Artificial Intelligence and Machine Learning, for Thyroid Disease and Thyroid Disorder Prediction and Detection: A Comprehensive Review of Techniques Clinical Features, Applications, Explainability, Challenges and Future Directions

Roshani Koli; Hemangini Patel

doi:10.55524/ijirem.2026.13.5.6

Abstract

Thyroid problems like hypothyroidism and hyperthyroidism are some of the common hormone-related issues. Getting the diagnosis quickly is very important for proper treatment. Traditional ways of diagnosing these problems rely on checking how a person feels and doing blood tests, for TSH, T3 and T4. As more digital health data becomes available, artificial intelligence, machine learning, deep learning and data mining are being used to help predict thyroid diseases faster and better. This review puts together the developments in AI-based thyroid prediction. It covers the kinds of data sets that are used how the data is processed and which features are chosen, machine learning algorithms, ways to combine models and some methods that are inspired by biology. It also looks at deep learning models ways to create data when there is an imbalance and tools that make the AI easier to understand for doctors. Research from studies shows that models such as Random Forest, XGBoost Gradient Boosting, SVM, Decision Trees, KNN and different types of combined models can have very good prediction results. Some of these models have reported accuracy rates of over 98 to 99 percent when they use feature selection methods, like Featurewiz and SelectKBest and techniques to balance the classes like making more data or using GANs to create new data. More recent work has started to use ideas from nature to create optimization algorithms. It also uses types of structures like Kolmogorov–Arnold Networks. It uses methods that make results easier to understand such as SHAP and LIME. These changes help improve both accuracy and transparency. However there are still problems. These problems include data sets that're small and not very different from each other. There is also a problem with some categories having more examples than others. There is not testing done in different places or with different groups. The data sets are not the same in places. Some data is missing. There is not information about patients over time.. There is very little use of these methods, in real hospitals or clinics. Future research should focus on building clear, multimodal and long-term aware AI systems. These systems need to be tested outside the setting and must help doctors make steady real-world decisions, in thyroid care every day.

Keywords

Artificial Intelligence, Machine Learning, Deep Learning, Thyroid Disease, Thyroid Disorder

References

[1] J. Zhong and Y. Wang, “A comparative empirical analysis for robust thyroid disorder detection using machine learning techniques,” Journal of Electrical Systems and Information Technology, vol. 13, Art. no. 12, 2026. Available from: https://doi.org/10.1186/s43067-026-00315-3

[2] H. Ali, S. Ullah, S. Ullah, R. Ahmad, Q. Ali, S. T. Shah, and S. Johar, “Evaluating the relationship between TSH and T4 in thyroid dysfunction,” Journal of Medical & Health Sciences Review, vol. 2, no. 3, 2025. Available from: https://doi.org/10.62019/wtj65f49

[3] M. R. Islam, A. Islam Chowdhury, S. Shama, and M. M. H. Lamyea, “Enhanced thyroid disease prediction using ensemble machine learning: A high-accuracy approach with feature selection and class balancing,” Discover Artificial Intelligence, vol. 5, Art. no. 9, 2025. Available from: https://doi.org/10.1007/s44163-025-00225-9

[4] D. Kesavulu and R. Kannadasan, “Improved bio-inspired with machine learning computing approach for thyroid prediction,” Scientific Reports, vol. 15, Art. no. 22524, 2025. Available from: https://doi.org/10.1038/s41598-025-03299-8

[5] A. Top?ir, F. Güler, E. Çetin, M. F. Burak, and M. A?raz, “Thyroid disease classification using generative adversarial networks and Kolmogorov-Arnold network for three-class classification,” BMC Medical Informatics and Decision Making, vol. 25, Art. no. 284, 2025. Available from: https://doi.org/10.1186/s12911-025-03014-7

[6] M. B. Hossain, A. Shama, A. Adhikary, A. D. Raha, K. M. A. Uddin, M. A. Hossain, I. Islam, S. A. Murad, M. S. Munir, and A. K. Bairagi, “An explainable artificial intelligence framework for the predictive analysis of hypo and hyper thyroidism using machine learning algorithms,” Human-Centric Intelligent Systems, vol. 3, pp. 211–231, 2023. Available from: https://doi.org/10.1007/s44230-023-00027-1

[7] S. T. Lim, K. W. Khaw, X. Y. Chew, and W. C. Yeong, “Prediction of thyroid disease using machine learning approaches and Featurewiz selection,” Journal of Telecommunication, Electronic and Computer Engineering, vol. 15, no. 3, 2023. Available from: https://doi.org/10.54554/jtec.2023.15.03.002

[8] T. Alyas, M. Hamid, K. Alissa, T. Faiz, N. Tabassum, and A. Ahmad, “Empirical method for thyroid disease classification using a machine learning approach,” BioMed Research International, vol. 2022, Art. no. 9809932, 2022. Available from: https://doi.org/10.1155/2022/9809932

[9] R. Chaganti, F. Rustam, I. De La Torre Díez, J. L. V. Mazón, C. L. Rodríguez, and I. Ashraf, “Thyroid disease prediction using selective features and machine learning techniques,” Cancers, vol. 14, no. 16, Art. no. 3914, 2022. Available from: https://doi.org/10.3390/cancers14163914

[10] S. S. Islam, M. S. Haque, M. S. U. Miah, T. B. Sarwar, and R. Nugraha, “Application of machine learning algorithms to predict the thyroid disease risk: An experimental comparative study,” PeerJ Computer Science, vol. 8, Art. no. e898, 2022. Available from: https://doi.org/10.7717/peerj-cs.898

[11] L. Aversano, M. L. Bernardi, M. Cimitile, M. Iammarino, P. E. Macchia, I. C. Nettore, and C. Verdone, “Thyroid disease treatment prediction with machine learning approaches,” Procedia Computer Science, vol. 192, pp. 1031–1040, 2021. Available from: https://doi.org/10.1016/j.procs.2021.08.106

[12] H. A. U. Rehman, C. Y. Lin, Z. Mushtaq, and S. F. Su, “Performance analysis of machine learning algorithms for thyroid disease,” Arabian Journal for Science and Engineering, vol. 46, no. 10, pp. 9437–9449, 2021. Available from: https://doi.org/10.1007/s13369-020-05206-x

[13] K. Salman and E. Sonuç, “Thyroid disease classification using machine learning algorithms,” Journal of Physics: Conference Series, vol. 1963, no. 1, Art. no. 012140, 2021. Available from: https://doi.org/10.1088/1742-6596/1963/1/012140

[14] U. Pawar, D. O’Shea, S. Rea, and R. O’Reilly, “Incorporating explainable artificial intelligence (XAI) to aid the understanding of machine learning in the healthcare domain,” CEUR Workshop Proceedings, vol. 2771, pp. 169–180, 2020. Available from: https://ceur-ws.org/Vol-2771/

[15] A. Begum and A. Parkavi, “Prediction of thyroid disease using data mining techniques,” in Proceedings of the 5th International Conference on Advanced Computing & Communication Systems (ICACCS), 2019, pp. 342–345. Available from: https://doi.org/10.1109/ICACCS.2019.8728320

[16] D. Dua and C. Graff, UCI Machine Learning Repository. Irvine, CA, USA: University of California, Irvine, School of Information and Computer Sciences, 2019. Available from: https://doi.org/10.24432/C5D010

[17] A. Tyagi, R. Mehra, and A. Saxena, “Interactive thyroid disease prediction system using machine learning technique,” in Proceedings of the 5th IEEE International Conference on Parallel, Distributed and Grid Computing (PDGC), 2018, pp. 689–693. Available from: https://doi.org/10.1109/PDGC.2018.8745910

[18] W. Ahmad, L. Huang, A. Ahmad, F. Shah, A. Iqbal, and A. Saeed, “Thyroid diseases forecasting using a hybrid decision support system based on ANFIS, k-NN and information gain method,” Journal of Applied Environmental and Biological Sciences, vol. 7, no. 10, pp. 78–85, 2017. Available from: https://www.researchgate.net/publication/320350249_Thyroid_Diseases_Forecasting_Using_a_Hybrid_Decision_Support_System_Based_on_ANFIS_k-NN_and_Information_Gain_Method

[19] K. Chandel, V. Kunwar, S. Sabitha, T. Choudhury, and S. Mukherjee, “A comparative study on thyroid disease detection using K-nearest neighbor and Naive Bayes classification techniques,” CSI Transactions on ICT, vol. 4, no. 2, pp. 313–319, 2016. Available from: https://doi.org/10.1007/s40012-016-0100-5

[20] I. Ionit? and L. Ionit?, “Prediction of thyroid disease using data mining techniques,” BRAIN. Broad Research in Artificial Intelligence and Neuroscience, vol. 7, no. 3, pp. 115–124, 2016. Available from: https://editora.ro/index.php/brain/article/view/211

Cites this article as

R. Koli, H. Patel, "Artificial Intelligence and Machine Learning, for Thyroid Disease and Thyroid Disorder Prediction and Detection: A Comprehensive Review of Techniques Clinical Features, Applications, Explainability, Challenges and Future Directions", International Journal of Innovative Research in Engineering and Management (IJIREM), Vol-13, Issue-5, Page No-47-53, 2026. Available from: https://doi.org/10.55524/ijirem.2026.13.5.6