Journal of Health and Biomedical Informatics
Volume 13, Issue 1 (6-2026)
jhbmi 2026, 13(1): 12-23 |
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:
Valizadeh N, Bolouki Speily O R, Ghorbanzadeh P. An Intelligent Framework Based on Health Informatics for Diagnosing Liver Diseases Using Generative Adversarial Network-Based Data Augmentation and Ensemble Learning. jhbmi 2026; 13 (1) :12-23
URL: http://jhbmi.ir/article-1-941-en.html
URL: http://jhbmi.ir/article-1-941-en.html
Ph.D in Information Technology Engineering, Assistant Proffessor, Department of Information Technology, Faculty of Modern Science & Technology, Urmia University of Technology, Urmia, Iran
Abstract: (2459 Views)
Introduction: Health is one of the fundamental aspects of human life, and chronic diseases pose a serious threat in this realm, entailing far-reaching consequences. With rapid advancements in artificial intelligence, intelligent medical diagnostic systems have seen significant development, facilitating early diagnosis, cost reduction, and more accurate clinical decision-making. Despite the application of various data mining techniques, the complexity and high dimensionality of medical data remain challenging. In this context, deep learning methods and generative models serve as powerful tools for analyzing such data.
Methods: This study proposes a novel method for improving the diagnosis of liver diseases, based on data preprocessing and augmentation using Generative Adversarial Networks (GANs) alongside a stacking ensemble learning approach. The Indian Liver Patient Dataset (ILPD) was used for evaluation. In the initial phase, the ensemble model achieved an accuracy of 77% using 10-fold cross-validation. Subsequently, the model was retrained using a GAN-based data augmentation technique, yielding an impressive accuracy of 99%.
Results: The results indicate that the appropriate application of preprocessing and data augmentation methods has a significant impact on improving model performance and enhancing the accuracy of early liver disease detection.
Conclusion: The proposed intelligent framework successfully overcomes the challenges of complexity and data imbalance in medical datasets, delivering a highly accurate and reliable model. This approach can serve as an efficient clinical decision-support system within the field of health informatics, offering significant assistance to healthcare professionals in the early diagnosis and improved management of liver diseases.
Methods: This study proposes a novel method for improving the diagnosis of liver diseases, based on data preprocessing and augmentation using Generative Adversarial Networks (GANs) alongside a stacking ensemble learning approach. The Indian Liver Patient Dataset (ILPD) was used for evaluation. In the initial phase, the ensemble model achieved an accuracy of 77% using 10-fold cross-validation. Subsequently, the model was retrained using a GAN-based data augmentation technique, yielding an impressive accuracy of 99%.
Results: The results indicate that the appropriate application of preprocessing and data augmentation methods has a significant impact on improving model performance and enhancing the accuracy of early liver disease detection.
Conclusion: The proposed intelligent framework successfully overcomes the challenges of complexity and data imbalance in medical datasets, delivering a highly accurate and reliable model. This approach can serve as an efficient clinical decision-support system within the field of health informatics, offering significant assistance to healthcare professionals in the early diagnosis and improved management of liver diseases.
Keywords: Early Liver Disease Diagnosis, Deep learning, Adversarial Generating Network, Machine Learning
Type of Study: Original Article |
Subject:
Artificial Intelligence in Healthcare
Received: 2025/08/3 | Accepted: 2026/03/30
Received: 2025/08/3 | Accepted: 2026/03/30
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |