BACKGROUND: Virtual dental casts have been recently introduced to orthodontics. The problem of capturing the shapes of teeth on study casts may be complicated by the presence of undercut areas and deep grooves. AIMS: This study aimed to develop a 3-D laser scanner and associated software, and to evaluate the reproducibility and validity of mesiodistal crown width measurements based on slice distinction. METHODS: To evaluate reproducibility: a cube was scanned with readily available equipment and the digital measurements compared to caliper measurements of the cube; and the mesiodistal widths of artificial upper teeth were measured, set-up in a dental model, scanned, measured on the virtual study models and compared with the reference measurements. Custom software was used to capture and process the resulting images. To determine validity artificial teeth were measured with calipers, set-up in 20 different malocclusions, duplicated and scanned. The caliper measurements of groups of teeth (incisors, canines, premolars, molars) were compared with digital measurements of the same teeth. In the second method to determine validity, 10 dental casts were scanned and measured by two examiners and the mesio-distal widths compared with the reference values, and with each other. RESULTS: The digital measurements of the cube fell within 0.1 mm of the reference value. The absolute error in repeated measures of the dental model was 0.32 +/- 0.25 mm. The overall correlation between 3-D images of the teeth and the reference values, using the intraclass correlation coefficient (ICC), was 0.84. The most and least valid results for groups of teeth were the premolars and canines, respectively. The inter-observer ICC was 0.60, and each examiner's ICC versus the reference values were 0.833 and 0.855. CONCLUSIONS: The mesiodistal widths of teeth and groups of teeth can be reliably and validly measured on virtual study models using the scanner and software developed for this purpose. Inaccuracy in the canine region may be overcome by using a smaller rotational angle during scanning.
BACKGROUND: Virtual dental casts have been recently introduced to orthodontics. The problem of capturing the shapes of teeth on study casts may be complicated by the presence of undercut areas and deep grooves. AIMS: This study aimed to develop a 3-D laser scanner and associated software, and to evaluate the reproducibility and validity of mesiodistal crown width measurements based on slice distinction. METHODS: To evaluate reproducibility: a cube was scanned with readily available equipment and the digital measurements compared to caliper measurements of the cube; and the mesiodistal widths of artificial upper teeth were measured, set-up in a dental model, scanned, measured on the virtual study models and compared with the reference measurements. Custom software was used to capture and process the resulting images. To determine validity artificial teeth were measured with calipers, set-up in 20 different malocclusions, duplicated and scanned. The caliper measurements of groups of teeth (incisors, canines, premolars, molars) were compared with digital measurements of the same teeth. In the second method to determine validity, 10 dental casts were scanned and measured by two examiners and the mesio-distal widths compared with the reference values, and with each other. RESULTS: The digital measurements of the cube fell within 0.1 mm of the reference value. The absolute error in repeated measures of the dental model was 0.32 +/- 0.25 mm. The overall correlation between 3-D images of the teeth and the reference values, using the intraclass correlation coefficient (ICC), was 0.84. The most and least valid results for groups of teeth were the premolars and canines, respectively. The inter-observer ICC was 0.60, and each examiner's ICC versus the reference values were 0.833 and 0.855. CONCLUSIONS: The mesiodistal widths of teeth and groups of teeth can be reliably and validly measured on virtual study models using the scanner and software developed for this purpose. Inaccuracy in the canine region may be overcome by using a smaller rotational angle during scanning.