Reproducibility of landmarks on three-dimensional cephalometric analysis for orthognathic surgery

Objective To study the Reproducibility of landmarks of soft and hard tissueso as to provide reference to accurate three-dimensional cephalometric analysis for orthognathic surgery.Methods The study population consisted of 59 volunteers examined with spiral 3D computed tomography (CT).For each analysis,54 cephalometric reference landmarks were identified on the three spatial axes(x,y,z).All the identifications were performed by the same observer.This was repeated three times during 3 sessions at least 1 week apart.In total,3186 imaging measurements were taken,as each determination has 3 coordinates,9558 data were processed with SAS9.2 statistical package.To discover the reproducibility of the method on landmark location,an ICC was undertaken using one variation factors:time (t1,t2 and t3) for each axis (x,y,z) and landmark.Results The ICC was＞0.9 for 86.48％ of all the landmarks,which with high reliability.The ICC was ＞0.9 for 86.38 ％of the hard tissue landmarks,and for 87 ％ of the soft tissue landmarks.The reproducibility of landmarks in three axis was z＞x＞y.Those with lower ICC were Me*,Zy* R,Zy* L and ZyL.Conclusions Three dimensional landmark identification using CT can offer consistent and reproducible data so as to create cephalometric analyses reflectingthe 3D reality. Key words: Three dimensional cephalometry;  Landmark;  Reproducibility;  Orthognathic surgery