Automated Measurements of Spinal Parameters for Scoliosis Using Deep Learning

• DOI: 10.1097/brs.0000000000005280
• Publication Date: 2025-03-28T15:09:49Z
• Authors: Xianghong Meng; Shan Zhu; Qilong Yang; Fengling Zhu; Zhi Wang; Xiaoming Liu; Pei Dong; Shuaikun Wang; Lianxi Fan

Study Design. Retrospective single-institution study.

Objective. To develop and validate an automated convolutional neural network (CNN) to measure the Cobb angle, T1 tilt angle, coronal balance, clavicular angle, height of the shoulders, T5–T12 Cobb angle, and sagittal balance for accurate scoliosis diagnosis.

Summary of Background Data. Scoliosis, characterized by a Cobb angle >10°, requires accurate and reliable measurements to guide treatment. Traditional manual measurements are time-consuming and have low interobserver and intraobserver reliability. While some automated tools exist, they often require manual intervention and focus primarily on the Cobb angle.

Materials and Methods. In this study, we utilized four data sets comprising the anterior-posterior (AP) and lateral radiographs of 1682 patients with scoliosis. The CNN includes coarse segmentation, landmark localization, and fine segmentation. The measurements were evaluated using the dice coefficient, mean absolute error (MAE), and percentage of correct key-points (PCK) with a 3-mm threshold. An internal testing set, including 87 adolescent (7–16 yr) and 26 older adult patients (≥60 yr), was used to evaluate the agreement between automated and manual measurements.

Results. The automated measures by the CNN achieved high mean dice coefficients (>0.90), PCK of 89.7%–93.7%, and MAE for vertebral corners of 2.87–3.62 mm on AP radiographs. Agreement on the internal testing set for manual measurements was acceptable, with an MAE of 0.26 mm or degree-0.51 mm or degree for the adolescent subgroup and 0.29 mm or degree-4.93 mm or degree for the older adult subgroup on AP radiographs. The MAE for the T5–T12 Cobb angle and sagittal balance, on lateral radiographs, was 1.03° and 0.84 mm, respectively, in adolescents, and 4.60° and 9.41 mm, respectively, in older adults. Automated measurement time was significantly shorter compared with manual measurements.

Conclusion. The deep learning automated system provides rapid, accurate, and reliable measurements for scoliosis diagnosis, which could improve clinical workflow efficiency and guide scoliosis treatment.

Level of Evidence. Level III.