Related Experiment Video
Updated: Jul 13, 2026
04:19
Minimally Invasive Treatment for Thoracolumbar Burst Fracture Using Sagittal Alignment Screws and A Trauma Reduction Device
Published on: November 8, 2024
Predicting proximal junctional failure in adult spinal deformity patients using machine learning models based on
Akihiko Hiyama1, Daisuke Sakai2, Hiroyuki Katoh2
1Department of Orthopaedic Surgery, Surgical Science, Tokai University School of Medicine, 143 Shimokasuya, Isehara, Kanagawa, 259-1193, Japan. a.hiyama@tokai.ac.jp.
Scientific Reports
|November 20, 2025
Summary
Machine learning models can predict proximal junctional failure (PJF) after adult spinal deformity (ASD) surgery. The Random Forest model showed the highest accuracy in identifying patients at risk for this complication.
Area of Science:
- Spine surgery
- Biomechanical engineering
- Machine learning in medicine
Background:
- Proximal junctional failure (PJF) is a common complication after adult spinal deformity (ASD) surgery.
- Early identification of patients at high risk for PJF is crucial but challenging due to complex risk factors.
Purpose of the Study:
- To evaluate the predictive performance of five machine learning (ML) models for identifying PJF risk.
- To assess the utility of preoperative and postoperative spinal alignment parameters in PJF prediction.
Main Methods:
- Retrospective analysis of 92 ASD patients undergoing two-stage corrective surgery with lateral lumbar interbody fusion (LLIF).
- Radiographic parameters were measured preoperatively and postoperatively.
- Six alignment features were selected, and five ML models (Random Forest, Logistic Regression, SVM, Decision Tree, Naive Bayes) were trained and tested.
Main Results:
- Random Forest model achieved the highest accuracy (78.4%) and AUC (0.704).
- Predicted PJF probabilities were significantly higher in the PJF group compared to the non-PJF group (p=0.0057).
- Cross-validation demonstrated model robustness (fivefold: 79.4%, tenfold: 77.3%).
Conclusions:
- The Random Forest model shows promise as a reliable tool for stratifying PJF risk.
- Preoperative and early postoperative spinal alignment parameters are valuable predictors of PJF.
- Future research should include bone mineral density and comorbidities for enhanced clinical applicability.
