BDBComp
Parceria:
SBC
Análise da Variação de Textura em Imagens Mamográficas para Classificação de Massas Suspeitas

P. B. RibeiroH. SchiabelA.C. PatrocinioR. A. F. Romero

In this paper, a method for the classification of mammary mass is proposed aiming to help the specialists as a second opinion. For this, initially the texture, very important for analyze of the mass, intensity and geometric descriptors have been extracted from Regions of Interest. Next, some methods of selection of attributes have been considered. They are Self-Organizing Map (SOM), Gaussian funcitons. Finally, the selected attributes have been used for training of neural networks: Multi-Layer Perceptron (MLP) for classifying 2 and 5 different kinds of breast masse.

http://www.lbd.dcc.ufmg.br/colecoes/wvc/2008/0023.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD