Diagnosis of COVID-19 Using Deep Learning Techniques

Please enable javascript in your browser.

Nourparvar, Azadeh | 2023

58 Viewed

Type of Document: M.Sc. Thesis
Language: English
Document No: 56176 (52)
University: Sharif University of Technology, International Campus, Kish Island
Department: Science and Engineering
Advisor(s): Hemmatyar, Ali Mohammad Afshin
Abstract:
In the last few years, the most important problem to find a solution for the COVID-19 disease, and tests show that many factors are effective in the health and recovery of people with this disease. Therefore, scientists all over the world are looking to prevent the spread of this disease by identifying the effective factors in the recovery of corona patients and finding solutions for their health. The proposed algorithm is hybrid deep learning model CNN+GRU and appeal them to the laboratory test data from hospital. In this research, the goal is to be able to diagnose this disease in time with routine laboratory test, which consist of three main stages. In the first stage: initial preprocessing and deletion of null data from the dataset are done. In the second stage: important and effective features in diagnosis are identified with the appropriate algorithm. Finally, in the third stage: the deep learning algorithm suitable for disease diagnosis is applied to the features. The result show that this method gives high accuracy and the value obtained in these four evaluation criteria of accuracy, precision, recall, and F1-score 93%
Keywords:
COVID-19 ; Deep Learning ; Laboratory Test ; Hybrid Deep Learning Model CNN+GRU ; Disease Diagnosis

No TOC