JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Development and validation of machine learning models for predicting the risk of refracture after percutaneous kyphoplasty in OVCF patients

  • 0Second Affiliated Hospital of Soochow University, Suzhou, China.
European Spine Journal : Official Publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society +

|

September 15, 2025

|

September 15, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

Adjacent vertebral refracture after Percutaneous Kyphoplasty (PKP) is linked to bone mineral density, preoperative vertebral height, and osteoporosis treatment. A Balanced Bagging machine learning model accurately predicts refracture risk, aiding personalized patient care.

Area Of Science

  • Orthopedics
  • Spine Surgery
  • Medical Imaging

Background

  • Osteoporotic vertebral compression fractures (OVCF) are common.
  • Percutaneous Kyphoplasty (PKP) is a common treatment for OVCF.
  • Adjacent vertebral refracture is a significant postoperative complication.

Purpose Of The Study

  • Identify risk factors for adjacent vertebral refracture post-PKP.
  • Develop and validate a predictive model for refracture risk.
  • Evaluate the impact of clinical features on postoperative outcomes.

Main Methods

  • Retrospective analysis of 3,942 OVCF patients undergoing PKP (2018-2023).
  • Classification into non-refracture and refracture groups.
  • Logistic regression for risk factor identification.
  • Construction and validation of multiple machine learning models for prediction.

Main Results

  • 10.75% of patients experienced adjacent vertebral refracture.
  • Independent risk factors identified: bone mineral density (BMD), preoperative anterior vertebral height (AVH), anterior vertebral height restoration rate (AVHRR), and osteoporosis treatment.
  • Balanced Bagging model achieved high accuracy (96.58%), sensitivity (94.12%), specificity (96.88%), and F1 score (0.8556).

Conclusions

  • AVHRR, BMD, and osteoporosis treatment are associated with refracture risk after PKP.
  • The Balanced Bagging model is optimal for predicting postoperative refracture.
  • This predictive model can guide personalized postoperative management, improve outcomes, and reduce revision surgeries.

Keywords: