Title :
Medical images segmentation by neuro-genetic approach
Author :
Benamrane, Nacéra ; Fekir, Abdelkader
Author_Institution :
Dept. of Comput. Sci., Univ. of Sci. & Technol. of "Mohamed Boudiaf" Oran, Algeria
Abstract :
In this paper, we propose an approach based on the RBF neural networks and the genetic algorithms for magnetic resonance (MR) brain images segmentation. In the feature extraction stage, nine features are calculated and used as RBF network inputs. The genetic algorithm build automatically a RBF NN (It determines the number of hidden neurons, and their respective centers and radii). Nets´ fitness is calculated by using the mean square error.
Keywords :
biomedical MRI; feature extraction; genetic algorithms; image segmentation; magnetic resonance imaging; medical image processing; radial basis function networks; MR brain image segmentation; RBF neural network; feature extraction; genetic algorithm; hidden neurons; magnetic resonance; mean square error; medical image segmentation; medical imagery; neuro-genetic approach; Biological neural networks; Biomedical imaging; Brain; Feature extraction; Genetic algorithms; Image segmentation; Magnetic resonance; Neural networks; Neurons; Radial basis function networks; Genetic algorithm; Medical imagery; RBF; Regions; Segmentation;
Conference_Titel :
Information Visualisation, 2005. Proceedings. Ninth International Conference on
Print_ISBN :
0-7695-2397-8
