Biomedical active segmentation guided by edge saliency

Deformable models are very popular approaches in biomedical image segmentat ion. Classical snake models are edge-oriented and work well if the target o bjects have distinct gradient values. This is not always true in biomedical imagery, which makes the model very dependent on initial conditions. In th is work we propose an edge-based potential aimed at the elimination of loca l minima due to undesired edges. The new approach integrates knowledge abou t the features of the desired boundaries apart from gradient strength and u ses a new method to eliminate local minima, which makes the segmentation le ss sensitive to initial contours. (C) 2000 Elsevier Science B.V. All rights reserved.