Journal of Soft Computing Paradigm
IRO Journals
Volume 7 — Issue 4 — December 2025

A Multimodal Approach to Detect the Tuberculosis Using Deep Learning Technique

https://doi.org/10.36548/jscp.2025.4.004
Saranya K. G.
Associate Professor, Department of Computer Science, PSG College of Technology, Coimbatore, Tamil Nadu, India
Akalya A.
Student, Department of Computer Science, PSG College of Technology, Coimbatore, Tamil Nadu, India
PDF
PDF

How to Cite

G., Saranya K., and Akalya A. 2026. “A Multimodal Approach to Detect the Tuberculosis Using Deep Learning Technique”. Journal of Soft Computing Paradigm 7 (4): 362-85. https://doi.org/10.36548/jscp.2025.4.004.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

— Tuberculosis Detection
— Deep Learning
— Hybrid Ensemble Model
— DenseNet
— MobileNet
— Graph Neural Network (GNN)
— Explainable AI (XAI)
— Grad-CAM
— Chest X-ray
— Medical Image Analysis
Published: 31-01-2026

Abstract

The biggest challenges faced byTB (tuberculosis) patients are difficulty in managing TB particularly for those living in low socioeconomic areas and especially in rural communities. Chest X-rays are valuable in diagnosing TB because they are inexpensive and non-invasive for obtaining the data needed. There are many limitations to existing applications using deep learning algorithms to detect TB, including a limited number of images available for use (i.e., an X-ray or radiograph image of the chest), and these applications may require significant computational power. Additionally, many applications do not have a variety of characteristics to support the for diagnosis of TB. Existing deep learning TB detection methods do not provide interpretable output because they produce high rates of false negatives, especially when the disease is clinically unclear. This paper proposes a hybrid multimodal ensemble approach that uses both X-ray images and a systematic set of clinical symptoms to improve the accuracy of clinical-quality TB detection. By combining the advantages of the DenseNet and MobileNet architectures, this model is able to create additional radiographic features. It also uses graph neural networks to learn additional features related to the clinical symptoms, allowing the model to learn the contextual and clinical connections within the input data. The multimodal representations will be used to create overall diagnostic predictions using an attention-based ensemble to combine these representations. The overall prediction accuracy for the multimodal approach is 96% with precision/recall statistics nearly the same resulting in more robust and reduced false positive/false negative predictions compared to predictions from unimodal, image-based approaches. Examples of XAI techniques may be implemented to support transparency and build trust in clinical decisions using visualizations (e.g. Grad-CAM) to illustrate areas of the lung that contribute to the diagnosis of the disease. The Unified Multimodal TB Screening method will provide useful, interpretable and computationally efficient solutions to help automate TB diagnoses with the technologies needed to implement screening in real world, resource-constrained settings.

References

  1. Khaing, Aung Phyo. "A Study on Economic Burden of Tuberculosis Patients (Case Study: South Dagon Township, Yangon Region)(Aung Phyo Khaing, 2024)." PhD diss., MERAL Portal, 2024.
  2. Galbusera, Fabio, and Andrea Cina. "Image annotation and curation in radiology: an overview for machine learning practitioners." European Radiology Experimental 8, no. 1 (2024): 11.
  3. Hansun, Seng, Ahmadreza Argha, Siaw-Teng Liaw, Branko G. Celler, and Guy B. Marks. "Machine and deep learning for tuberculosis detection on chest X-rays: Systematic literature review." Journal of medical Internet research 25 (2023): e43154.
  4. Heiliger, Lars, Anjany Sekuboyina, Bjoern Menze, Jan Egger, and Jens Kleesiek. "Beyond medical imaging-a review of multimodal deep learning in radiology." TechRxiv 19103432 (2022).
  5. Huy, Vo Trong Quang, and Chih-Min Lin. "An improved densenet deep neural network model for tuberculosis detection using chest x-ray images." IEEE Access 11 (2023): 42839-42849.
  6. Kotei, Evans, and Ramkumar Thirunavukarasu. "A comprehensive review on advancement in deep learning techniques for automatic detection of tuberculosis from chest X-ray images." Archives of computational methods in engineering 31, no. 1 (2024): 455-474.
  7. Sharma, Vinayak, Sachin Kumar Gupta, and Kaushal Kumar Shukla. "Deep learning models for tuberculosis detection and infected region visualization in chest X-ray images." Intelligent Medicine 4, no. 2 (2024): 104-113.
  8. Urooj, Shabana, S. Suchitra, Lalitha Krishnasamy, Neelam Sharma, and Nitish Pathak. "Stochastic learning-based artificial neural network model for an automatic tuberculosis detection system using chest X-ray images." IEEE Access 10 (2022): 103632-103643.
  9. Mehrrotraa, Rajat, M. A. Ansari, Rajeev Agrawal, Pragati Tripathi, Md Belal Bin Heyat, Mohammed Al-Sarem, Abdullah Yahya Mohammed Muaad, Wamda Abdelrahman Elhag Nagmeldin, Abdelzahir Abdelmaboud, and Faisal Saeed. "Ensembling of efficient deep convolutional networks and machine learning algorithms for resource effective detection of tuberculosis using thoracic (chest) radiography." IEEE Access 10 (2022): 85442-85458.
  10. Munadi, Khairul, Kahlil Muchtar, Novi Maulina, and Biswajeet Pradhan. "Image enhancement for tuberculosis detection using deep learning." IEEE Access 8 (2020): 217897-217907.
  11. Alsdurf, H., B. Empringham, C. Miller, and A. Zwerling. "Tuberculosis screening costs and cost-effectiveness in high-risk groups: a systematic review." BMC infectious diseases 21, no. 1 (2021): 935.
  12. Sharma, Anubhav, Karamjeet Singh, and Deepika Koundal. "A novel fusion based convolutional neural network approach for classification of COVID-19 from chest X-ray images." Biomedical Signal Processing and Control 77 (2022): 103778.
  13. Bhosale, Yogesh H., and K. Sridhar Patnaik. "IoT deployable lightweight deep learning application for COVID-19 detection with lung diseases using RaspberryPi." In 2022 International conference on IoT and blockchain technology (ICIBT), pp. 1-6. IEEE, 2022.
  14. Malik, Hassaan, Tayyaba Anees, Muhammad Umar Chaudhry, Radomir Gono, Michał Jasiński, Zbigniew Leonowicz, and Petr Bernat. "A novel fusion model of hand-crafted features with deep convolutional neural networks for classification of several chest diseases using X-ray images." IEEE Access 11 (2023): 39243-39268.
  15. Xu, Tianhao, and Zhenming Yuan. "Convolution neural network with coordinate attention for the automatic detection of pulmonary tuberculosis images on chest x-rays." IEEE Access 10 (2022): 86710-86717.
  16. Uslu, Fatmatülzehra, and Anil A. Bharath. "TMS-Net: A segmentation network coupled with a run-time quality control method for robust cardiac image segmentation." Computers in Biology and Medicine 152 (2023): 106422.
  17. Abideen, Zain Ul, Mubeen Ghafoor, Kamran Munir, Madeeha Saqib, Ata Ullah, Tehseen Zia, Syed Ali Tariq, Ghufran Ahmed, and Asma Zahra. "Uncertainty assisted robust tuberculosis identification with bayesian convolutional neural networks." Ieee Access 8 (2020): 22812-22825.
  18. Rahman, Tawsifur, Amith Khandakar, Muhammad Abdul Kadir, Khandaker Rejaul Islam, Khandakar F. Islam, Rashid Mazhar, Tahir Hamid et al. "Reliable tuberculosis detection using chest X-ray with deep learning, segmentation and visualization." Ieee Access 8 (2020): 191586-191601.
  19. Hooda, Rahul, Ajay Mittal, and Sanjeev Sofat. "Tuberculosis detection from chest radiographs: a comprehensive survey on computer-aided diagnosis techniques." Current Medical Imaging Reviews 14, no. 4 (2018): 506-520.
  20. National Library of Medicine. (2024). Open-i: Open Access Biomedical Image Search Engine (Tuberculosis Chest X-ray Dataset). U.S. National Library of Medicine, National Institutes of Health. Available at: https://openi.nlm.nih.gov/