An Approach for Prediction of Osteoporosis and Its Association with Osteoporotic Fracture Risk Using Machine Learning

Muayad H.  Abdul-Zahra; Salam  Al-augby

doi:10.31642/JoKMC/2018/130103

Authors

Muayad H. Abdul-Zahra Faculty of Computer Science and Mathematics - University of Kufa
Salam Al-augby Department of Computer Science Faculty of Computer Science and Mathematics,University of Kufa

DOI:

https://doi.org/10.31642/JoKMC/2018/130103

Keywords:

Osteoporosis, Machine Learning, DEXA, Osteoporotic fracture, FRAX tool., Big Data

Abstract

Osteoporosis is a prevalent condition characterized by decreased bone mineral density and increased fracture risk, representing a significant global health concern, especially in aging populations. Early detection and risk stratification are critical for preventing fractures and reducing healthcare burdens associated with osteoporosis. A dataset of 1,958 individuals with demographic, lifestyle, and clinical features is analyzed, with the data explicitly split into 75% for training and 25% for testing. Three machine learning models—Random Forest, Decision Tree, and Multilayer Perceptron (MLP)—are developed and evaluated by using accuracy, precision, recall, F1-score, ROC-AUC, and balanced accuracy. The Decision Tree model achieved the highest accuracy (0.884) and perfect precision (1.0), while Random Forest and MLP models demonstrated competitive performance with ROC-AUC values of 0.876 and 0.875, respectively. Feature importance analysis confirmed age as the most influential predictor, with prior fractures, low BMI, glucocorticoid use, and smoking contributing to osteoporosis risk classification. On another side, to determine the degrees of fracture risks and their categories using the Fracture Risk Assessment Tool (FRAX) medical tool.

So, we suggested a system using three machine learning (ML) techniques known for their success in diseases prediction, significant results were obtained compared to previous related studies according to the usual evaluation criteria for ML techniques which used to reach the best one to be relied upon in building the future model for prediction of osteoporosis, among all the algorithms, all evaluation metrics indicated that the Decision Tree (DT) was the best model (Accuracy: 0.884, Precision: 1.000, Recall: 0.767, F1-score: 0.868, and Area Under the Receiver Operating characteristic Curve (AUROC): 0.889), followed by the Random Forest (RF) and then the Feedforward Neural Network (Multi-Layer Perceptron MLP). On another side, to determine the degrees of fracture risks and their categories using the Fracture Risk Assessment Tool (FRAX) medical tool.

This study is, to the best of our information, unique in linking osteoporosis with osteoporotic fracture risks by weighting important factors within using historical healthcare datasets as big data, which provides an integrated approach a solid foundation for building accurate predictive models that support researchers and healthcare professionals in enhancing diagnosis and prevention, thus saving effort, time, and cost for both the patient and the healthcare communities

Downloads

Download data is not yet available.

References

[1] M. Jararweh, M. Daraghmeh, and M. Z. Ali, “Exploiting machine learning for osteoporosis risk prediction and early intervention,” in Proc. 2024 Int. Conf. Multimedia Comput., Netw. Appl. (MCNA), 2024, pp. 1–6.

[2] Z. Si, et al., “PrOsteoporosis: predicting osteoporosis risk using NHANES data and machine learning approach,” BMC Med. Inform. Decis. Mak., vol. 18, no. 1, pp. 1–10, 2025.

[3] M. J. Almohaimeed, “Enhancing prediction of osteoporosis using supervised and unsupervised learning: new approach to disease subtyping,” Int. J. Inf. Manage., vol. 17, no. 2, pp. 31–47, 2025.

[4] B. Bouvard, C. Annweiler, and E. J. Legrand, “Osteoporosis in older adults,” Bone, vol. 88, no. 3, p. 105135, 2021.

[5] W.-Y. Ou Yang, et al., “Development of machine learning models for prediction of osteoporosis from clinical health examination data,” Int. J. Environ. Res. Public Health, vol. 18, no. 14, p. 7635, 2021.

[6] S. Tabib, et al., “Diagnosis osteoporosis risk: using machine learning algorithms among Fasa adults cohort study (FACS),” PLOS Digit. Health, vol. 8, no. 1, p. e70023, 2025.

[7] G. S. O. O., “Epileptic detection from EEG recordings based on machine learning techniques,” unpublished, 2023.

[8] O. Johnell and J. A. Kanis, “An estimate of the worldwide prevalence and disability associated with osteoporotic fractures,” Osteoporos. Int., vol. 17, pp. 1726–1733, 2006.

[9] J. A. Kanis, et al., “FRAX™ and the assessment of fracture probability in men and women from the UK,” Osteoporos. Int., vol. 19, pp. 385–397, 2008.

[10] G. S. Ohannesian and E. J. Harfash, “Epileptic seizures detection from EEG recordings based on a hybrid system of Gaussian mixture model and random forest classifier,” Inform. Med. Unlocked, vol. 46, no. 6, 2022.

[11] Y.-T. Lin, et al., “Can machine learning predict pharmacotherapy outcomes? An application study in osteoporosis,” Comput. Biol. Med., vol. 225, p. 107028, 2022.

[12] H. M. Bui, et al., “Predicting the risk of osteoporosis in older Vietnamese women using machine learning approaches,” Sci. Rep., vol. 12, no. 1, p. 20160, 2022.

[13] L. Ji, et al., “Osteoporosis, fracture and survival: application of machine learning in breast cancer prediction models,” Front. Oncol., vol. 12, p. 973307, 2022.

[14] X. Wu and S. J. Park, “A prediction model for osteoporosis risk using a machine-learning approach and its validation in a large cohort,” J. Korean Med. Sci., vol. 38, no. 21, 2023.

[15] C. Lee, et al., “Prediction of osteoporosis in patients with rheumatoid arthritis using machine learning,” Sci. Rep., vol. 13, no. 1, p. 21800, 2023.

[16] W.-Y. Ou Yang, et al., “Development of machine learning models for prediction of osteoporosis from clinical health examination data,” Int. J. Environ. Res. Public Health, vol. 18, no. 14, p. 7635, 2021.

[17] J.-B. Tu, et al., “Using machine learning techniques to predict the risk of osteoporosis based on nationwide chronic disease data,” Front. Public Health, vol. 14, no. 1, p. 5245, 2024.

[18] C. Qiu, et al., “Developing and comparing deep learning and machine learning algorithms for osteoporosis risk prediction,” Front. Endocrinol., vol. 7, p. 1355287, 2024.

[19] W.-C. Huang, et al., “A simple and user-friendly machine learning model to detect osteoporosis in health examination populations in Southern Taiwan,” Arch. Osteoporos., vol. 24, p. 101826, 2025.

[20] Y. Peng, C. Zhang, and B. Zhou, “A cross-sectional study comparing machine learning and logistic regression techniques for predicting osteoporosis in a group at high risk of cardiovascular disease among old adults,” BMC Geriatr., vol. 25, no. 1, pp. 1–15, 2025.

[21] A. Cabrera, et al., “Application of machine learning algorithms to predict 30-day hospital readmission following cement augmentation for osteoporotic vertebral compression fractures,” J. Clin. Neurosci., vol. 23, p. 100338, 2024.

[22] Y. Liu, et al., “Body mass index and the risk of all-cause and site-specific fractures: a systematic review and dose-response meta-analysis of cohort studies,” Bone, vol. 162, p. 116441, 2022.