Explainable Transformer and Machine Learning Models in Predicting Tuberculosis Treatment Outcomes. A Systematic Review
DOI:
https://doi.org/10.30871/jaic.v10i1.11846Keywords:
Tuberculosis, Treatment Outcomes, Transformer Models, Explainable AI, Machine LearningAbstract
Tuberculosis (TB) remains a major health challenge, and predicting treatment outcomes continues to be difficult in real-world settings. Recent advances in Artificial Intelligence (AI), particularly transformer-based models, have shown promise in modelling longitudinal, multimodal, and heterogeneous TB data. However, their clinical adoption is constrained by limited interpretability, fairness concerns, and deployment challenges. This study presents a systematic literature review of explainable transformer and machine learning models used for TB prognosis. Following PRISMA guidelines, searches across ACM, IEEE Xplore, PubMed, and ScienceDirect identified 17 peer-reviewed studies published between 2020 and 2025 that met the inclusion criteria. The review synthesises evidence on predictive performance, explainability techniques, and deployment considerations. Findings indicate that transformer-based and deep learning models generally outperform conventional machine learning approaches on longitudinal and multimodal data. In contrast, traditional models remain competitive for tabular clinical datasets. Explainability approaches are dominated by feature importance methods and SHAP, with limited use of intrinsic transformer interpretability mechanisms. Persistent challenges include data scarcity, limited generalisability, computational overhead, insufficient evaluation of fairness, and weak alignment with real-world TB care workflows. Building on these findings, the study proposes the Explainable Transformer Adoption Model for TB Prognosis (ETAMTB) as a conceptual clinical adoption framework integrating multimodal transformers, explainability layers, clinician-facing interfaces, and deployment enablers. Overall, the review concludes that effective AI adoption in TB care requires balancing predictive performance, interpretability, and equity, and that explainable transformers should currently be viewed as promising but largely experimental tools rather than deployment-ready solutions.
Downloads
References
[1] J. Chakaya et al., “Global Tuberculosis Report 2020 – Reflections on the Global TB burden, treatment and prevention efforts,” Int. J. Infect. Dis., vol. 113, pp. S7–S12, Dec. 2021, doi: 10.1016/j.ijid.2021.02.107.
[2] Global Tuberculosis Report 2023. World Health Organization, 2023.
[3] S. I. Nafisah and G. Muhammad, “Tuberculosis detection in chest radiograph using convolutional neural network architecture and explainable artificial intelligence,” Neural Comput. Appl., vol. 36, no. 1, pp. 111–131, Jan. 2024, doi: 10.1007/S00521-022-07258-6/METRICS.
[4] “Global tuberculosis report 2024 World Health Organization,” Nov. 2024.
[5] Y. Luo et al., “Development of diagnostic algorithm using machine learning for distinguishing between active tuberculosis and latent tuberculosis infection,” BMC Infect. Dis., vol. 22, no. 1, Dec. 2022, doi: 10.1186/s12879-022-07954-7.
[6] I. Shah, V. Poojari, and H. Meshram, “Multi-Drug Resistant and Extensively-Drug Resistant Tuberculosis,” Indian J. Pediatr., vol. 87, no. 10, pp. 833–839, Oct. 2020, doi: 10.1007/S12098-020-03230-1/METRICS.
[7] M. C. Hosu, L. M. Faye, and T. Apalata, “Predicting Treatment Outcomes in Patients with Drug-Resistant Tuberculosis and Human Immunodeficiency Virus Coinfection, Using Supervised Machine Learning Algorithm,” Pathogens, vol. 13, no. 11, Nov. 2024, doi: 10.3390/pathogens13110923.
[8] L. S. Peetluk et al., “A Clinical Prediction Model for Unsuccessful Pulmonary Tuberculosis Treatment Outcomes,” Clin. Infect. Dis., vol. 74, no. 6, pp. 973–982, Mar. 2022, doi: 10.1093/cid/ciab598.
[9] A. Gupta, V. Kumar, S. Natarajan, and R. Singla, “Adverse drug reactions & drug interactions in MDR-TB patients,” Indian J. Tuberc., vol. 67, no. 4, pp. S69–S78, Dec. 2020, doi: 10.1016/J.IJTB.2020.09.027.
[10] V. A. Dartois and E. J. Rubin, “Anti-tuberculosis treatment strategies and drug development: challenges and priorities,” Nat. Rev. Microbiol., vol. 20, no. 11, pp. 685–701, Nov. 2022, doi: 10.1038/S41579-022-00731-Y;SUBJMETA.
[11] A. Z. Peng et al., “Explainable machine learning for early predicting treatment failure risk among patients with TB-diabetes comorbidity,” Sci. Rep., vol. 14, no. 1, Dec. 2024, doi: 10.1038/s41598-024-57446-8.
[12] L. Pantaleon, “Why measuring outcomes is important in health care,” J. Vet. Intern. Med., vol. 33, no. 2, pp. 356–362, Mar. 2019, doi: 10.1111/jvim.15458.
[13] A. Sambarey et al., “Integrative analysis of multimodal patient data identifies personalized predictors of tuberculosis treatment prognosis,” iScience, vol. 27, no. 2, Feb. 2024, doi: 10.1016/j.isci.2024.109025.
[14] S. K. Pandey, K. U. Singh, R. S. Dingankar, K. Jadhav, K. Gupta, and R. K. Yadav, “Prediction of Tuberculosis Disease Progression with AI Analysis of Clinical Data,” 2023, doi: 10.1109/ICAIIHI57871.2023.10489091.
[15] M. C. Hosu, L. M. Faye, and T. Apalata, “Comorbidities and Treatment Outcomes in Patients Diagnosed with Drug-Resistant Tuberculosis in Rural Eastern Cape Province, South Africa,” Diseases, vol. 12, no. 11, Nov. 2024, doi: 10.3390/diseases12110296.
[16] M. C. Hosu, L. M. Faye, and and T. Apalata, “Predicting Treatment Outcomes in Patients with Drug-resistant Tuberculosis and HIV Co-infection Using a Supervised Machine Learning Algorithm.” Sep. 2024, doi: 10.20944/preprints202409.1747.v1.
[17] O. A. Hussain and K. N. Junejo, “Predicting treatment outcome of drug-susceptible tuberculosis patients using machine-learning models,” Informatics Heal. Soc. Care, vol. 44, no. 2, pp. 135–151, Apr. 2019, doi: 10.1080/17538157.2018.1433676.
[18] S. Hadebe, B. Ndlovu, and K. Maguraushe, “Managing Diabetes Using Machine Learning and Digital Twins,” Indones. J. Innov. Appl. Sci., vol. 5, no. 2, pp. 145–162, 2025, doi: 10.47540/ijias.v5i2.1981.
[19] Y. Wang, Y. Xu, Z. Yang, X. Liu, and Q. Dai, “Using Recursive Feature Selection with Random Forest to Improve Protein Structural Class Prediction for Low-Similarity Sequences,” Comput. Math. Methods Med., vol. 2021, 2021, doi: 10.1155/2021/5529389.
[20] Y. Yang, L. Xu, L. Sun, P. Zhang, and S. S. Farid, “Machine learning application in personalised lung cancer recurrence and survivability prediction,” Comput. Struct. Biotechnol. J., vol. 20, pp. 1811–1820, Jan. 2022, doi: 10.1016/j.csbj.2022.03.035.
[21] M. Asad, A. Mahmood, and M. Usman, “A machine learning-based framework for Predicting Treatment Failure in tuberculosis: A case study of six countries,” Tuberculosis, vol. 123, p. 101944, Jul. 2020, doi: 10.1016/J.TUBE.2020.101944.
[22] W. L. W. Foh, S. L. Ang, C. Y. Lim, A. A. L. Alaga, and G. H. Yeap, “Prediction of Tuberculosis Patients’ Treatment Outcomes Using Multinomial Naive Bayes Algorithm and Class-Imbalanced Data,” 2023, doi: 10.1109/GlobConET56651.2023.10150132.
[23] Y. S. Zakaria, N. A. Ariffin, A. Ahmad, R. Rainis, A. M. Muslim, and W. M. M. Wan Ibrahim, “Optimizing Tuberculosis Treatment Predictions: A Comparative Study of XGBoost with Hyperparameter in Penang, Malaysia,” Sains Malaysiana, vol. 54, no. 1, pp. 3741–3752, Jan. 2025, doi: 10.17576/jsm-2025-5401-22.
[24] B. Nansamba, J. Nakatumba-Nabende, A. Katumba, and D. P. Kateete, “A Systematic Review on Application of Multimodal Learning and Explainable AI in Tuberculosis Detection,” IEEE Access, vol. 13. Institute of Electrical and Electronics Engineers Inc., pp. 62198–62221, 2025, doi: 10.1109/ACCESS.2025.3558878.
[25] A. Sambarey et al., “Integrative analysis of multimodal patient data identifies efficacious drug regimens and personalized predictors of tuberculosis treatment prognosis.”
[26] S. Madan, M. Lentzen, J. Brandt, D. Rueckert, M. Hofmann-Apitius, and H. Fröhlich, “Transformer models in biomedicine,” BMC Medical Informatics and Decision Making, vol. 24, no. 1. BioMed Central Ltd, Dec. 2024, doi: 10.1186/s12911-024-02600-5.
[27] T. Lin, Y. Wang, X. Liu, and X. Qiu, “A survey of transformers,” AI Open, vol. 3, pp. 111–132, Jan. 2022, doi: 10.1016/J.AIOPEN.2022.10.001.
[28] S. Nerella et al., “Transformers in Healthcare: A Survey.”
[29] A. Vaswani et al., “Attention Is All You Need,” 2017.
[30] S. Yadav, N. Jeyaraman, M. Jeyaraman, and G. Rawal, “Artificial intelligence in tuberculosis diagnosis: Revolutionizing detection and treatment,” IP Indian J. Immunol. Respir. Med., vol. 9, no. 2, pp. 85–87, 2024, doi: 10.18231/j.ijirm.2024.017.
[31] “CASP Checklists - Critical Appraisal Skills Programme.” .
[32] S. Santhanam, V. Ravindran, and C. Wincup, “Critical appraisal of an original research article,” J. R. Coll. Physicians Edinb., Sep. 2025, doi: 10.1177/14782715251369964.
[33] M. Nijiati et al., “Deep learning and radiomics of longitudinal CT scans for early prediction of tuberculosis treatment outcomes,” Eur. J. Radiol., vol. 169, Dec. 2023, doi: 10.1016/j.ejrad.2023.111180.
[34] G. Mate Landry, R. Nsimba Malumba, F. C. Balanganayi Kabutakapua, and B. Boluma Mangata, “PERFORMANCE COMPARISON OF CLASSICAL ALGORITHMS AND DEEP NEURAL NETWORKS FOR TUBERCULOSIS PREDICTION,” J. Techno Nusa Mandiri, vol. 21, no. 2, pp. 126–133, Sep. 2024, doi: 10.33480/techno.v21i2.5609.
[35] M. Zhan et al., “A clinical indicator-based prognostic model predicting treatment outcomes of pulmonary tuberculosis: a prospective cohort study,” BMC Infect. Dis., vol. 23, no. 1, Dec. 2023, doi: 10.1186/s12879-023-08053-x.
[36] L. S. Peetluk et al., “A clinical prediction model for unsuccessful pulmonary tuberculosis treatment outcomes Sterling 4* , and the Regional Prospective Observational Research in Tuberculosis (RePORT)-Brazil network,” doi: 10.1093/cid/ciab598/6313211.
[37] A. D. Orjuela-Cañón, A. F. Romero-Gómez, A. L. Jutinico, C. E. Awad, E. Vergara, and M. A. Palencia, “Data Fusion of Medical Records and Clinical Data to Enhance Tuberculosis Diagnosis in Resource-Limited Settings,” Appl. Sci., vol. 15, no. 10, May 2025, doi: 10.3390/app15105423.
[38] H. Chefer, S. Gur, and L. Wolf, “Transformer Interpretability Beyond Attention Visualization.”
[39] F. Zhang, F. Zhang, L. Li, and Y. Pang, “Clinical utilization of artificial intelligence in predicting therapeutic efficacy in pulmonary tuberculosis,” Journal of Infection and Public Health, vol. 17, no. 4. Elsevier Ltd, pp. 632–641, Apr. 2024, doi: 10.1016/j.jiph.2024.02.012.
[40] O. Boursalie, R. Samavi, T. E. Doyle, and O. Boursalie, “Decoder Transformer for Temporally-Embedded Health Outcome Predictions,” in Proceedings - 20th IEEE International Conference on Machine Learning and Applications, ICMLA 2021, 2021, pp. 1461–1467, doi: 10.1109/ICMLA52953.2021.00235.
[41] S. A. Fayaz et al., “Machine learning algorithms to predict treatment success for patients with pulmonary tuberculosis,” PLoS One, vol. 19, no. 10 October, Oct. 2024, doi: 10.1371/journal.pone.0309151.
[42] S. You et al., “Predicting resistance to fluoroquinolones among patients with rifampicin-resistant tuberculosis using machine learning methods,” PLOS Digit. Heal., vol. 1, no. 6 June, Jun. 2022, doi: 10.1371/journal.pdig.0000059.
[43] L. Shichman et al., “Predictive Modeling and Machine Learning Insights into Multi-Drug Resistant Tuberculosis Treatment Outcomes,” in 2025 IEEE Systems and Information Engineering Design Symposium, SIEDS 2025, 2025, pp. 185–190, doi: 10.1109/SIEDS65500.2025.11021160.
[44] S. Chishakwe, N. Moyo, B. M. Ndlovu, and S. Dube, “Intrusion Detection System for IoT environments using Machine Learning Techniques,” 2022 1st Zimbabwe Conf. Inf. Commun. Technol. ZCICT 2022, pp. 1–7, 2022, doi: 10.1109/ZCICT55726.2022.10045992.
[45] B. Ndlovu, F. J. Kiwa, M. Muduva, and C. T. Chipfumbu, “Developing a Logistic Regression Machine Learning Model that Predicts Viral Load Outcomes for Children Living with HIV in Gutu District , Zimbabwe,” Indones. J. Innov. Appl. Sci., pp. 277–304, 2025, doi: 10.47540/ijias.v5i3.2275.
[46] H. Chamboko and B. Ndlovu, “Twitter ( X ) Sentiment Analysis on Monkeypox : A Systematic Literature Review,” vol. 14, no. 2, pp. 629–639, 2025, doi: 10.14421/ijid.2025.5196.
[47] N. W.C Mukura and B. Ndlovu, “Performance Evaluation of Artificial Intelligence in Decision Support System for Heart Disease Risk Prediction,” no. Who 2018, pp. 83–93, 2023, doi: 10.46254/ap04.20230043.
[48] M. Nijiati et al., “Deep learning on longitudinal CT scans: automated prediction of treatment outcomes in hospitalized tuberculosis patients,” iScience, vol. 26, no. 11, Nov. 2023, doi: 10.1016/j.isci.2023.108326.
[49] A. K. Mondal, A. Bhattacharjee, P. Singla, and A. P. Prathosh, “XViTCOS: Explainable Vision Transformer Based COVID-19 Screening Using Radiography,” IEEE J. Transl. Eng. Heal. Med., vol. 10, 2022, doi: 10.1109/JTEHM.2021.3134096.
[50] B. Ndlovu, K. Maguraushe, and O. Mabikwa, “A Comparative Analysis of Machine Learning Techniques and Explainable AI on Voice Biomarkers for Effective Parkinson’s Disease Prediction,” J. Inf. Syst. Informatics, vol. 7, no. 3, pp. 2196–2228, Sep. 2025, doi: 10.51519/JOURNALISI.V7I3.1172.
[51] T. Ngwazi and B. Ndlovu, “Early Detection of Diabetic Retinopathy Through Explainable AI Models : A Systematic Review,” vol. 14, no. 2, pp. 616–628, 2025, doi: 10.14421/ijid.2025.5200.
[52] B. Ndlovu, K. Maguraushe, and O. Mabikwa, “Machine Learning and Explainable AI for Parkinson’s Disease Prediction: A Systematic Review,” Indones. J. Comput. Sci., vol. 14, no. 2, 2025, doi: https://doi.org/10.33022/ijcs.v14i2.4837.
[53] S. Sathitratanacheewin, P. Sunanta, and K. Pongpirul, “Deep learning for automated classification of tuberculosis-related chest X-Ray: dataset distribution shift limits diagnostic performance generalizability,” Heliyon, vol. 6, no. 8, Aug. 2020, doi: 10.1016/j.heliyon.2020.e04614.
[54] D. Saraswat et al., “Explainable AI for Healthcare 5.0: Opportunities and Challenges,” IEEE Access, vol. 10. Institute of Electrical and Electronics Engineers Inc., pp. 84486–84517, 2022, doi: 10.1109/ACCESS.2022.3197671.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Shumirai Sibanda, Belinda Ndlovu
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License (Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) ) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
