Imaging of facial soft tissues in multislice computerized tomography: a new geometric method of analysis and its...

The main purpose of this preliminary study was to propose an original analysis of facial soft tissues' depth to detect their topographic variability, which causes are intrinsic (depth) and extrinsic (bone shape and reciprocal correlations with soft tissues themselves). To check the utility and accuracy of the method in recognizing measures' differences, other aims were comparison of two different groups of people and statistical validation.

The study sample was composed of 100 consecutive patients (56 females and 44 males) undergoing a routine CT scanning. The images, obtained through the reference plane of the palate, were post-processed on a PC workstation. A goniometer construction was superimposed and centered on the posterior nasal spine, where a grid of eighteen rays was developed to calculate the superficial soft tissues' depths of the face. To test the validity of the method, three of the rays were measured twice, by the same as by another observer. Step-by-step procedures were attained to get maximum standard in measures' reliability.

Superficial soft tissues' depths were obtained and sex differences were analyzed (Student's t and Wilcoxon rank-sum test). The statistical reliability was proven with Bland-Altman statistics and the upper 95% limit of agreement was 1.459 mm for intraobserver repeatability and 1.886 mm for interobserver reproducibility. Validation of the method was proven by intraclass correlation coefficient (0.99 both for repeatability and reproducibility) and mean differences (respectively 0.6% and 0.4%).

The method appeared easy to be applied, reliable and not observer dependent, so suitable for study single as multiple patients with CT images. Future analyses shall be possible.