Accuracy and precision of computer tomographic scanning of dental models

Objectives: To ascertain the accuracy and precision of Computer Tomography (CT) scanning of dental models for rapid and mass digital archival. Methods: In the accuracy experiment, a maxillary arch plaster model was subjected to one round of scanning with the Siemens Somatom CT scanner (Siemens AG, Germany) and separately with the Minolta VIVID900 surface laser scanner (Konica-Minolta, Japan). Measurements on the CT scanned image were made using the Osirix viewer (Osirix Foundation, Switzerland). Identical measurements of the laser scanned image were made on the software RAPIDFORM, version 2006 (INUS Technology, South Korea). The direct measurements of the physical plaster model were made using a Coordinate Measuring Machine (Brown & Sharpe, Hexagon Metrology, USA). Two variables, the mesio-distal width of tooth 26, and the buccal cuspal tip-gingival height of tooth 14, were measured. Each measurement was repeated consecutively 10 times by one investigator for each image type and on the physical plaster model. The Precision experiment compared 10 repeated measurements, twice taken 3 days apart, by the same investigator of the inter-canine width on the same plaster model images. Bland Altman test with a 95% confidence interval was used to compare the accuracy and precision of measurements. Results: Mean accuracy error for the CT image was 0.23mm (95% CI 0.07 to 0.39) and 0.63mm (95% CI 0.46 to 0.80) for the buccal cusp tip to gingival height and mesio-distal width measurements, respectively. Mean accuracy error for surface laser scanning was 0.04mm (95% CI -0.09 to 0.16) and 0.06mm (95% CI -0.08 to 0.20), respectively. The mean precision error for both CT and surface laser scanned images was 0.04mm (95%CI -0.14 to 0.23) and 0.11mm (95% CI -0.09 to 0.30), respectively. Conclusion: Accurate and consistent linear measurements can be obtained from CT scanned images of the dental plaster model.