Convolutional Neural Network Model to Detect COVID-19 Patients Utilizing Chest X-ray Images

Abstract

This study aims to investigate if applying machine learning and deep learning approaches on chest X-ray images can detect cases of coronavirus. The chest X-ray datasets were obtained from Kaggle and Github and pre-processed into a single dataset using random sampling. We applied several machine learning and deep learning methods including Convolutional Neural Networks (CNN) along with classical machine learners. In deep learning procedure, several pre-trained models were also employed transfer learning in this dataset. Our proposed CNN model showed the highest accuracy (94.03%), AUC (95.52%), f-measure (94.03%), sensitivity (94.03%) and specificity (97.01%) as well as the lowest fall out (4.48%) and miss rate (2.98%) respectively. We also evaluated specificity and fall out rate along with accuracy to identify non-COVID-19 individuals more accurately. As a result, our new models might help to early detect COVID-19 patients and prevent community transmission compared to traditional methods.

A deep learning approach using effective preprocessing techniques to detect COVID-19 from chest CT-scan and X-ray images

“…It may also be possible to detect early mild symptomatic cases that give false negative RT-PCR results due to low viral particle numbers in the upper respiratory tract. Two kinds of chest radiographic images have been used for this approach: X-ray & computed tomography (CT) [ 15 , 16 ]. CT scans provide very fine detail but require substantially more radiation exposure than X-ray images and require high cost equipment [ 17 ].…”

An intelligent deep convolutional network based COVID-19 detection from chest X-rays

“…Researchers intended to use exclusively genuine X-ray, ultrasonography, and computed tomography information in this project, and they will not explore creating and using simulated data at this time. For the model experiments, researchers also utilized a rather equal database of images, with imbalances resolved through the use of computed class training values [23] . Researchers constructed a master database for all tests using the source information listed in Table 2 .…”

Transfer Learning for the Detection and Diagnosis of Types of Pneumonia including Pneumonia Induced by COVID-19 from Chest X-ray Images

“…These methods are varied in their use of different pipelines and ML techniques from feature preprocessing to the choice of architecture under different contexts and considerations, thus yielding different performance results. Ahammed et al [30] in their comparative survey of ML and deep learning approaches for the detection of COVID-19 using a dataset of all publicly available chest X-ray images of COVID-19 patients, reported 94.03% accuracy, 95.52% AUC, and 94.03% sensitivity. Ng et al [21] created a massive dataset of 13,1975 XCR images and used a deep neural network model to classify the images which resulted to an accuracy of 93.30%.…”