A statistical-genetic algorithm to select the most significant features in mammograms

TitleA statistical-genetic algorithm to select the most significant features in mammograms
Publication TypeBook Chapter
Year of Publication2007
AuthorsSanchez-Ferrero, G. V., and J. I. Arribas
Book TitleLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4673 LNCS
ISBN Number9783540742715
KeywordsBreast cancer, Diagnosis, Feature extraction, Genetic algorithms, Mammography, Microcalcification classification, Network complexity, Neural network classifiers, Neural networks, Tumors

An automatic classification system into either malignant or benign microcalcification from mammograms is a helpful tool in breast cancer diagnosis. From a set of extracted features, a classifying method using neural networks can provide a probability estimation that can help the radiologist in his diagnosis. With this objective in mind, this paper proposes a feature selection algorithm from a massive number of features based on a statistical distance method in conjunction with a genetic algorithm (GA). The use of a statistical distance as optimality criterion was improved with genetic algorithms for selecting an appropriate subset of features, thus making this algorithm capable of performing feature selection from a massive set of initial features. Additionally, it provides a criterion to select an appropriate number of features to be employed. Experimental work was performed using Generalized Softmax Perceptrons (GSP), trained with a Strict Sense Bayesian cost function for direct probability estimation, as microcalcification classifiers. A Posterior Probability Model Selection (PPMS) algorithm was employed to determine the network complexity. Results showed that this algorithm converges into a subset of features which has a good classification rate and Area Under Curve (AUC) of the Receiver Operating Curve (ROC). © Springer-Verlag Berlin Heidelberg 2007.