Segmentation and Classification of Breast Cancer Using Independent Component Analysis, Texture Features and Neural Networks

Lucio Flávio A. CamposEmanuel C. M. LemosLuis C. O. SilvaDaniel D.CostaAllan Kardec Barros

We propose a method for segmentation and classification of breast cancer in digital mammograms using Independent Component Analysis (ICA), Texture Features and Multilayer Perceptron (MLP) Neural Networks. The method was tested for a mammogram set from MIAS database, resulting in 90.15% success rate, with 92% of specificity and 88.3% of sensitivity.

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

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: