Diagnosis of Covid-19 Disease in Lung Images Using Siamese Deep Neural Network and Meta-Heuristic Algorithm for Feature Selection
27 Pages Posted: 28 Sep 2022
Abstract
COVID-19 disease, which is a member of the coronavirus family, has become a global pandemic and has had many adverse individual and social effects. The most basic method of controlling this disease is early diagnosis to reduce the mortality rate. The results of various studies on the diagnosis of COVID-19 show that deep learning can be successful in diagnosing this disease. In this regard, due to the importance of early diagnosis of this disease, in this article, a physician assistant system for the diagnosis of COVID-19 based on the Siamese neural network and the meta-heuristic algorithm of the Great Pyramid of Giza is presented. The deep Siamese neural network consists of two subunits of the same convolution and CT images of patients' lungs have been used to extract deep features. Also, to achieve the optimal subset of features, the meta-heuristic algorithm of the Great Pyramid of Giza based on archeology was used and then to diagnose COVID-19 disease, a combination of three K-Nearest Neighbor algorithms including simple, weighted and fuzzy version with majority voting technique was used. . The evaluation results of this method in the CT-COV database of patient’s lungs show that this method has been able to achieve 98.64% accuracy in diagnosing the disease, which indicates its success.
Note:
Funding Information: None to declare.
Declaration of Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Keywords: COVID-19 disease, Siamese Neural Networks, Giza Pyramid Algorithm, Fuzzy Classification
Suggested Citation: Suggested Citation