CT-based assessment of posterior ligamentous complex integrity in AO spine type A1–A2 thoracolumbar junction fractures under conditions of diagnostic uncertainty
Open Access
- 1 January 2026
- journal article
- Published by MedCrave Group Kft. in MOJ Applied Bionics and Biomechanics
- Vol. 10 (1) , 10-20
Abstract
Purpose: The diagnosis of posterior ligamentous complex (PLC) injury in thoracolumbar fractures remains a major challenge under conditions of diagnostic uncertainty, particularly when magnetic resonance imaging (MRI) is unavailable or delayed. This study aimed to develop and internally validate a quantitative CT-based model and a clinical nomogram for predicting PLC status in AO Spine type A1–A2 fractures using machine learning–guided morphometric analysis. Methods: CT and MRI data from 90 patients with thoracolumbar junction injuries were retrospectively analyzed, including 43 patients with AO Spine type A1–A2 fractures used for model development. Key morphometric parameters, including the acute interspinous expansion angle (AIEA), anterior-to-posterior height ratio (A/P), and bone density expressed in Hounsfield units (HU), were measured. An extreme gradient boosting (XGBoost) algorithm was used to estimate feature importance, which subsequently informed the development of an interpretable 10-point scoring scale and a logistic regression–based nomogram. Model performance was assessed using the area under the receiver operating characteristic curve (AUC) and calibration analysis with bootstrap correction. Results: The scoring scale identified AIEA and the A/P ratio as the most influential predictors, contributing 10.0 and 8.5 points, respectively. The score-based model demonstrated high discriminatory performance with an AUC of 0.944 (95% CI: 0.912–0.976), exceeding that of the baseline machine learning model (AUC = 0.836). Each additional point on the scale was associated with a 31% increase in the odds of PLC injury (odds ratio 1.31, p < 0.05). Calibration analysis showed good agreement between predicted and observed probabilities across the clinically relevant risk range (MAE = 0.041). Conclusion: The proposed CT-based prognostic model provides an objective and transparent tool for risk stratification in AO Spine type A1–A2 thoracolumbar fractures. By quantifying the probability of occult posterior instability, the nomogram supports evidence-based decision-making regarding the need for MRI and surgical stabilization in acute thoracolumbar trauma.Keywords
This publication has 46 references indexed in Scilit:
- Deep learning model for automated detection of fresh and old vertebral fractures on thoracolumbar CTEuropean Spine Journal, 2024
- Towards a standardized reporting of the impact of magnetic resonance imaging on the decision-making of thoracolumbar fractures without neurological deficit: Conceptual framework and proposed methodologyBrain and Spine, 2024
- Hounsfield units as predictor for cage subsidence and loss of reduction: following posterior-anterior stabilization in thoracolumbar spine fracturesEuropean Spine Journal, 2018
- CT for thoracic and lumbar spine fractures: Can CT findings accurately predict posterior ligament complex injury?European Spine Journal, 2018
- Reliability assessment of AOSpine thoracolumbar spine injury classification system and Thoracolumbar Injury Classification and Severity Score (TLICS) for thoracolumbar spine injuries: results of a multicentre studyEuropean Spine Journal, 2016
- Applied Predictive Modeling2013
- Guidelines for Reporting Reliability and Agreement Studies (GRRAS) were proposedJournal of Clinical Epidemiology, 2011
- Another look at measures of forecast accuracyInternational Journal of Forecasting, 2006
- Presentation of multivariate data for clinical use: The Framingham Study risk score functionsStatistics in Medicine, 2004
- Planning a reproducibility study: how many subjects and how many replicates per subject for an expected width of the 95 per cent confidence interval of the intraclass correlation coefficientStatistics in Medicine, 2001