Neural Network in Developing Software for Indentifying Arch Form

The treatment of Class I malocclusion treatment is to arrange the teeth position in a good arch form. Arch form consists of tooth size and arch dimension (intercanine width, canine depth, intermolar width, molar depth). Several ways are developed to describe arch form. A lot of methods used to describe arch form qualitatively. The objective of this study is to develop qualitative arch form diagnostic references using artificial neural network from pre-post treatment dental cast scanning result. Pre-post orthodontic treatment dental casts (1990-2006) from Post Graduate Clinic Faculty of Dentistry University of Indonesia and 3 other orthodontists were gathered and scanned. Data were measured using Image Tool and analyzed using Stata 9. ANOVA was used to compare arch forms (square, oval, tapered) and gender (male and female), with each component of arch dimension upper and lower jaw, before and after treatment; and also arch perimeter to kinds of treatment The results were compiled to determine variables in building the software for analyzing arch form qualitatively. The data from190 pre-post orthodontic treatment dental casts consisted of 42 male (22.1%) and 148 female (77.9%) treated without extraction (32.63%), 4 Premolars extraction (48.42%), Upper Premolars extraction (11.05%), atypical extraction (7.90%). Gender and all variables from pre treatment did not influence arch form, except kinds of treatment. Therefore, only post treatment data are included for arch form analysis. The shape of arch form (square, oval and tapered) can be described qualitatively by software using artificial neural network. This software could describe arch form with the accuracy of 76.3158%. This study concluded that Intercanine width, Canine depth, Intermolar width, and molar depth were variables that influenced arch form. A software using artificial neural network to describe arch form qualitatively could be used for diagnostic reference to Class I malocclusion orthodontic post treatment.