Detection of Covid-19 using PDE based filter, GLSZM Features and Classifiers

Abstract

Radiologists can diagnose Covid-19 using information from a chest X-ray scan. However, X-ray visual analysis takes time. Consequently, it's essential to create algorithms for the automatic detection of COVID-19 from CT scans. To identify Covid-19 cases, we provide a logistic regression, KNN, and boosted tree classifiers in this study. The proposed method uses the fourth-order PDE-based filter as a preprocessing before extracting GLSZM features and making the classification decision. In comparison to results from the classifiers, these results are more trustworthy and explicable. According to experimental findings, the technique can produce good performance with an accuracy of 98.42%.