Open AccessCovid-19 Detection using Deep Learning
Author(s) -
Michael S Reddy
Publication year - 2021
Publication title -
international journal for research in applied science and engineering technology
Language(s) - English
Resource type - Journals
ISSN - 2321-9653
DOI - 10.22214/ijraset.2021.35813
Subject(s) - covid-19 , computer science , support vector machine , artificial intelligence , pneumonia , random forest , histogram , gold standard (test) , histogram of oriented gradients , machine learning , pattern recognition (psychology) , statistics , disease , medicine , image (mathematics) , pathology , mathematics , infectious disease (medical specialty)
Recently, the virus (COVID-19) has spread widely throughout the world and has led to the examination of large numbers of suspected cases using standard COVID-19 tests and has become pandemic. Everyday life, public health and the global economy have been destroyed. The pathogenic laboratory tests such as Polymerase chain reaction (PCR) take a long time with false negative results and are considered the gold standard for diagnosis. Therefore, there was an urgent need for rapid and accurate diagnostic methods to detect COVID-19 cases as soon as possible to prevent the spread of this epidemic and combat it. Applying advanced artificial intelligence techniques along with radiography may be helpful in detecting this disease. In this study, we propose a classification model that detect the infected condition through the chest X-ray images. A dataset containing chest x-ray images of normal people, people with pneumonia such as SARS, streptococcus and pneumococcus and other patients with COVID- 19 were collected. Histogram of oriented gradients (HOG) is used for image features extraction. The images are then classified using Support Vector Machines (SVM), random forests and K- nearest neighbours (KNN), with classification rate 98.14%, 96.29% and 88.89% respectively. These results may contribute efficiently in detecting COVID-19 disease. The input dataset is taken from Kaggle which provides the dataset to analyse and helps to get the best possible solutions from the set of problems. Kaggle is launching a companion COVID-19 forecasting challenges to help answer a subset of the NASEM/WHO questions. While the challenge involves forecasting confirmed cases and fatalities between April 1 and April 30 by region, the primary goal isn't only to produce accurate forecasts. It’s also to identify factors that appear to impact the transmission rate of COVID-19.