Parameter Estimation with a Bayesian Network in Medical Image Segmentation

P.S. Rodrigues and G.A. Giraldi (Brazil)

Keywords

Medical Image Segmentation, Bayesian Networks, Snakes.

Abstract

Parameter estimation is a hard problem for image pro cessing and segmentation tasks. In the case of 3D med ical images, the segmentation can be performed slice by-slice, extracting the Region of Interest (RI) in each slice. This task can be accomplished by using edge en hancement methods followed by Active Contour Mod els, also called snakes. In this case, we must set for each slice several parameters such as thresholds, kernel size for spatial filters, snake parameters (tension, rigid ity), etc., which can be a tedious and time consuming task. In this paper we present a new methodology to adjust the parameter values in the slice + 1 from the extracted region in the slice . For each slice, several sets of candidate parameter values and consequently sets of estimated regions are randomly generated, and a bayesian network is used to maximize the probability of the RI in the slice + 1 given the RI in slice . We have tested this methodology in a large medical image database.

Important Links:



Go Back