Related Experiment Video
Updated: Jun 18, 2026
Evaluation of Patients' Posture and Gait Profile After Lumbar Fusion Surgery by Video Rasterstereography and Treadmill Gait Analysis
Published on: March 23, 2019
A Decision Tree Algorithm to Predict Postoperative Kyphosis After Cervical Laminoplasty for Cervical Spondylotic
Karma Barot1, Miguel A Ruiz-Cardozo1, Karan Joseph1
1Department of Neurological Surgery, Washington University School of Medicine, Saint Louis, Missouri, USA.
Background And Objectives:
Approximately 15% of patients develop kyphotic deformity after cervical laminoplasty. Understanding preoperative risk factors of postlaminoplasty kyphosis (PLK) could allow surgeons to tailor their treatment according to patient's risk. The objective of this study was to develop a decision-tree algorithm to evaluate the risk of developing PLK 1-year postlaminoplasty.
Methods:
This retrospective study included cervical spondylotic myelopathy patients who underwent laminoplasty with a complete 1-year clinical follow-up. Radiographic measurements included T1 slope, C2-7 Cobb angle, C2-7 sagittal vertical axis, McGregor's slope, occiput to C2 angle (O-C2 angle), neck tilt, and C2-C3 disk angle. Development of PLK was defined as a loss of cervical lordosis of >10° in the preoperative and postoperative C2-7 Cobb angle 1 year after surgery. Regression analyses assessed the predictive performance of radiographic measurements for PLK. Radiographic measurements, age, sex, and body mass index were used to build a decision tree model to predict PLK. The leave-one-out cross-validation procedure was used to validate the model.
Results:
Seventy-six patients (54 men, 71%) met the inclusion criteria, with 9 developing PLK. The cohort had a mean age of 59 years and a body mass index of 29.0 kg/m2. When controlling for age, body mass index, and sex, C2-7 Cobb angle (P = .028, odds ratio 0.837), O-C2 angle (P = .005, odds ratio 1.23), and C2-C3 disk angle (P = .01, odds ratio 1.16) were significant predictors for PLK. A decision tree model calculated the importance of the features and identified that an O-C2 angle (<27°) and C2-7 Cobb (>9°) offered the best classification performance (area under the curve: 0.76, CI 0.61-0.92, sensitivity of 78% and specificity of 80%).
Conclusion:
Radiographical predictors identified the risk of PLK with high metric performance. The O-C2 angle, and C2-C7 angle may be pivotal radiographic indicators to evaluate the preoperative risk of PLK.
