Abstract: Objective To understand the effect of cone-beam CT(CBCT) voxel overlap at different stages on the accuracy of quantitative analysis during longitudinal efficacy assessment of two-dimensional images. Methods Thirty-six adult patients with bimaxillary protrusion who had finished extraction orthodontic treatment were selected, and CBCT data before and after treatment were imported into Dolphin software. The experimental group underwent voxel overlap based on the anterior cranial base, then were reconstructed respectively in layers to form the combined lateral film. In the control group they were directly converted to lateral films after cranial repositioning. 16 indexes were selected for measurement and one senior orthodontist completed all the measurements. All the data were measured twice. Differences of the indexes before and after treatment were calculated respectively. Paired t-test was used for inter-group analysis. Results Voxel overlap helped improve the fitting degree of skull base and the marker points in the experimental group were easier to identify. Inter-group comparison of the differences showed that there were significant differences in SNA, U1-NA, U1-Y, U6-Y, L1-Y and L6-Y (P<0.05). Conclusion CBCT voxel overlap can reduce errors and improve the accuracy of longitudinal studies.

Key words: voxel overlap, reconstruction in layers, error, longitudinal study