Discrete-Time Recurrent Neural Networks for Medical Image Segmentation Based on Competitive Layer Model with LT Neurons

This paper discusses a class of Discrete-Time Recurrent Neural Networks with LT neurons based on Competitive Layer Model (CLM-DT-LT-RNNs). It first addresses the boundedness and complete stability of the networks, then a theorem is given to let the networks have CLM phenomena. Such networks are applied to medical image segmentation by using the global gray-level information and the contextual information of pixels. In order to alleviate time and storage consuming, a technique of divide-and-merge (DAM) is used. Simulation results are used to illustrate the application in image segmentation.