Sheshadri Thiruvenkadam

A novel model for boundary determination that incorporates prior shape information into geometric active contours is presented. The basic idea of this model is to minimize the energy functional depending on the information of the image gradient and the shape of interest, so that the boundary of the object can be captured either by higher magnitude of the image gradient or by the prior knowledge of its shape. The level set form of the proposed model is also provided. We present our experimental results on some synthetic images, functional MR brain images, and ultrasound images for which the existing active contour methods are not applicable. The existence of the solution to the proposed minimization problem is also discussed.

In this work, a level set energy for segmenting the lungs from digital Posterior-Anterior (PA) chest x-ray images is presented. The primary challenge in using active contours for lung segmentation is local minima due to shading effects and presence of strong edges due to the rib cage and clavicle. We have used the availability of good contrast at the lung boundaries to extract a multi-scale set of edge/corner feature points and drive our active contour model using these features. We found these features when supplemented with a simple region based data term and a shape term based on the average lung shape, able to handle the above local minima issues. The algorithm was tested on 1130 clinical images, giving promising results.