Signal and Data Processing Journal A scientific journal officially licensed by the Commission for Scientific Publications of the (MSRT). Publisher: Research Ceter for Developmen of Technologies
Volume 23, Issue 1 (6-2026)
JSDP 2026, 23(1): 61-82 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Mohammadi S, Taghavian Asiabi Z, Babagoli M. Multi-class Lung Disease Detection from Enhanced CXR Images Using Ensemble Learning. JSDP 2026; 23 (1) : 5
URL: http://jsdp.rcisp.ac.ir/article-1-1475-en.html
URL: http://jsdp.rcisp.ac.ir/article-1-1475-en.html
KN Toosi University of Technology
Abstract: (9 Views)
In recent years, with the increase in lung disease mortality, especially following the COVID-19 pandemic, early and accurate detection has become a focal point in image processing. While considerable research has been conducted on lung disease detection, a notable gap is observed in the literature regarding the combined and individual effects of image enhancement, lung segmentation, and the application of ensemble learning techniques. In this paper, a novel multi-classification approach based on advanced deep learning algorithms is proposed. The COVID-19 Database, a publicly available CXR image dataset, is utilized to investigate the effects of plain, enhanced, and segmented imaging techniques. Additionally, the potential benefits of combining segmented-enhanced and enhanced-segmented images are investigated. Lung segmentation is performed using a U-Net model, while enhancement of both plain and segmented CXR images is conducted with an advanced residual dense neural network. The preprocessing step involves resizing and normalizing the CXR images before they are input into seven transfer learning models: VGG16, VGG19, ResNet50V2, InceptionV3, MobileNet, InceptionResNetV2, and DenseNet201. The strengths of these models are aggregated using an ensemble approach, and the enhanced framework demonstrates superior performance with 97.03% accuracy, 97.12% precision, 97.03% recall, a 97.1% F1-score, and an AUC of 98.98%, outperforming state-of-the-art research.
Article number: 5
Type of Study: Research |
Subject:
Paper
Received: 2025/06/23 | Accepted: 2026/05/5 | Published: 2026/06/21 | ePublished: 2026/06/21
Received: 2025/06/23 | Accepted: 2026/05/5 | Published: 2026/06/21 | ePublished: 2026/06/21
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |