An alternative approach for medical image segmentation
Furkan Nur Deniz, Barış Baykant Alagöz, Mehmet Emin Tagluk · 2010
In this study a segmentation algorithm controlled by a dissimilarity function composed of a few criterion for segmentation of medical images, starting from a seed point and growing toward the boundary of the considered anatomical tissue, is proposed. It has been shown that the algorithm can be controlled to overcome the complications such as noise, variations in the pixel values and weak connectivity encountered on the targeted anatomical region in the medical image.