COVID-19 Disease Classification Model Using Deep Dense Convolutional Neural Networks

ABSTRACT

Preventing the transmission of COVID-19 necessitates diagnosis and identification. Researchers have developed algorithms to detect the presence of COVID-19 in X-ray and CT scans and images. These methodologies produce skewed data and incorrect disease detection. So, in the case of COVID-19 forecasting utilizing CT scans in an IoT setting, the current study paper established an oppositional-based deep dense convolutional neural network (DDCNN) and chimp optimization algorithm. The framework proposed is divided into two stages preprocessing and estimation. Previously, a CT scan pictures generated from anticipated COVID-19 are acquired utilizing IoT devices from an open-source system. After that, the photos are preprocessed with a Gaussian function. A Gaussian filter can be used to remove undesirable noise from CT scan pictures that have been obtained. The preprocessed photos are then transmitted to the prediction process. DDCNN is applied to the images preprocessed in this step. The recommended classifier is designed to be as efficient as possible using the oppositional-based chimp optimization algorithm (OCOA). This approach is used to choose the best classifier parameters under consideration. Furthermore, the suggested method is applied to forecast COVID-19 and categorizes the findings as COVID-19 or non-COVID-19. The proposed technique was used in Python, and results were assessed using statistical analysis. CNN-EPO and CNN-FA were compared to the new method. The results proved that the proposed model was optimal. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.