Investigation of an Intraoral Scanner and Analysis Workflows for Measuring Surface Loss on Planar and Freeform Surfaces in vitro

Student thesis: Doctoral ThesisDoctor of Philosophy


Objective – The aim of this thesis was to determine the accuracy and measurement thresholds of an intraoral scanner, IOS-TD (True Definition, Midmark Corp., Ohio, USA) for measuring surface change on planar and freeform surfaces such as seen in polished and natural enamel.

Methods – The IOS-TD’s accuracy for measuring a groove on polished enamel was investigated using two handling techniques (handheld Vs jig-guided scanning) at different surface-to-camera distances (3, 4, 5, and 7 mm), and its titanium dioxide scanning powder was characterised. Thereafter, the IOS-TD was tested for measuring Sq roughness on flat textured surfaces (Sq range: 1.2 – 269.0 μm) and for measuring step heights and 𝑋𝑌 areas of grooves (0.0 - 86.5 μm depth range) on polished enamel, as well as surface skewness and kurtosis. Moreover, two techniques (surface-subtraction and surface-registration) were compared to the gold standard single-scan analysis for measuring enamel grooves (2 - 19 μm depth). Consequently, the errors of four bi-scan analyses were investigated, namely: best-fit surface-registration (BF-Reg), reference-based surface-registration (Ref-Reg), a combination of best-fit surface-registration and surface-subtraction (BF-Sub) and reference-based surface registration and surface subtraction (Ref-Sub), using freeform softgauges with craters of known depths. Using the Ref-Sub analysis thereafter, step height and 𝑋𝑌 areas of craters (11 – 81 μm depth) on natural enamel were measured by the IOS-TD. Finally, the effect of different scan sizes (cusp/tooth/sextant/quadrant/full-arch) on the accuracy of the IOS-TD for measuring different crater depths was investigated. Throughout the thesis, the IOS-TD measurements were compared to a gold standard non-contacting laser profilometer, NCLP (TaiCaan Technologies™, XYRIS 2000CL, UK).

Results – No statistically significant differences were observed between handheld and jig-guided scanning at any surface-to-camera distances 3-, 4-, 5-, and 7-mm (p=0.8946-0.9999). Scanning powder application increased surface roughness significantly (p<0.0001); however, the change in surface form was < 1 μm. The IOS-TD demonstrated significantly different roughness measurements than the NCLP (p<0.0001). It also demonstrated significant 𝑋𝑌 area measurement differences compared to NCLP below 44 μm (p<0.05). Enamel groove depths ≥ 44 μm were reliably measured by the IOS-TD whilst no significant differences were observed compared to NCLP for step heights and 𝑋𝑌 area measurements above this depth. The bi-scan surface subtraction and single-scan analyses were not statistically different; whilst the bi-scan surface-registration analysis resulted in significant overestimation (p=0.0001 – 0.0119) of groove depths. The methods of BF-Sub and Ref-Sub resulted in a significant reduction of errors, achieving 0.0 to -0.3% error, compared to -29.7 – -32.5% for the BF-Reg and -2.4 – -3.6 % for the Ref-Reg (p < 0.0001). Using the Ref-Reg, the IOS-TD measurements were significantly different than the NCLP for crater depths below 73 μm (p=0.0001-0.0005), whilst above this threshold the IOS-TD could reliably measure all craters. Finally, the highest accuracy (±5% error) was observed using scan sizes restricted to single teeth, even at the smallest crater depth of 83 μm. This decreased significantly (p<0.0001) as the scan size increased to a sextant, quadrant and full-arch. Using full-arch scanning, the accuracy plateaued to approximately 10-12% underestimation for crater depths ≥195μm.

Conclusions – This thesis demonstrated the accuracy, thresholds and limitations of the IOS-TD in combination with software analysis workflows for measuring change on planar and freeform surfaces. The limitations are linked to the scanner’s lower spatial resolution and propagation of image-stitching errors as well as errors arising during surface alignment. The findings showed potential for intraoral scanners to measure surface loss in the order of ≥ 80 μm when using datasets of single-teeth. However, the accuracy decreased using full arch scanning even up to depths ≤195μm. Further research is needed to optimise the resolution and accuracy of intraoral scanners and software workflows.

Date of Award1 Jun 2023
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorRupert Austin (Supervisor) & David Bartlett (Supervisor)

Cite this