Integrating Clinical and Radiological Features for Lumbosacral Radiculopathy (Sciatica) Prediction and a Comparative Analysis of Various Machine Learning Approaches

Sciatica is a neurological condition characterized by Compression of the sciatic nerve causes pain that radiates from the legs to the lower back. Conventional diagnostic approaches, including physical examinations and MRI analysis, are time-consuming, prone to human error, and limited by subjective interpretation. AI and ML have revolutionized industries with their emergence medical diagnostics, offering data-driven solutions to improve accuracy and reduce diagnostic uncertainty. The study assesses various ML models—such as Decision Trees, SVM, Random Forests, Neural Networks, and Gradient Boosting—in predicting sciatica using clinical and imaging data. Recent research suggests that ensemble methods like Random Forest and Gradient Boosting often outperform conventional models in predictive performance, making them strong candidates for sciatica diagnosis. However, the interpretability of complex models, such as deep learning architectures, remains a crucial factor in clinical adoption. The study further evaluates the trade-offs between predictive accuracy and model explainability to determine the most suitable ML approach for real-world clinical applications. Additionally, AI-driven diagnostic systems can facilitate early detection, reduce the risk of chronic pain, and minimize the need for invasive procedures. To the research, this contributes findings the advancement of intelligent diagnostic tools in musculoskeletal healthcare, enhancing clinical decision-making, optimizing diagnostic workflows, and improving patient outcomes. Study highlights the potential of AI in revolutionizing sciatica diagnosis and provides insights into selecting an optimal ML model for effective implementation in clinical practice. In addition, the study incorporates insights from recent deep learning research. Furthermore, AI-driven diagnostic systems offer the potential for early detection, reduced risk of chronic pain progression, and minimized reliance on invasive procedures. Integrating these predictive tools into telemedicine platforms could also enhance access to specialized care in underserved regions. The study underscores the transformative role of AI, particularly machine learning, in the future of sciatica diagnosis.

Sciatica-Detection, Machine Learning, Nerve root Compression, Predictive Analytics, Healthcare AI.

1. Valat, J. P., Genevay, S., Marty, M., Rozenberg, S., & Koes, B. (2010). Sciatica. Best practice & research Clinical rheumatology, 24(2), 241-252. [Google Scholar] [Crossref]

2. Demiryol, D. (2022). Case study of physiotherapeutic treatment about a patient with disc herniation with L5/S1 with radiculopathy. [Google Scholar] [Crossref]

3. Euro, U. (2019). Risk factors for sciatica [Google Scholar] [Crossref]

4. Baloh, R. W. (2019). Sciatica and chronic pain. Neuropathic pain and sciatica. Springer Berlin Heidelberg. [Google Scholar] [Crossref]

5. Ayimbetova, U. (2025). Recent Advances in AI-Driven Diagnostic Systems. Confrencea, 1, 50-54. [Google Scholar] [Crossref]

6. Baur, D., Kroboth, K., Heyde, C. E., & Voelker, A. (2022). Convolutional neural networks in spinal magnetic resonance imaging: a systematic review. World Neurosurgery, 166, 60-70. [Google Scholar] [Crossref]

7. Maleki Varnosfaderani, S., & Forouzanfar, M. (2024). The role of AI in hospitals and clinics: transforming healthcare in the 21st century. Bioengineering, 11(4), 337 [Google Scholar] [Crossref]

8. Custers, P., Van de Kelft, E., Eeckhaut, B., Sabbe, W., Hofman, A., Debuysscher, A., ... & Maes, G. (2024). Clinical examination, diagnosis, and conservative treatment of chronic low back pain: a narrative review. Life, 14(9), 1090. [Google Scholar] [Crossref]

9. Staartjes , V. E., de Wispelaere, M. P., Vandertop, W. P., & Schröder, M. L. (2019). Deep learning-based preoperative predictive analytics for patient-reported outcomes following lumbar discectomy: feasibility of center-specific modeling. The Spine Journal, 19(5), 853-861. [Google Scholar] [Crossref]

10. 10.Berg, B., Gorosito, M. A., Fjeld, O., Haugerud, H., Storheim, K., Solberg, T. K., & Grotle, M. (2024). Machine learning models for predicting disability and pain following lumbar disc herniation surgery. JAMA Network Open, 7(2), e2355024-e2355024. [Google Scholar] [Crossref]

11. 11.Das, A. (2024). Logistic regression. In Encyclopedia of quality of life and well-being research (pp. 3985-3986). Cham: Springer International Publishing. [Google Scholar] [Crossref]

12. 12.Berkhout, S. W., Haaf, J. M., Gronau, Q. F., Heck, D. W., & Wagenmakers, E. J. (2024). A tutorial on Bayesian model-averaged meta-analysis in JASP. Behavior Research Methods, 56(3), 1260-1282. [Google Scholar] [Crossref]

13. 13.Yin, H., Aryani, A., Petrie, S., Nambissan, A., Astudillo, A., & Cao, S. (2024). A rapid review of clustering algorithms. arXiv preprint arXiv:2401.07389. [Google Scholar] [Crossref]

14. 14.Guido, R., Ferrisi, S., Lofaro, D., & Conforti, D. (2024). An overview on the advancements of support vector machine models in healthcare applications: a review. Information, 15(4), 235. [Google Scholar] [Crossref]

15. 15.Kern, C., Klausch, T., & Kreuter, F. (2019, April). Tree-based machine learning methods for survey research. In Survey research methods (Vol. 13, No. 1, p. 73). [Google Scholar] [Crossref]

16. 16.Bentéjac, C., Csörgő, A., & Martínez-Muñoz, G. (2021). A comparative analysis of gradient boosting algorithms. Artificial Intelligence Review, 54, 1937-1967. [Google Scholar] [Crossref]

17. 17.Koes, B. W., Van Tulder, M. W., & Peul, W. C. (2007). Diagnosis and treatment of sciatica. Bmj, 334(7607), 1313-1317. [Google Scholar] [Crossref]

18. 18.Stynes, S., Konstantinou, K., Ogollah, R., Hay, E. M., & Dunn, K. M. (2018). Clinical diagnostic model for sciatica developed in primary care patients with low back-related leg pain. PLoS One, 13(4), e0191852. [Google Scholar] [Crossref]

19. 19.Lu, J. T., Pedemonte, S., Bizzo, B., Doyle, S., Andriole, K. P., Michalski, M. H., ... & Pomerantz, S. R. (2018, November). Deep spine: automated lumbar vertebral segmentation, disc-level designation, and spinal stenosis grading using deep learning. In Machine Learning for Healthcare Conference (pp. 403-419). PMLR. [Google Scholar] [Crossref]

20. 20.Xu, X., Li, J., Zhu, Z., Zhao, L., Wang, H., Song, C., ... & Pei, Y. (2024). A comprehensive review on synergy of multi-modal data and ai technologies in medical diagnosis. Bioengineering, 11(3), 219. [Google Scholar] [Crossref]

21. 21.Abd-Elsayed, A., Robinson, C. L., Marshall, Z., Diwan, S., & Peters, T. (2024). Applications of artificial intelligence in pain medicine. Current Pain and Headache Reports, 28(4), 229-238. [Google Scholar] [Crossref]

22. 22.Marcílio, W. E., & Eler, D. M. (2020, November). From explanations to feature selection: assessing SHAP values as feature selection mechanism. In 2020 33rd SIBGRAPI conference on Graphics, Patterns and Images (SIBGRAPI) (pp. 340-347), IEEE. [Google Scholar] [Crossref]

23. 23.Wang, H., Liang, Q., Hancock, J. T., & Khoshgoftaar, T. M. (2024). Feature selection strategies: a comparative analysis of SHAP-value and importance-based methods. Journal of Big Data, 11(1), 44. [Google Scholar] [Crossref]

24. 24.Jensen, R. K., Kongsted, A., Kjaer, P., & Koes, B. (2019). Diagnosis and treatment of sciatica. Bmj, 367. [Google Scholar] [Crossref]

25. 25.Palanivinayagam, A., & Damaševičius, R. (2023). Effective handling of missing values in datasets for classification using machine learning methods. Information, 14(2), 92. [Google Scholar] [Crossref]

26. 26.Gulowaty, B., & Ksieniewicz, P. (2019). SMOTE algorithm variations in balancing data streams. In Intelligent Data Engineering and Automated Learning–IDEAL 2019: 20th International Conference, Manchester, UK, November 14–16, 2019, Proceedings, Part II 20 (pp. 305-312). Springer International Publishing. [Google Scholar] [Crossref]

27. 27.Ghotra, B., McIntosh, S., & Hassan, A. E. (2017, May). A large-scale study of the impact of feature selection techniques on defect classification models. In 2017 IEEE/ACM 14th International Conference on Mining Software Repositories (MSR) (pp. 146-157). IEEE. [Google Scholar] [Crossref]

• A Survey on Sickle Cell Disease Detection and Analysis