The aim of this research was to evaluate the reliability and validity of measurements obtained from clinical standardized occlusal photographs compared with dental cast measurements. This study comprised a consecutive sample of 16 patients (eight males and eight females, aged 15–24 years) in the permanent dentition without agenesis and/or tooth loss. A paired t-test and intraclass correlation coefficient (ICC) were used to examine validity and reliability. Both statistics were applied for intra- and inter-methods error analysis, at P < 0.05.
Random error for the photogrammetric method (less than 0.48 mm) was similar to that for the dental cast measurements (less than 0.43 mm). ICC revealed excellent reliability for both methods (P < 0.01) and no significant difference for any variables, with the exception of upper inter-canine width obtained on the dental casts (P = 0.0038) and photogrammetry (P = 0.01). However, differences were less than 1 per cent of the mean inter-canine width. Inter-method analysis showed a significant correlation for all variables (P < 0.001), with good to excellent reliability (r = 0.66–0.93). A significant mean inter-method difference was consistently observed for the upper first molars (0.33 mm, P < 0.01). For the remaining teeth (left to right second premolars), the largest mean difference was approximately equal to the resolution of the human eye (0.2 mm or less). Minor differences (around 2 per cent of the mean) and an excellent ICC (0.75–0.93, P < 0.01) were observed for arch dimensions.
Except for the mesio-distal width of the upper first molars, the recently developed photogrammetric method showed accuracy, validity, and reliability acceptable for clinical and scientific purposes.