Three-dimensional reconstruction of ultrasound imaging in dentistry: a systematic review

BACKGROUND: Three-dimensional ultrasound (3D-US) reconstruction has emerged as a potential solution to the limited field of view and operator dependence of two-dimensional intraoral ultrasonography. Multiple acquisition, tracking, and reconstruction strategies have been proposed, yet no prior systematic review has synthesized the methods used specifically within dentistry. OBJECTIVE: To identify and compare 3D-US reconstruction techniques applied in dentistry and evaluate how acquisition strategy, spatial tracking, and reconstruction methodology influence accuracy, reproducibility, feasibility, and clinical applicability. METHODS: A systematic search of 5 databases was performed. Studies were included in the review if they used ultrasound imaging with any 3D-US reconstruction method for a defined dental application. Risk of bias was assessed using a customized hybrid QUADAS-2/QUIN/STARD framework. RESULTS: Thirteen studies (2010-2025) met the inclusion criteria; 10 were in vitro and 3 in vivo. Acquisition methods included mechanized linear-stage scanning (n = 6), robotic arm control (n = 2), tracked freehand (n = 2), untracked freehand (n = 1), and custom intraoral jigs (n = 2). Accuracy ranged from 12 to 50 μm for digital impressions to sub-millimeter deviations for periodontal and ridge assessments, with most error rates below 10%. Reproducibility and clinical feasibility were both inversely related to hardware complexity: mechanized and robotic systems achieved the highest precision and enabled precise positioning for repeated evaluations; however, they had the lowest clinical usability. Conversely, freehand approaches demonstrated the strongest clinical potential but at the expense of reproducible positioning and highly precise measurements. Risk of bias was high or unclear in all studies due to heterogeneity, small sample sizes, and limited in vivo validation. CONCLUSIONS: Over 15 years, 3D-US reconstruction in dentistry has advanced substantially, demonstrating promising accuracy for digital impressions, periodontal assessment, ridge measurement, and implant planning. However, most studies remain feasibility-based and in vitro, and methodological variation limits comparability. Future progress will depend on standardization of validation frameworks, increased in vivo accuracy studies, and development of compact, automated acquisition workflows. With continued refinement, 3D-US has the potential to become a clinically practical, radiation-free imaging modality for diagnostic and procedural applications in dentistry.