Assessment of 3D facial scan integration in 3D digital workflows using radiographic markers and the iterative closest point algorithm

PURPOSE: Integration of three-dimensional (3D) facial scanning into digital workflows has become increasingly important for enhancing treatment planning and esthetic evaluation. However, limited data exists on the accuracy of various methods of merging facial scans with cone beam computed tomography (CBCT) scans. The purpose of this study is to compare the accuracy of integrating 3D facial scans with CBCT scans using the iterative closest point (ICP) algorithm versus a radiopaque (RO) marker technique. MATERIALS AND METHODS: This prospective clinical study included 15 patients who received CBCT scans and 3D facial scans in repose and smile positions. The Digital Imaging and Communications in Medicine (DICOM) datasets from the CBCT scans containing RO markers were integrated with the standard tessellation language (STL) files from the facial scans in Exocad software using two methods: RO alignment and the ICP algorithm. The datasets from both groups were statistically compared using a paired t-test (α = 0.05). RESULTS: The means for the six subsets merged by the ICP algorithm ranged from 1.47 to 2.0 mm, and the means for the RO markers were 0.13 to 0.15 mm. The novel RO markers method was statistically significant and more accurate than the ICP algorithm (p < 0.001). CONCLUSIONS: Merging 3D facial scans to CBCT using radiopaque markers demonstrated higher trueness compared to the ICP algorithm under the conditions tested. This technique may serve as a reliable alternative for improving integration accuracy in digital dental workflows, particularly where precise facial landmark preservation is essential.