AI-Powered Intraoperative Navigation Photogrammetry for Complete-Arch Implant Impression and Immediate Loading With a 3D-Printed Temporary Prosthesis: A Prospective Clinical Study.

OBJECTIVE

Despite the widespread clinical use of digital scanning, its accuracy for complete-arch implant impressions remains controversial. This study aimed to investigate the feasibility and clinical performance of an artificial intelligence (AI)-guided navigation photogrammetry system for intraoperative complete-arch impression and immediate loading.

MATERIALS AND

METHODS

Patients requiring complete-arch rehabilitation in either jaw were treated between July and December 2024. After navigation surgery, photogrammetry scanning was performed, and AI automatically compiled patient data, including implant coordinates, preoperative anatomy, digital wax-up, all referenced within the same X, Y, Z system. A monolithic three-dimensional (3D) printed screw-retained temporary prosthesis was delivered. Prosthetic fit was clinically and radiographically assessed and the extent of occlusal adjustment was measured. Scanning time, patient-reported outcomes (comfort, esthetics), and any biological and mechanical complications at 6 months were reported with descriptive statistics. To investigate whether the number of implants (4 vs. 6) and type of arch (maxilla vs. mandible) significantly affected impression-taking time, independent two-tailed t-tests were conducted (p < 0.05).

RESULTS

Nineteen patients were treated with 24 complete-arch prostheses (maxillary n = 14, mandibular n = 10) supported by 6 (n = 10) or 4 (n = 14) implants. The scanning time was 10.21 ± 1.02 and 27.73 ± 9.29 s for four and six implants, respectively. Type of arch did not affect scanning time (p > 0.05). A high level of esthetic satisfaction (8.91 ± 0.93) and comfort (8.08 ± 0.88) were reported. All prostheses were deemed to have passive fit, with no detectable marginal or radiographic gaps. Mean occlusal adjustment was 692.17 ± 247.43 μm, classified as moderate. No biological and mechanical complications were reported.

CONCLUSIONS

Within the limitations of this study, navigation photogrammetry and its AI-driven digital workflow were reliable for quick and efficient intraoperative complete-arch implant impression in both jaws. The implant number was the only variable to significantly influence scanning time. All 3D-printed screw-retained temporary prostheses were accurately fitted with moderate occlusal adjustments and immediately loaded. This approach eliminates chairside conversion or relining, potentially reducing treatment time. High patient satisfaction was reported.