Tuberculosis Disease Prediction System based on Machine Learning and Generative Data
Journal of Artificial Intelligence in Medical Applications · 2025
DOI: 10.65511/jaima.v1i1.741 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Artificial Intelligence in Medical Applications, 2025. Verfasst von Enas A. Raheem und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Tuberculosis (TB) continues to be a prominent global health issue that necessitates accurate diagnostic methods. This research presents a novel (TB) prediction system that classifies individuals into three distinct categories: individuals with recently diagnosed TB, individuals experiencing TB relapses, and healthy subjects. The system utilizes Deoxyribonucleic Acid (DNA)-based biomarkers and clinical tests obtained from TB stage prognosis, thereby enhancing accuracy in classification. An organized dataset comprising 99 subjects guarantees fair class representation. In order to enhance the quality of data, an autoencoder is utilized, which effectively increases the number of classes instances and enables more resilient training. The DNA-based biomarkers that have been extracted play a crucial role as discriminative features, allowing for precise differentiation between the specified groups. For performance evaluation, two classification algorithms were utilized, namely K-Nearest Neighbors (KNN) and an ensemble of bagged decision trees. Both methodologies result in a remarkable approximately 81% success rate. To overcome the limitations imposed by data constraints, Generative Adversarial Network (GAN), which has demonstrated significant enhancements, was used. The Naïve Bayes classifier achieves an accuracy of approximately 91% when trained with GAN-augmented data. Our efforts aim to contribute to the early detection of tuberculosis to enhance patient outcomes and strengthen the global battle against this widespread affliction.
Zitieren
APA
Raheem, E. A., Akbar, M. F., M.Dinar, A., Al-Fahdawi, F. H., & Baig, I. (2025). Tuberculosis Disease Prediction System based on Machine Learning and Generative Data. Journal of Artificial Intelligence in Medical Applications. https://doi.org/10.65511/jaima.v1i1.741
Vancouver
Raheem EA, Akbar MF, M.Dinar A, Al-Fahdawi FH, Baig I. Tuberculosis Disease Prediction System based on Machine Learning and Generative Data. Journal of Artificial Intelligence in Medical Applications. 2025. doi:10.65511/jaima.v1i1.741
BibTeX
@article{Raheem2025,
author = {Raheem, Enas A. and Akbar, Muhammad Firdaus and M.Dinar, Ahmed and Al-Fahdawi, Fawzi Hassan and Baig, Imran},
title = {Tuberculosis Disease Prediction System based on Machine Learning and Generative Data},
journal = {Journal of Artificial Intelligence in Medical Applications},
year = {2025},
doi = {10.65511/jaima.v1i1.741}
}
EUVIMED