An explainable machine learning model for predicting osteoporotic fragility fractures: a retrospective study in South China - Takeaways - MDSpire

An explainable machine learning model for predicting osteoporotic fragility fractures: a retrospective study in South China

  • By

  • Zebing Si

  • Konghe Hu

  • Huajun Wang

  • Xiaofei Zheng

  • June 19, 2026

  • 0 min

Share

  • 1

    The study analyzed 1125 hospitalized patients with osteoporotic fractures at Yuebei People's Hospital from 2019 to 2025.

  • 2

    A total of 73 clinical, laboratory, and radiological features were collected and filtered using Recursive Feature Elimination.

  • 3

    LightGBM outperformed other machine learning algorithms with a training-set ROC-AUC of 0.903 and an independent test-set ROC-AUC of 0.840.

  • 4

    SHAP analysis identified age, Hepatitis C virus IgG antibody, and serum sodium as key factors influencing fracture risk.

  • 5

    The model aims to enhance prediction of osteoporotic fragility fractures, facilitating earlier intervention for at-risk patients.

Original Source(s)

An explainable machine learning model for predicting osteoporotic fragility fractures: a retrospective study in South China

  • by Zebing Si, Konghe Hu, Huajun Wang, Xiaofei Zheng

  • June 19, 2026

Related Content

MiR-1a-3p/Fcgr4-dependent osteoclast activation regulates pathological bone loss

Body composition-derived principal components partially explain sex and age effects on bone mineral density in type 2 diabetes mellitus

Comparative efficacy and safety of odanacatib, abaloparatide, denosumab, teriparatide, and bisphosphonates for male osteoporosis: a systematic review and network meta-analysis