Assessment of the AI-assisted Registration Versus Conventional Point-based Registration on CBCTs With Heavy Metal Artifacts
Assessment of the AI-assisted Registration Versus Conventional Point-based Registration on CBCTs With Heavy Metal Artifacts
Our study investigates the accuracy and duration needed for 3D model registration using artifical intelligence (AI) assistance compared to conventional point-based registration. Manual segmentation of all cone beam computed tomography (CBCT) scans will be performed before the registration procedure.
CBCT images and intraoral scans will be screened following specific eligibility criteria. 16 CBCT images and intraoral scans that will meet the inclusion criteria will undergo manual segmentation via 3D medical image processing software. Afterward, point-based registration and AI-assisted registration will be performed by a single operator using specialized implant planning software. Then, the registration accuracy will be examined by measuring the distances between the three-dimensional models of CBCT data and intraoral scans. Also, the duration required for registration will be calibrated and recorded by a stopwatch.
Inclusion Criteria:
CBCT scans for either the maxilla or mandible or both and intraoral scans or manual impressions with metal restorations.
Exclusion Criteria:
Scans without metal restorations.