Title :
Identification of variegated coloring in skin tumors: neural network vs. rule-based induction methods
Author :
Durg, Ajaya ; Stoecker, William Aiaya V ; Cookson, John P. ; Umbaugh, S.E. ; Moss, Randy H.
Author_Institution :
Missouri Univ., Rolla, MO, USA
Abstract :
The use of neural networks for automatic identification of variegated coloring, which is believed to be one of the most predictive features for malignant melanoma, is described. The Nestor development system (NDS) was chosen for neural network implementation. At the heart of NDS is a three-layer neural network called a restricted Coulomb energy (RCE) network. The learning scheme and the database for detection of variegated coloring are discussed. Results are reported.<>
Keywords :
colour; medical image processing; neural nets; skin; 3-layer neural network; Nestor development system; database; learning scheme; malignant melanoma predictive features; medical diagnosis; restricted Coulomb energy network; rule-based induction methods; variegated coloring identification; Artificial neural networks; Biological neural networks; Cancer; Decision trees; Humans; Intelligent networks; Malignant tumors; Neural networks; Neurons; Skin neoplasms;
Journal_Title :
Engineering in Medicine and Biology Magazine, IEEE
