AUTONOMOUS CT REPLACEMENT METHOD FOR THE SKULL PROSTHESIS MODELLING

Abstract

The geometric modeling of prosthesis is a complex task from medical and engineering viewpoint. A method based on CT replacement is proposed in order to circumvent the related problems with the missing information to modeling. The method is based on digital image processing and swarm intelligence algorithm. In this approach, a missing region on the defective skull is represented by curvature descriptors. The main function of the descriptors is to simplify the skull’s contour geometry; and they are defined from the Cubic Bezier Curves using a meta-heuristic process for parameter’s estimation. The Artificial Bee Colony (ABC) optimization technique is applied in order to evaluate the best solution. The descriptors from a defective CT slice image are the searching parameters in medical image databases, and a similar image, i.e. with similar descriptors, can be retrieval and used to replace the defective slice. Thus, a prosthesis piece is automatically modeled with information extracted from distinct skulls with similar anatomical characteristics.