Accuracy of intraoral scanners in neonates cleft anatomy

Aim: This study investigated the accuracy of intraoral scanner (IOS) based on different image acquisition technologies in the field of presurgical-orthopedictreatment (PSOT) in neonates with cleft.
Methods: Dental cast models of clinical situations representing unilateral cleft-lip-palate(UCLP), bilateral cleft-lippalate( BCLP) and cleft-palate(CP) with reference PEEK-scanbodies (Cares RC Mono-Scankörper, Straumann, Switzerland) were scanned utilizing four IOS systems: CareStream-CS3600®(CS), Medit-i500®(MD), Cerec-Omnicam®(SO), 3Shape-Trios-3®(TS). One calibrated operator made 5 scans from each model using each IOS (N=60). Reference digital impressions were obtained by an industrialgrade laboratory scanner (Sirona inEos-X5) and superimposed using best fit algorithm. The divergence measure was extracted and the scanners were compared in view of their accuracy using generalized least squares statistical models that account for variance heterogeneity. Additionally, comparative 3D analysis of scans was performed using the reverse engineering software (Geomagic-ControlX) in order to measure the discrepancy between intraoral scans and the reference scan in different anatomic regions of interest: alveolar-crest(AC), cleft(CL), palate(PL), vestibulum(VS), premaxilla(PM).
Results: The four IOS showed relevant and significant differences in estimated trueness (P<0.001) and precision (P=0.009). Among all anatomical models and analysed area of interest TS had the best accuracy (trueness: -1.57μm; precision: 9.41μm), followed by MD (trueness: - 20.63μm; precision: 29.18μm), CS (trueness: -40.43μm; precision: 16.52μm) and SO (trueness: 81.27μm; precision: 40.32μm).
Conclusions: Impression of the maxilla in cleft lip and palate patients is challenging for the operator. Relevant and significant differences in trueness and precision were found between the four IOS. TS showed the best accuracy and was least influenced IOS under different anatomical situations.