ABSTRACT
The use of computer aided diagnostic (CAD) models has been proposed to aid in the detection and classification of breast cancer. In this work, we evaluated the performance of neural network models of multilayered perceptrons and nonlinear support vector machines to classify breast cancer nodules. Ten morphological characteristics, from the outline of 569 samples, were used as input to the classifiers. The average results obtained in the set of 50 simulations showed that the proposed models presented good performance (all exceeded 90.0 %) in terms of the accuracy of the test set. The nonlinear support vector machine algorithm stands out when compared to the proposed multilevel perceptrons neural network algorithm, with accuracy of ≈ 99,0 % and false negative rate of approx 2.0 %. The neural network model presented inferior performance to the non-linear support vector machine classifier. The average results, with the application of the proposed models, are shown to be promising in the classification of breast cancer.
Keywords:
neural network; support vector machine; breast cancer
