Title :
Unsupervised segmentation of multi-echo MR images with an ART-based neural network
Author :
Li, Wanqing ; Attikouzel, Y.
Author_Institution :
Centre for Intelligent Inf. Process. Syst., Western Australia Univ., Nedlands, WA, Australia
Abstract :
This paper investigates the suitability of an ART-based neural network for unsupervised segmentation of multi-echo MR images. The ART2A network was used to segment standard dual-echo MR images. Two problems were identified with the basic ART2A: one, the network was hardly convergent; and two, the categorization depended on the order of presentation of the patterns. In order to solve these two problems, a dynamic learning parameter and random pattern presentation method were introduced. Results using a number of actual dual-echo MR images with the modified ART2A network show that ART-based networks can be used for segmentation of multi-echo MR images
Keywords :
ART neural nets; biomedical NMR; image segmentation; medical image processing; unsupervised learning; ART-based neural network; ART2A network; dynamic learning parameter; multi-echo MR images; random pattern presentation method; unsupervised segmentation; Adaptive filters; Biological neural networks; Data mining; Humans; Image converters; Image segmentation; Information processing; Neural networks; Pathology; Subspace constraints;
Conference_Titel :
Neural Networks, 1995. Proceedings., IEEE International Conference on
Conference_Location :
Perth, WA
Print_ISBN :
0-7803-2768-3
DOI :
10.1109/ICNN.1995.487819
