The effect of magnetic resonance image slice thickness and voxel geometry on digital dental implant planning: an in vitro study

Abstract Objectives To evaluate the effect of magnetic resonance imaging (MRI) slice thickness and voxel geometry on digital dental implant planning compared to cone-beam computed tomography (CBCT). Methods Edentulous phantoms were scanned using CBCT and two MRI protocols: anisotropic and isotropic. Two calibrated examiners digitally planned the position of four implants for each sample ( n = 8) per exam (CBCT, anisotropic and isotropic). Diagnostic accuracy was assessed using implant angular deviation and linear measurements. Image quality was evaluated using a questionnaire regarding visualization of anatomical structures. Cost-to-planning-time ratio was assessed by exam costs per minute of planning. The Friedman and Wilcoxon tests were used to evaluate differences in diagnostic accuracy, planning time, and cost-to-planning-time ratio. Image quality ratings on a 5-point scale were classified as non-valid (≤ 2) or valid (≥ 3) for clinical use. Reliability was assessed by percentage agreement, and differences were evaluated using the chi-square test. Results A total of 96 implants were planned, with 32 in each exam. No significant differences were observed between exams for angular deviation ( p > 0.197), linear measurements ( p > 0.305), image quality ( p > 0.132), or planning time ( p = 0.093). However, CBCT had lower cost-to-planning-time ratio than anisotropic and isotropic MRI ( p = 0.012), which did not differ from each other ( p = 1.000). Inter-examiner agreement on image quality for anisotropic and isotropic MRI varied widely, with lower ratings for the mandibular nerve canal. Conclusion No statistically significant differences were detected among the imaging protocols in diagnostic accuracy or image quality under the investigated in vitro conditions.