EfficientNet deep learning model for pneumothorax disease detection in chest X-ray images

Aya Migdady, Ahmad Alaiad and Ra'ed M. Al-Khatib

International Journal of Business Information Systems, 2025, vol. 50, issue 1, 1-21

Abstract: Pneumothorax is a serious, life-threatening disease resulting from a collapsed lung and respiratory distress that requires prompt and effective diagnosis. The common diagnoses of pneumothorax disease are reliant on chest X-rays. Automatic ways that can fulfill the accurate detection of pneumothorax from chest X-ray images assist radiologists in diagnoses and treatment. In this paper, convolution neural network (CNN) model instituted on the EfficientNet-B3 architecture was proposed to train a dataset of pneumothorax chest X-ray images. The used pneumothorax dataset has 2,027 chest X-ray images. Data augmentation processes have been applied before feeding the main proposed model with these data. This prepossessing technique is required to solve data imbalanced issues and to enhance utmost accuracy. The dataset is split into training and testing data with 80:20 ratio. The experimental outputs of this study proved that the proposed approach can accurately detect the afflicted pneumothorax chest X-ray images with the best accuracy results reaching 97.26% at the testing dataset, when compared with other state-of-the-art models. Consequently, these strong outcomes with the efficient performance of final obtained outcomes will increase the impact of clinical training and patient nursing.

Keywords: artificial; intelligence;; convolution; neural; network;; CNN;; EfficientNet-B3;; data; augmentation;; X-ray; images;; pneumothorax. (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.inderscience.com/link.php?id=148498 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ids:ijbisy:v:50:y:2025:i:1:p:1-21

Access Statistics for this article

More articles in International Journal of Business Information Systems from Inderscience Enterprises Ltd
Bibliographic data for series maintained by Sarah Parker ().