International Journal of Computer Techniques Volume 12 Issue 3 | Tuberculosis Detection using Machine Learning
Tuberculosis Detection using Machine Learning
International Journal of Computer Techniques – Volume 12 Issue 3, May – June 2025
Abstract
Tuberculosis (TB) remains a global health threat, particularly in resource-limited regions. This study presents a **computer-aided detection system based on the EfficientDet network**, enhanced with a **fused attention mechanism**. The model effectively **analyzes chest X-ray images**, minimizing false positives and improving detection accuracy. Evaluations on the **TBX11K dataset** demonstrate significant improvements in precision and recall, making the system highly applicable for **clinical applications**.
Keywords
Tuberculosis Detection, Machine Learning, Deep Learning, EfficientDet, Attention Mechanism, AI in Healthcare, Chest X-ray Analysis.
Conclusion
The **attention-based EfficientDet network** achieved **high detection accuracy**, effectively identifying **TB-related lesions** while overcoming challenges posed by **lung abnormalities and image variability**. Future improvements will focus on **enhancing real-time deployment, refining lesion segmentation accuracy, and integrating multi-modal AI diagnostics**.
References
- World Health Organization (2024). “Global Tuberculosis Report 2024.”
- Carabalí-Isajar ML et al. (2023). “Clinical Manifestations and Immune Response to Tuberculosis.” World J Microbiol Biotechnol.
- Chang, R. I., Chiu, Y. H., & Lin, J. W. (2023). “Two-Stage Classification of Tuberculosis Culture Diagnosis using CNN.” Journal of Supercomputing.
Post Comment