Convolutional Neural Network for Predicting COVID-19 from Chest x-ray Images

dc.contributor.advisorÜnveren, Ahmet (Supervisor)
dc.contributor.authorAlbariqi, Anwar Ali A.
dc.date.accessioned2025-04-08T13:08:08Z
dc.date.available2025-04-08T13:08:08Z
dc.date.issued2022-12
dc.date.submitted2022-12
dc.departmentEastern Mediterranean University, Faculty of Engineering, Dept. of Computer Engineeringen_US
dc.descriptionMaster of Science in Computer Engineering. Institute of Graduate Studies and Research. Thesis (M.S.) - Eastern Mediterranean University, Faculty of Engineering, Dept. of Computer Engineering, 2022. Supervisor: Assist. Prof. Dr. Ahmet Ünveren.en_US
dc.description.abstractThe world has recently witnessed many deaths for all age groups due to the respiratory COVID-19 but detecting this disease in its early stages helps to recover, avoid negative effects, and reduce the outbreak of the disease quickly. Many symptoms of this disease were found, most notably chest infections and shortness of breath resulting from infection with this disease. The goal of this project is to use chest x-rays images to predict whether a person has the COVID-19 or not. In this study, we tested the solution performances for our problem on different versions of the CNN. Such as Mobile Net, CNN with Adam optimizer, CNN with Data Augmentation, CNN with Batch Normalization, CNN with Leaky Relu, CNN with Dropout, CNN with Early Stopping, CNN with Hyper-parameter Tuning, RESNET50, VGG-16, and VGG-19. The results showed that the VGG-19 model outperformed all the models in detecting infection with MERS-Cove quickly and with high accuracy instead of regular examinations that take a long time and thus limit the spread of the disease.en_US
dc.description.abstractÖZ: Dünya son zamanlarda solunum yolu kaynaklı COVID-19 nedeniyle tüm yaş grupları için birçok ölüme tanık oldu, ancak bu hastalığın erken evrelerinde tespit edilmesi iyileşmeye, olumsuz etkilerden kaçınmaya ve hastalığın salgınının hızla azalmasına yardımcı oluyor. Bu hastalığın birçok semptomu bulundu, özellikle de göğüs enfeksiyonları ve bu hastalığa bağlı enfeksiyondan kaynaklanan nefes darlığı. Bu projenin amacı, bir kişinin COVID-19'a sahip olup olmadığını tahmin etmek için göğüs röntgeni görüntülerini kullanmaktır. Bu çalışmada, problemimizin çözüm performanslarını CNN'nin farklı versiyonları üzerinde test ettik. MobileNet, Adam optimizer ile CNN, Veri Artırma ile CNN, Toplu Normalleştirme ile CNN, LeakyRelu ile CNN, Bırakma ile CNN, Early Stopping ile CNN, Hiperparametre Ayarlama ile CNN, RESNET-50, VGG-16 ve VGG-19. Sonuçlar, VGG-19 modelinin, uzun zaman alan ve böylece hastalığın yayılmasını sınırlayan düzenli muayeneler yerine MERS-Cove ile enfeksiyonu hızlı ve yüksek doğrulukla tespit etmede tüm modellerden daha iyi performans gösterdiğini gösterdi.en_US
dc.identifier.citationAlbariqi, Anwar Ali A. (2022). Convolutional Neural Network for Predicting COVID-19 from Chest x-ray Images. Thesis (M.S.), Eastern Mediterranean University, Institute of Graduate Studies and Research, Dept. of Computer Engineering, Famagusta: North Cyprus.en_US
dc.identifier.urihttps://hdl.handle.net/11129/6232
dc.language.isoen
dc.publisherEastern Mediterranean University (EMU) - Doğu Akdeniz Üniversitesi (DAÜ)en_US
dc.relation.publicationcategoryTez
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectComputer Engineering Departmenten_US
dc.subjectComputational intelligence--Deep Learningen_US
dc.subjectComputer assisted--Image processingen_US
dc.subjectIdentification--Data processingen_US
dc.subjectCOVID-19, deep learning, VGG-19, conventional neural network, chest x-raysen_US
dc.titleConvolutional Neural Network for Predicting COVID-19 from Chest x-ray Imagesen_US
dc.typeMaster Thesis

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
AlbariqiAnwarAli_MS.pdf
Size:
1.71 MB
Format:
Adobe Portable Document Format
Description:
Thesis, Master

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.77 KB
Format:
Item-specific license agreed upon to submission
Description: