Predictive Model Development to Identify Failed Healing in Patients after Non–Union Fracture Surgery

2024-04-17

Annual International Conference of the IEEE Engineering in Medicine and Biology Society

AI Summary

This study addresses the challenge of predicting nonunion risk following initial revision surgery in patients with long-bone nonunions. Leveraging the TRUFFLE clinical dataset (n = 797), we developed and rigorously evaluated multiple machine learning models—marking the first validation of XGBoost’s predictive performance for nonunion outcomes in a small-sample orthopedic cohort. Through clinically informed feature engineering, we trained binary classifiers to distinguish union from nonunion. The XGBoost model achieved 70% sensitivity and 66% specificity—substantially outperforming SVM (49% sensitivity) and logistic regression (43% sensitivity)—thereby overcoming the sensitivity limitations of conventional statistical methods in low-event-rate settings. Importantly, the model provides interpretable, high-sensitivity risk stratification, enabling early identification of high-risk patients and timely, personalized intervention. These findings advance the clinical translation of precision prognostic tools in orthopedic trauma care.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Bone non-union is among the most severe complications associated with trauma surgery, occurring in 10–30 % of cases after long bone fractures. Treating non-unions requires a high level of surgical expertise and often involves multiple revision surgeries, sometimes even leading to amputation. Thus, more accurate prognosis is crucial for patient well-being.Recent advances in machine learning (ML) hold promise for developing models to predict non-union healing, even when working with smaller datasets, a commonly encountered challenge in clinical domains. To demonstrate the effectiveness of ML in identifying candidates at risk of failed non-union healing, we applied three ML models—logistic regression, support vector machine, and XGBoost—to the clinical dataset TRUFFLE, which includes 797 patients with long bone non-union.The models provided prediction results with 70% sensitivity, and the specificities of 66 % (XGBoost), 49 % (support vector machine), and 43 % (logistic regression). These findings offer valuable clinical insights because they enable early identification of patients at risk of failed non-union healing after the initial surgical revision treatment protocol.

Problem

• Fracture Surgery
• Wound Healing
• Predictive Model

Innovation

• Machine Learning
• XGBoost
• Fracture Healing Prediction