Validating rasterstereography to evaluate thoracic kyphosis in patients with Scheuermann's disease.

Purpose: Evaluate the validity of rasterstereography compared to low-dose, biplanar spine radiography for assessing thoracic kyphosis (TK) angles in Scheuermann's disease patients.

Methods: This prospective study included all the Scheuermann's disease patients consulting our clinic from 2016 to 2018. Recruited patients underwent two-dimensional low-dose biplanar anteroposterior full-length spine radiography and rasterstereography on the same day. Relationships between the TK angles measured were evaluated using Pearson correlation coefficients. Agreement between radiographic and rasterstereographic TK angles was evaluated using two-way intraclass correlation coefficients (ICCs) and Bland-Altman plots. Proportional biases were assessed using the slope regression lines fitted to Bland-Altman plots.

Results: The mean demographic and radiological characteristics of the 52 patients (20 girls; 39%) included were: age 13.1 ± 2 years; BMI 17.3 ± 2.8; and TK max. 50.4° ± 10°. Rasterstereographic TK angles were strongly correlated with radiographic TK angles evaluated from T2-T12 (r = 0.677) and from C7-Max (r = 0.704), with 'good' agreement (ICC > 0.75). A proportional bias was revealed in the slope regression line fitted to the Bland-Altman plot from the C7-Max radiography and the rasterstereography measurements (p = 0.034) but not from the T2-T12 and rasterstereographic TK angles (p = 0.997).

Conclusion: Rasterstereography is a reliable means of quantifying TK angles in Scheuermann's disease patients. It could directly reduce the number of radiographic scans patients need over time, minimising their radiation exposure.