Purpose: To compare, in vitro, the trueness of three photogrammetry-based systems-intraoral (IPG), extraoral (EPG), and navigation photogrammetry (NPG)-for complete-arch digital implant impressions.

Materials and methods: An actual edentulous maxillary dental cast with four multi-unit abutment analogs was digitized with a desktop scanner to obtain reference dataset. The cast was scanned using IPG, EPG, and NPG (10 scans each). Test and reference standard tessellation language (STL) files were superimposed using a best-fit algorithm. Linear (ΔX, ΔY, ΔZ), angular (ΔANGLE), three-dimensional deviations (ΔEUC, global ΔRMS) were calculated. Data normality was assessed with Shapiro-Wilk test. Median regression models evaluated effects of photogrammetry system, implant position (anterior/posterior), and their interaction. Effective scanning time was recorded.

Results: Minimal differences were observed among systems (ΔRMS: EPG 18.10 µm, IPG 20 µm, NPG 20.70 µm; ΔEUC: IPG 34.35 µm, EPG 44.05 µm, NPG 46.80 µm; ΔAngle: IPG 0.43°, EPG 0.45°, NPG 0.46°). Scanning time was similar for EPG and NPG (8, 8.5 s), while with IPG was significantly longer (46.5 s). Median regression showed significant effects of photogrammetry system and implant position on ΔX and ΔY (P <0.01), with significant interaction terms for ΔY and ΔZ (P <0.001).

Conclusion: Intraoral, extraoral, and navigation photogrammetry demonstrated high accuracy suitable for complete-arch digital impressions. Differences among systems were small, and their clinical implications require further validation under clinical conditions. The IPG system showed slightly lower global linear deviations, whereas EPG and NPG provided significantly shorter implant-registration times. Both photogrammetry system and implant position significantly influenced linear deviation patterns, particularly along the horizontal axes, demonstrating system-dependent behavior between anterior and posterior sites.