Extraction of Skull and Face Surfaces from CT Images

摘要

Craniofacial reconstruction is a method to estimate the shape of a person's face based on the shape of his skull. The primary data in craniofacial reconstruction are the skull and face surfaces, but often human head data is taken using a CT scanner. The size of the CT image files is enormous, so we need to represent the CT image in a more straightforward format with small file size. In this paper, we propose a method for skull and face surfaces extraction from CT images. This study aims to obtain point clouds with small file size. In general, the method consists of two stages. First, the segmentation of the skull and face to get the outermost contour of the skull and face. Second, the point cloud construction. The proposed method has succeeded in extracting skull and face surfaces. The extraction process produces point clouds with an average file size of 9.1% of the size of the original CT image file.