Constraints setting method for skeleton CT image segmentation using structural analysis

Abstract

We present a constraint condition setting method for segmentation of CT images based on structural analysis. This proposed method could be added to our previous works. In first report, our method finds boundaries of clusters by repetition of structural analysis and removing voxels with higher von Mises strain. In second report, our method find optimal loading condition for first report. We limit the number of boundary condition setting pattern and evaluate the con Mises strain distribution. By using this method, we can compute the optimal boundary condition and strain distribution. Indeed, we demonstrate segmentation results for various types of CT images including living objects.