COVID-CNNnet: Convolutional Neural Network for Coronavirus Detection

Abstract

The coronavirus disease (COVID-19) is the most recent severe diseases that has spread globally at an exponential rate. During this crisis, any technological approach that allows highly precise early detection of COVID-19 infection will save many lives. The main clinical technique for COVID-19 recognition is the reverse transcription polymerase chain reaction (RT-PCR). However, the RT-PCR testing tool is time-consuming, inaccurate and requires skilled medical staff. Therefore, auxiliary diagnostic tools should be developed to stop the spread of COVID-19 amongst people. Chest X-ray imaging is a readily available method that able to serve as an extremely good alternative for RT-PCR in identifying patients with COVID-19 diseases because it provides salient COVID-19 virus information. In this study, the COVID-CNNnet model proposed based on a convolutional neural network (CNN) deep learning (DL) algorithm, to detect COVID-19 cases rapidly and accurately based on patient chest X-ray images. The proposed COVID-CNNnet model aims to provide an accurate binary diagnostic classification for COVID-19 cases versus normal cases. To validate the proposed model, 3540 chest X-ray images were obtained from multiple sources, including 1770 images for COVID-19 cases. Results show that the COVID-CNNnet model can identify all classes (COVID-19 cases versus normal cases) with an accuracy of 99.86%. The proposed method can assist doctors diagnose COVID-19 cases effectively using chest X-ray images.

Ähnliche Arbeiten