Diagnosis of Covid-19 using Fractional B-spline Wavelet Transform in Lung Ultrasound images

The coronavirus spread rapidly in the world and caused disease of Covid-19. proposed research was conducted with aim classifying people Covid-19 from other people. Among all imaging modalities, lung ultrasound images were used to diagnose covid-19. open source Point-Care-of-Ultrasound (POCUS) database collected, which contained 59 Lung Ultrasound (LUS) images. In this research, KNN classifier K-fold cross validation classify feature matrix obtained Fractional B-spline Wavelet Transform (FBSWT). method, Block-Matching 3D filtering (BM3D) filter some methods, had acceptable results. method healthy patients results show that when features based on wavelet transform (WT) are used, can be achieved 90.90% sensitivity, 92.30% specificity, 91.42% accuracy. While extracted FBSWT 95.45% 93.30% 94.28% accuracy.Fractional transforms, especially FBSWT, a useful tool for image processing. They various purposes such as detection classification. By using it is possible accurately