Detection of COVID-19 and Viral Pneumonia from Chest X-Ray Images with Deep Learning

Pınar Tüfekçi; Burak Gezici

Article Details

Detection of COVID-19 and Viral Pneumonia from Chest X-Ray Images with Deep Learning

Authors:

Pınar Tüfekçi | Burak Gezici

Publisher:

Afyon Kocatepe University

Main Topics:

Yapay Zeka

Journal:

Afyon Kocatepe Üniversitesi Fen Ve Mühendislik Bilimleri Dergisi Vol.23 Issue 1, pp.89-100

Document Type:

Article

Publication Date:

1/2023

Document Language:

Turkish

Document ID (Org No.):

2001284431

pISSN:

2149-3367

Keywords:

Abstract:

 In todays conditions, although the pandemic has lost its effects and is in the process of ending, COVID-19 still shows its effects on people as mild. With the developments in image processing and Artificial Intelligence technologies, the correct detection of such viruses in the early stages will both help the healing process of the disease quickly by applying the right treatment, and will alleviate the burden on health systems. In this study, it has been tried to create a high-accuracy and reliable model that tries to distinguish COVID-19 and viral pneumonia diseases from chest X-ray images. For this purpose, a comprehensive modeling study has been carried out by applying the AlexNet and GoogleNet special architectures of convolutional neural networks, which are deep learning algorithms, directly to their original versions and to their improved versions with transfer learning. The data set used in the modeling process COVID-19 Radiography Database is a popular data set and an unbalanced data set with 3 classes and a different number of samples in each class. By applying data reduction and increase methods to this data set, 2 new balanced data sets containing equal number of samples in each class were created. By dividing the original dataset and newly created datasets into training and test datasets at a rate of 80-20, and also by cross validating 3, 5, and 10 times, model performances were measured and the model with the best performance was tried to be found. As a result, the best model was found with 99.90 accuracy, as the data set balanced by data augmentation method was divided according to cross validation 10 times and AlexNet architecture developed with transfer learning was applied.

 In todays conditions, although the pandemic has lost its effects and is in the process of ending, COVID-19 still shows its effects on people as mild. With the developments in image processing and Artificial Intelligence technologies, the correct detection of such viruses in the early stages will both help the healing process of the disease quickly by applying the right treatment, and will alleviate the burden on health systems. In this study, it has been tried to create a high-accuracy and reliable model that tries to distinguish COVID-19 and viral pneumonia diseases from chest X-ray images. For this purpose, a comprehensive modeling study has been carried out by applying the AlexNet and GoogleNet special architectures of convolutional neural networks, which are deep learning algorithms, directly to their original versions and to their improved versions with transfer learning. The data set used in the modeling process COVID-19 Radiography Database is a popular data set and an unbalanced data set with 3 classes and a different number of samples in each class. By applying data reduction and increase methods to this data set, 2 new balanced data sets containing equal number of samples in each class were created. By dividing the original dataset and newly created datasets into training and test datasets at a rate of 80-20, and also by cross validating 3, 5, and 10 times, model performances were measured and the model with the best performance was tried to be found. As a result, the best model was found with 99.90 accuracy, as the data set balanced by data augmentation method was divided according to cross validation 10 times and AlexNet architecture developed with transfer learning was applied.

 In todays conditions, although the pandemic has lost its effects and is in the process of ending, COVID-19 still shows its effects on people as mild. With the developments in image processing and Artificial Intelligence technologies, the correct detection of such viruses in the early stages will both help the healing process of the disease quickly by applying the right treatment, and will alleviate the burden on health systems. In this study, it has been tried to create a high-accuracy and reliable model that tries to distinguish COVID-19 and viral pneumonia diseases from chest X-ray images. For this purpose, a comprehensive modeling study has been carried out by applying the AlexNet and GoogleNet special architectures of convolutional neural networks, which are deep learning algorithms, directly to their original versions and to their improved versions with transfer learning. The data set used in the modeling process COVID-19 Radiography Database is a popular data set and an unbalanced data set with 3 classes and a different number of samples in each class. By applying data reduction and increase methods to this data set, 2 new balanced data sets containing equal number of samples in each class were created. By dividing the original dataset and newly created datasets into training and test datasets at a rate of 80-20, and also by cross validating 3, 5, and 10 times, model performances were measured and the model with the best performance was tried to be found. As a result, the best model was found with 99.90 accuracy, as the data set balanced by data augmentation method was divided according to cross validation 10 times and AlexNet architecture developed with transfer learning was applied.

 Günümüz şartlarında her ne kadar pandemi etkilerini kaybetmiş ve bitme sürecinde olsa da, COVID-19 halen etkilerini insanlar üzerinde hafif olarak göstermektedir. Yapay Zeka ve görüntü işleme teknolojilerindeki gelişmelerle birlikte, bu tür virüslerin erken aşamalarda doğru bir şekilde tespit edilmesi, hem doğru tedavinin uygulanarak hastalığın iyileşme sürecine hızlı bir şekilde yardımcı olacak hem de sağlık sistemlerinin üzerindeki yükü hafifletmiş olacaktır. Bu çalışmada, göğüs röntgeni görüntülerinden, COVID-19 ve viral pnömoni hastalıklarını ayırt etmeye çalışan, yüksek doğruluklu ve güvenilir bir model oluşturulmaya çalışılmıştır. Bu amaçla, derin öğrenme algoritmalarından olan Evrişimli Sinir Ağlarının AlexNet ve GoogleNet özel mimarilerinin, doğrudan orijinal halleri ve transfer öğrenmeyle geliştirilmiş halleri uygulanarak, geniş kapsamlı bir modelleme çalışması yapılmıştır. Modelleme sürecinde, kullanılan veri seti COVID-19 Radiography Database popüler bir veri seti olup, 3 sınıflı ve her sınıfta farklı sayıda örnek bulunduran dengesiz bir veri setidir. Bu veri setine, veri azaltma ve arttıma yöntemleri uygulanarak, her sınıfta eşit sayıda örnek içeren dengeli 2 yeni veri seti oluşturulmuştur. Orijinal veri seti ve yeni oluşturulan veri setleri, 80-20 oranında eğitim ve test veri setine bölünerek ve aynı zamanda 3, 5 ve 10 kez çapraz doğrulamaya göre bölünerek, model performansları ölçülmüştür ve en iyi performansa sahip model bulunmaya çalışılmıştır. Sonuç olarak, en iyi model, veri arttırma yöntemiyle dengeli hale getirilmiş veri setinin, 10 kez çapraz doğrulamaya göre bölünerek, transfer öğrenme ile geliştirilmiş AlexNet mimarisinin uygulandığı model olarak,  99.90 doğruluk başarısı ile bulunmuştur.

Article Details

Detection of COVID-19 and Viral Pneumonia from Chest X-Ray Images with Deep Learning

Authors:

Pınar Tüfekçi | Burak Gezici

Publisher:

Afyon Kocatepe University

Main Topics:

Yapay Zeka

Journal:

Afyon Kocatepe Üniversitesi Fen Ve Mühendislik Bilimleri Dergisi Vol.23 Issue 1, pp.89-100

Document Type:

Article

Publication Date:

1/2023

Document Language:

Turkish

Document ID (Org No.):

2001284431

pISSN:

2149-3367

COVID-19

Viral pneumonia

Deep learning

AlexNet

GoogleNet

COVID-19

Viral pnömoni

Derin öğrenme

AlexNet

GoogleNet

Issues

Article Details

Authors: Pınar Tüfekçi | Burak Gezici

Publisher: Afyon Kocatepe University

Main Topics: Yapay Zeka

Journal: Afyon Kocatepe Üniversitesi Fen Ve Mühendislik Bilimleri Dergisi Vol.23 Issue 1, pp.89-100

Document Type: Article

Publication Date: 1/2023

Document Language: Turkish

Document ID (Org No.): 2001284431

pISSN: 2149-3367

Issues

Authors:

Pınar Tüfekçi | Burak Gezici

Publisher:

Afyon Kocatepe University

Main Topics:

Yapay Zeka

Journal:

Afyon Kocatepe Üniversitesi Fen Ve Mühendislik Bilimleri Dergisi Vol.23 Issue 1, pp.89-100

Document Type:

Article

Publication Date:

1/2023

Document Language:

Turkish

Document ID (Org No.):

2001284431

pISSN:

2149-3367