Related Experiment Video
Updated: Sep 27, 2025
12:04
Assessment of Bone Fracture Healing Using Micro-Computed Tomography
Published on: December 9, 2022
2.0K
Machine Learning Algorithms: Prediction and Feature Selection for Clinical Refracture after Surgically Treated
Hirokazu Shimizu1,2, Ken Enda2, Tomohiro Shimizu1
1Department of Orthopaedic Surgery, Faculty of Medicine and Graduate School of Medicine, Hokkaido University, Sapporo 060-8638, Japan.
Journal of Clinical Medicine
|April 12, 2022
Summary
Predicting refracture in fragility fracture patients is crucial. Rheumatoid arthritis and chronic kidney disease are identified as key predictors, aiding in better patient management.
Area of Science:
- Orthopedics
- Gerontology
- Data Science
Background:
- Increasing incidence of fragility fractures globally.
- Refracture rates are rising, yet predictors remain unclear.
- Over 7000 surgically treated fragility fracture patients analyzed.
Purpose of the Study:
- Identify predictors of clinical refracture.
- Develop machine learning models for refracture prediction.
- Enhance management strategies for fragility fracture patients.
Main Methods:
- Utilized a registry-based longitudinal dataset.
- Developed automatic prediction models using machine learning.
- Employed a decision-tree-based model (LightGBM) for analysis.
Main Results:
- Rheumatoid arthritis (RA) identified as a significant predictor.
- Chronic kidney disease (CKD) identified as a significant predictor.
- Both RA and CKD are associated with secondary osteoporosis.
Conclusions:
- LightGBM model accurately predicts clinical refracture.
- RA and CKD are key clinical predictors of refracture.
- Understanding these predictors can optimize patient care.
