Widiasri, Monica and Siswantoro, Joko and Valentina Shen, Calista Pneumonia Detection System with GAN-based Augmentation in Chest X-Ray Images Using Hybrid CNN-SVM. In: The 2026 International Conference on Advancement in Data Science, E-learning and Information System (ICADEIS) 2026, 3 – 24 June 2026, Bandung. (Submitted)
![[thumbnail of Final manuscript]](/style/images/fileicons/application_pdf.png) |
PDF (Final manuscript)
icadeis-final manuscript-1571256136-monica widiasri.pdf - Accepted Version Restricted to Registered users only Download (542kB) | Request a copy |
Abstract
Pneumonia is a lung infection caused by viruses, bacteria, or fungi, with symptoms including cough, fever, and shortness of breath. Pneumonia can be life-threatening if not treated promptly and appropriately. Early detection of pneumonia can be achieved through chest x-rays performed by radiologists, which are prone to subjective variability. This study proposes a pneumonia detection system from chest x-rays to assist in chest x-ray examinations. However, the number of publicly available chest X-ray datasets used to build detection models is very limited and imbalanced between classes. Therefore, in this study, traditional and Wasserstein-penalized gradient generative adversarial network (WGAN-GP) augmentation is applied to the training data to build a hybrid detection model of convolutional neural network (CNN) using DenseNet121 and support vector machine (SVM). This system was implemented in a Flask-based web application as a diagnostic tool that can aid medical decision-making. The best pneumonia detection results were obtained from a system that combined traditional augmentation and WGAN-GP into a CNN-SVM hybrid detection model achieving an accuracy of 93.23%, compared to a detection model without augmentation achieving an accuracy of 91.12%. Experimental results show that augmentation can improve the performance of the pneumonia detection system with limited and imbalanced dataset.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Faculty of Engineering > Department of Informatic |
| Depositing User: | Monica Widiasri 61151 |
| Date Deposited: | 03 Jun 2026 07:08 |
| Last Modified: | 03 Jun 2026 07:08 |
| URI: | http://repository.ubaya.ac.id/id/eprint/50779 |
Actions (login required)
 |
View Item |
