Application of Convolutional Neural Network in COVID-19 Diagnosis

Since the outbreak and spread of COVID-19 in large areas of the world, the importance of rapid diagnosis of COVID-19 has increased. In the first week after the onset of COVID-19, the density of lesions is uneven, and chest CT is often difficult to show local subpleural ground-glass shadows, resulting in missed diagnosis. The COVID-19 intelligent diagnosis system based on the convolutional neural network algorithm can not only accurately identify the feature points, reduce the workload of doctors and improve the diagnosis efficiency, but also reduce the rate of missed diagnosis and misdiagnosis, which is conducive to epidemic control. © 2022 ACM.