Evaluation of a 3D digital photographic imaging system of the human face.

The aim of this study was to investigate the accuracy of a 3D photographic imaging system for potential application in fabrication of maxillofacial prostheses. For validity estimates, computer digitized 3D photographic images of calibrated ruler attached to a model were analysed for linear and curvilinear distances in 1 mm increments (1-50 mm from the centre of the image). Distortion was evaluated from 1 mm distances measured 20, 40 and 60 mm from the centre of the field. To estimate reliability of measurements in vivo, two raters measured the right endocanthion-subnasale distance and the distance between right and left endocathion on images of 10 subjects, repeated three times (30 images total). From the digital file for one subject, a rapid prototyping (RP) machine produced a 3D model of the mid-face. Measures from the model were compared with those of the 3D computer image to estimate error in fabrication. Mean error values for calibrated distances ranged from 0.07 to 0.26 mm for linear distances, 0.08-0.34 mm for curvilinear distances, and 0.06-0.12 mm for distortion from the centre. Intra- and interexaminer correlation ranged from 0.92 to 1.00 and 0.94-1.00 respectively. Measures of the endocanthion and subnasale distances on the RP fabricated mid-face model were within 8% of the corresponding measures on 3D computer images. The accuracy of the photographic 3D imaging system tested was sufficient for clinical description of the mid-face structures and may be potentially useful for rapid prototyping of facial prostheses.