This study presents a machine-learning framework that has been created using routine lab data as a tool to support diagnostic decisions for rheumatologic and autoimmune diseases. A publicly available Iraqi dataset was used, comprising 12,085 biologically reliable patients, 14 diagnostic features, and 7 diagnostic examples. Both the XGBoost and CatBoost primary classifiers were employed, along with a hierarchical integration of multiple expert predictive models. The results were confirmed when the model was tested on all patient samples not used in training. The proposed machine learning framework achieved an accuracy of 0.8606; an overall F1 mean of 0.8535; a balanced accuracy of 0.8452; a Matthews’s correlation coefficient (MCC) of 0.8339; and a two-highest accuracy of 0.9880. These results support differential diagnosis in cases where specific tests may not be available. The implication of the study, based on a machine learning framework, proved to be a valuable tool for supporting diagnostic decision-making, as it provided an improved classification of probabilities, but it does not replace the physician's final decision in diagnosis.
M. F. Mahdi, A. Jahani, and D. H. Abd, “Diagnosis of rheumatic and autoimmune diseases dataset,” Data in Brief, vol. 60, Art. no. 111623, 2025, [Online]. Available: https://doi.org/10.1016/j.dib.2025.111623.
J. D. Reveille, “Biomarkers for diagnosis, monitoring of progression, and treatment responses in ankylosing spondylitis and axial spondyloarthritis,” Clinical Rheumatology, vol. 34, pp. 1009-1018, 2015, [Online]. Available: https://doi.org/10.1007/s10067-015-2949-3.
C. Mohan and S. Assassi, “Biomarkers in rheumatic diseases: how can they facilitate diagnosis and assessment of disease activity?,” BMJ, vol. 351, p. h5079, 2015, [Online]. Available: https://doi.org/10.1136/bmj.h5079.
M. F. Mahdi, A. Jahani, and D. H. Abd, “Fuzzy evaluation and explainable machine learning for diagnosis of rheumatic and autoimmune diseases,” PeerJ Computer Science, vol. 11, p. e3096, 2025, [Online]. Available: https://doi.org/10.7717/peerj-cs.3096.
A. M. Elmesiry et al., “Calibrated, explainable machine learning on routine laboratory data to characterize diagnostic assignment patterns in rheumatic diseases: a retrospective study of 12,085 patients,” BMC Rheumatology, vol. 10, no. 1, Art. no. 10, 2025, [Online]. Available: https://doi.org/10.1186/s41927-025-00607-7.
G. S. Collins et al., “TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods,” BMJ, vol. 385, p. e078378, 2024, [Online]. Available: https://doi.org/10.1136/bmj-2023-078378.
T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, pp. 785-794, 2016, [Online]. Available: https://doi.org/10.1145/2939672.2939785.
L. Prokhorenkova, G. Gusev, A. Vorobev, A. V. Dorogush, and A. Gulin, “CatBoost: unbiased boosting with categorical features,” in Advances in Neural Information Processing Systems, vol. 31, pp. 6638-6648, 2018.
S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” in Advances in Neural Information Processing Systems, vol. 30, pp. 4765-4774, 2017.
H. Peng, F. Long, and C. Ding, “Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 8, pp. 1226-1238, 2005, [Online]. Available: https://doi.org/10.1109/TPAMI.2005.159.
