期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2021
卷号：99
期号：11
语种：English
出版社：Journal of Theoretical and Applied
摘要：In this paper, a novel method is proposed for COVID-19 detection from chest images. The proposed method uses some important features from both spatial and the Fourier transform of the input images. The binary particle swarm optimization is used to select the most relevant features. Two common classifiers are used for testing; support vector machine and k-nearest neighbor. Results show that the k-nearest neighbor outperforms support vector machine. The accuracy of the proposed method outperforms other algorithms in the literature. The accuracy of the proposed method approximately equals 91% when using the proposed features combined with the binary particle swarm optimization (BPSO). The sensitivity exceeds 89%, and also outperforms that proposed in previous work. Specificity is also maintained. These important findings may represent physicians' importance in decreasing diagnosis time and cost using automated systems. These systems may be useful for physicians in case of resource limitation.
