Evaluation of speed, repeatability, and reproducibility of digital radiography with manual versus computer-assisted cephalometric analyses.

Abstract

The aims of this study were to evaluate intra-examiner repeatability and inter-examiner reproducibility of landmarks using two cephalometric analysing techniques, manual and computerized, and to compare these for speed. One hundred lateral cephalometric radiographs were randomly selected and 11 angular and six linear parameters were traced and measured by two examiners using the manual method and Dolphin Image Software 9.0 on each radiograph. A Student's t-test for paired and independent samples was used to compare the mean values of intra- and inter-examiner differences. Intraclass correlation coefficients (ICC) were calculated to determine intra- and inter-examiner correlation (r value). Both operators were generally consistent in the repeated measurements; however, for one examiner, the differences for Na is perpendicular to A (P < 0.001), Na is perpendicular Pog, and U1-NA (P < 0.01) distance measurements were found to be statistically significant. Intra-examiner repeatability of landmarks both with the manual and Dolphin techniques showed high correlation coefficients. While inter-examiner reproducibility of landmarks was unacceptable, measurement errors with the manual technique were generally comparable with the Dolphin technique. The mean tracing times of the two operators for a single tracing was 2 minutes 41 seconds for Dolphin and 6 minutes 51 seconds for manual tracings. Computer-assisted cephalometric analysis does not increase intra- and inter-examiner reliability but can result in time saving.