INVESTIGATION OF FILTER SETS FOR SUPERVISED PIXEL CLASSIFICATION OF CEPHALOMETRIC LANDMARKS BY SPATIAL SPECTROSCOPY

The diagnostic process of orthodontics requires the analysis or: a cep halometric radiograph, Image landmarks on this two-dimensional lateral projection image of the patient's head are manually identified and sp atial relationships are evaluated. This method is very time consuming. A reliable method for automatic computer landmark identification does not exist. Spatial Spectroscopy is a proposed method of automatic lan dmark identification on cephalometric radiographs, that decomposes an image by convolving it with a set of filters followed by a statistical decision process. The purpose of this paper is to discuss and lest ap propriate filter sets for the application of Spatial Spectroscopy for automatic identification of cephalometric radiographic landmarks. This study evaluated two different filter sets with 15 landmarks on fourte en images. Spatial Spectroscopy was able to consistently locale landma rks on all 14 cephalometric radiographs tested. The mean landmark iden tification error of 0.841+/-1.253 pixels for a Multiscale Derivative f ilter set and 0.912+/-1.364 pixels for an Offset Gaussian filter set w as not significantly different. Furthermore, there were no significant differences between identification of individual landmarks for the Mu ltiscale Derivative and the Offset Gaussian filter set (P>0.05). These results suggest that Spatial Spectroscopy may be useful in landmark i dentification tasks. (C) 1997 Elsevier Science Ireland Ltd.