TY - GEN
T1 - Segmentation of medical images through competitive learning
AU - Dhavan, Atam P.
AU - Arata, Louis
N1 - Funding Information:
This work was supportedin part by the NIH grant CA49976. Authors are deeply gratefult o ThomasD ufresneP, rashanthK ini, CharlesP eck III, and YateenC hitrefor manyv aluables ugges-tions and discussionds uringthe developmenotf this work.
PY - 1993
Y1 - 1993
N2 - In image analysis applications, segmentation of gray-level images into meaningful regions is an important low-level processing step. Various approaches to segmentation investigated in the literature, in general, use either local information of gray-level values of pixels (region growing based methods, for example) or the global information (histogram thresholding based methods, for example). Application of these approaches for segmenting medical images with large structural information often does not provide satisfactory results. We present a novel approach to image segmentation that combines local contrast as well as global feature information. The presented method adaptively learns useful features and regions through the use of a normalized contrast function as a measure of local information and a competitive learning based method to update region segmentation incorporating global information about the gray-level distribution of the image. In this paper, we present the framework of such a self-organizing feature map, and show the results on simulated as well as real medical images.
AB - In image analysis applications, segmentation of gray-level images into meaningful regions is an important low-level processing step. Various approaches to segmentation investigated in the literature, in general, use either local information of gray-level values of pixels (region growing based methods, for example) or the global information (histogram thresholding based methods, for example). Application of these approaches for segmenting medical images with large structural information often does not provide satisfactory results. We present a novel approach to image segmentation that combines local contrast as well as global feature information. The presented method adaptively learns useful features and regions through the use of a normalized contrast function as a measure of local information and a competitive learning based method to update region segmentation incorporating global information about the gray-level distribution of the image. In this paper, we present the framework of such a self-organizing feature map, and show the results on simulated as well as real medical images.
UR - http://www.scopus.com/inward/record.url?scp=0005238337&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0005238337&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:0005238337
SN - 0780312007
T3 - 1993 IEEE International Conference on Neural Networks
SP - 1277
EP - 1282
BT - 1993 IEEE International Conference on Neural Networks
PB - Publ by IEEE
T2 - 1993 IEEE International Conference on Neural Networks
Y2 - 28 March 1993 through 1 April 1993
ER -