad on the development of urban epidemics. Since the agent model marks the individual's time and space location and social relationship, this paper proposes a method for epidemiological investigation through the agent model, which is more convenient and more efficient than traditional epidemiological investigations.This article also visually displays the results of the infection chain, which is convenient for analyzing the activity trajectory of virus carriers and close contacts. This model provides valuable decision-making information for urban epidemic prevention and control. Moreover, the simulation results show that if there is another epidemic outbreak in the city, the epidemic will be controlled within 14-20 days so the citizens don't need to be panic. However, it is still necessary to improve self-protection awareness and protect individuals finely, especially the children and the elderly. When the epidemic comes again, it is recommended that schools and enterprises should establish a joint health monitoring mechanism to strengthen the health monitoring of children and employees, respectively. Relevant governmental departments have to strengthened the spread of epidemic prevention knowledge and persuaded retired people to reduce gatherings and wear masks reasonably. 2021, Science Press. All right reserved.