Accuracy of Automatic Cephalometric Software on Landmark Identification

N. Anuwongnukroh, S. Dechkunakorn, S. Damrongsri, C. Nilwarat, N. Pudpong, W. Radomsutthisarn, S. Kangern

Research output: Contribution to journalConference articlepeer-review

3 Citations (Scopus)

Abstract

This study was to assess the accuracy of an automatic cephalometric analysis software in the identification of cephalometric landmarks. Thirty randomly selected digital lateral cephalograms of patients undergoing orthodontic treatment were used in this study. Thirteen landmarks (S, N, Or, A-point, U1T, U1A, B-point, Gn, Pog, Me, Go, L1T, and L1A) were identified on the digital image by an automatic cephalometric software and on cephalometric tracing by manual method. Superimposition of printed image and manual tracing was done by registration at the soft tissue profiles. The accuracy of landmarks located by the automatic method was compared with that of the manually identified landmarks by measuring the mean differences of distances of each landmark on the Cartesian plane where X and Y coordination axes passed through the center of ear rod. One-Sample T test was used to evaluate the mean differences. Statistically significant mean differences (p<0.05) were found in 5 landmarks (Or, A-point, Me, L1T, and L1A) in horizontal direction and 7 landmarks (Or, A-point, U1T, U1A, B-point, Me, and L1A) in vertical direction. Four landmarks (Or, A-point, Me, and L1A) showed significant (p<0.05) mean differences in both horizontal and vertical directions. Small mean differences (<0.5mm) were found for S, N, B-point, Gn, and Pog in horizontal direction and N, Gn, Me, and L1T in vertical direction. Large mean differences were found for A-point (3.0 < 3.5mm) in horizontal direction and L1A (>4mm) in vertical direction. Only 5 of 13 landmarks (38.46%; S, N, Gn, Pog, and Go) showed no significant mean difference between the automatic and manual landmarking methods. It is concluded that if this automatic cephalometric analysis software is used for orthodontic diagnosis, the orthodontist must correct or modify the position of landmarks in order to increase the accuracy of cephalometric analysis.

Original languageEnglish
Article number012028
JournalIOP Conference Series: Materials Science and Engineering
Volume265
Issue number1
DOIs
Publication statusPublished - 29 Nov 2017
Event2017 International Conference on Material Engineering and Manufacturing, ICMEM 2017 - Chengdu, China
Duration: 9 Oct 201711 Oct 2017

Fingerprint

Dive into the research topics of 'Accuracy of Automatic Cephalometric Software on Landmark Identification'. Together they form a unique fingerprint.

Cite this