LSB neural network based segmentation of MR brain images

Least square backpropagation (LSB) algorithm is employed to train a three-layer neural network for segmentation of magnetic resonance (MR) brain images. The simulation results demonstrate the use of LSB algorithm as a promising method for the segmentation of multi-modal medical images. The training time has been dramatically reduced comparing with that of BP network. The influence of the number of neurones in the hidden layer of the network is discussed in the paper.