COVID-19 Classification Based Deep Convolutional Neural Network Using CT Scans | ||||
ERU Research Journal | ||||
Volume 3, Issue 2, April 2024, Page 1209-1222 PDF (509.13 K) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/erurj.2024.221776.1059 | ||||
View on SCiNiTO | ||||
Authors | ||||
Mahmoud Khaled 1; Amira Mofreh1; Heba Hamdy2; Asmaa Mohamed1 | ||||
1Faculty of Artificial Intelligence, Egyptian Russian University, Cairo 11829, Egypt | ||||
2Faculty of computers and Artificial Intelligence, Beni-suef university, Beni-suef, Egypt | ||||
Abstract | ||||
In the year 2020, a pandemic appeared threatening the whole world called COVID-19, the number of deaths due to this dreaded virus is constantly increasing over time. Therefore, many researchers, scientists, and professionals are seeking a solution to this problem. Diagnosis and confirmation of the presence of the virus in a person is done through CT scans. Because of the increased number of infected people, which is estimated in millions, there must be a computer system to help doctors with diagnosis to save time and effort and to help patients in the speed of treatment to preserve their lives and reduce the number of deaths. Thus, we suggested a smart computer method to automatically detect Coronavirus. A modified convolutional neural network (CNN) has been developed for automatic COVID-19 detection. The proposed technique contains three phases. In the first phase, the images are resized, data augmented, over-sampled, and normalized. In the second phase, the CNN extracts features from the images. In the classification phase, the features are used to classify the images as either COVID-19 or non-COVID. The method was evaluated on a database of 1230 non-COVID CT images and 1252 COVID CT images. The method achieved an accuracy of 93.81%, which is outperforming the other methods in terms of accuracy, sensitivity, and specificity. | ||||
Keywords | ||||
COVID-19; Deep learning; Convolutional Neural Network (CNN); CT-Scans; Image Classification | ||||
Statistics Article View: 26 PDF Download: 20 |
||||