A combined deep CNN-LSTM network for the detection of novel coronavirus (COVID-19) using X-ray images

• DOI: 10.1016/j.imu.2020.100412
• Publication Date: 2020-08-15T14:50:20Z
• Authors: Md. Zabirul Islam; Md. Milon Islam; Amanullah Asraf
• Keywords: Feature (linguistics)

Nowadays, automatic disease detection has become a crucial issue in medical science due to rapid population growth. An framework assists doctors the diagnosis of and provides exact, consistent, fast results reduces death rate. Coronavirus (COVID-19) one most severe acute diseases recent times spread globally. Therefore, an automated system, as fastest diagnostic option, should be implemented impede COVID-19 from spreading. This paper aims introduce deep learning technique based on combination convolutional neural network (CNN) long short-term memory (LSTM) diagnose automatically X-ray images. In this CNN is used for feature extraction LSTM using extracted feature. A collection 4575 images, including 1525 images COVID-19, were dataset system. The experimental show that our proposed system achieved accuracy 99.4%, AUC 99.9%, specificity 99.2%, sensitivity 99.3%, F1-score 98.9%. desired currently available dataset, which can further improved when more available. help treat patients easily.