COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings

The key component in deep learning research is the availability of training data sets. With a limited number of publicly available COVID-19 chest X-ray images, the generalization and robustness of deep learning models to detect COVID-19 cases developed based on these images are questionable. We aime...

Full description

Bibliographic Details
Main Authors: Che Azemin, Mohd Zulfaezal, Hassan, Radhiana, Mohd Tamrin, Mohd Izzuddin, Md. Ali, Mohd. Adli
Format: Article
Language:English
English
English
Published: 2020
Subjects:
Online Access:http://irep.iium.edu.my/82304/
http://irep.iium.edu.my/82304/1/8828855.pdf
http://irep.iium.edu.my/82304/7/82304_COVID-19%20Deep%20Learning%20Prediction%20Model%20Using%20Publicly_Scopus.pdf
http://irep.iium.edu.my/82304/12/82272_Synthesis%20and%20characterization%20of%20bubble_wos.pdf
_version_ 1848789273931677696
author Che Azemin, Mohd Zulfaezal
Hassan, Radhiana
Mohd Tamrin, Mohd Izzuddin
Md. Ali, Mohd. Adli
author_facet Che Azemin, Mohd Zulfaezal
Hassan, Radhiana
Mohd Tamrin, Mohd Izzuddin
Md. Ali, Mohd. Adli
author_sort Che Azemin, Mohd Zulfaezal
building IIUM Repository
collection Online Access
description The key component in deep learning research is the availability of training data sets. With a limited number of publicly available COVID-19 chest X-ray images, the generalization and robustness of deep learning models to detect COVID-19 cases developed based on these images are questionable. We aimed to use thousands of readily available chest radiograph images with clinical findings associated with COVID-19 as a training data set, mutually exclusive from the images with confirmed COVID-19 cases, which will be used as the testing data set. We used a deep learning model based on the ResNet-101 convolutional neural network architecture, which was pretrained to recognize objects from a million of images and then retrained to detect abnormality in chest X-ray images. The performance of the model in terms of area under the receiver operating curve, sensitivity, specificity, and accuracy was 0.82, 77.3%, 71.8%, and 71.9%, respectively. The strength of this study lies in the use of labels that have a strong clinical association with COVID-19 cases and the use of mutually exclusive publicly available data for training, validation, and testing.
first_indexed 2025-11-14T17:54:06Z
format Article
id iium-82304
institution International Islamic University Malaysia
institution_category Local University
language English
English
English
last_indexed 2025-11-14T17:54:06Z
publishDate 2020
recordtype eprints
repository_type Digital Repository
spelling iium-823042021-02-18T08:56:31Z http://irep.iium.edu.my/82304/ COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings Che Azemin, Mohd Zulfaezal Hassan, Radhiana Mohd Tamrin, Mohd Izzuddin Md. Ali, Mohd. Adli RC731 Specialties of Internal Medicine-Diseases of The Respiratory System TK7885 Computer engineering The key component in deep learning research is the availability of training data sets. With a limited number of publicly available COVID-19 chest X-ray images, the generalization and robustness of deep learning models to detect COVID-19 cases developed based on these images are questionable. We aimed to use thousands of readily available chest radiograph images with clinical findings associated with COVID-19 as a training data set, mutually exclusive from the images with confirmed COVID-19 cases, which will be used as the testing data set. We used a deep learning model based on the ResNet-101 convolutional neural network architecture, which was pretrained to recognize objects from a million of images and then retrained to detect abnormality in chest X-ray images. The performance of the model in terms of area under the receiver operating curve, sensitivity, specificity, and accuracy was 0.82, 77.3%, 71.8%, and 71.9%, respectively. The strength of this study lies in the use of labels that have a strong clinical association with COVID-19 cases and the use of mutually exclusive publicly available data for training, validation, and testing. 2020-08-18 Article PeerReviewed application/pdf en http://irep.iium.edu.my/82304/1/8828855.pdf application/pdf en http://irep.iium.edu.my/82304/7/82304_COVID-19%20Deep%20Learning%20Prediction%20Model%20Using%20Publicly_Scopus.pdf application/pdf en http://irep.iium.edu.my/82304/12/82272_Synthesis%20and%20characterization%20of%20bubble_wos.pdf Che Azemin, Mohd Zulfaezal and Hassan, Radhiana and Mohd Tamrin, Mohd Izzuddin and Md. Ali, Mohd. Adli (2020) COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings. International Journal of Biomedical Imaging, 2020. pp. 1-7. ISSN 1687-4188 E-ISSN 1687-4196 https://www.hindawi.com/journals/ijbi/2020/8828855/ https://doi.org/10.1155/2020/8828855
spellingShingle RC731 Specialties of Internal Medicine-Diseases of The Respiratory System
TK7885 Computer engineering
Che Azemin, Mohd Zulfaezal
Hassan, Radhiana
Mohd Tamrin, Mohd Izzuddin
Md. Ali, Mohd. Adli
COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title_full COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title_fullStr COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title_full_unstemmed COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title_short COVID-19 deep learning prediction model using publicly available radiologist-adjudicated chest X-ray images as training data: preliminary findings
title_sort covid-19 deep learning prediction model using publicly available radiologist-adjudicated chest x-ray images as training data: preliminary findings
topic RC731 Specialties of Internal Medicine-Diseases of The Respiratory System
TK7885 Computer engineering
url http://irep.iium.edu.my/82304/
http://irep.iium.edu.my/82304/
http://irep.iium.edu.my/82304/
http://irep.iium.edu.my/82304/1/8828855.pdf
http://irep.iium.edu.my/82304/7/82304_COVID-19%20Deep%20Learning%20Prediction%20Model%20Using%20Publicly_Scopus.pdf
http://irep.iium.edu.my/82304/12/82272_Synthesis%20and%20characterization%20of%20bubble_wos.pdf