Aim of the study was to verify the accuracy of rasterstereography (RST), as radiation-free alternative to plain radiography (RAD) in the monitoring of spine deformity and scoliosis progression in juvenile and adolescent subjects with idiopathic scoliosis. 192 subjects underwent RST (by Formetric 4D device) and low-dose RAD (EOS Imaging, France) in the same session. A sub-group of 30 subjects, selected for conservative treatment with corrective bracing, was assessed at 6-months follow-up. The Cobb angles (CA) obtained by the 3D spine reconstruction from RAD were compared with those provided by RST. Thoracic kyphosis (TK) and lumbar lordosis (LL) were compared as well. RST provided lower CA compared to RAD (15° vs. 33°, mean values). The average difference in measuring CA was 18°, and the correlation coefficient was 0.55. Comparable TK was observed, whereas LL resulted underestimated by RST compared to RAD (34° vs. 43°, average values). The within-subjects correlation, measuring the accuracy of RST in monitoring the scoliosis progression, was 0.3. Accuracy of RST in identifying increased or decreased CA was 67%. Sensitivity and specificity were 64% and 69%. RST demonstrated moderate accuracy in measuring the scoliosis degree and low accuracy in monitoring the curve progression. Accordingly, it cannot be considered as a valid alternative to radiographic evaluation. However, since demonstrated capable of revealing the presence of spine deformity, it could be in principle considered for the early screening in large adolescent populations, but after accounting for a cost-benefit analysis with respect to other traditional approaches. These slides can be retrieved under Electronic Supplementary Material.
Read full abstract