Multi-spectral magnetic resonance image segmentation using LVQ neural networks
Javad Alirezaie, Claude Nahmias, M. Edward Jernigan · 2002
Segmentation of images obtained from magnetic resonance (MR) imaging techniques is an important step in the analysis of MR images of the human body. The multi-spectral nature of MRI has been exploited in the past to obtain better performance in the segmentation process. The new emerging field of artificial neural networks promises to provide improved solutions for the pattern classification of medical images. The authors present the application of a learning vector quantization (LVQ) neural network for the multispectral supervised classification of MR images. The authors have modified the LVQ for better and more accurate classification. The authors compare the results using multispectral images to those with a single slice image. This comparison shows that the authors' method is insensitive to the gray-level variation of MR images between different slices. Also, a comparison with the classical maximum likelihood classifier (MLC) demonstrates the superiority of the authors' LVQ ANN approach.