Malignant-Benign Classification of Pulmonary Nodules by Bagging-Decision Trees

Abstract

Today, computer-aided detection systems have been highly needed in many clinical applications. In this study, a new Computer-aided Diagnosis system (CAD) was proposed for classifying pulmonary nodules as malignant and benign. The classifiers of the Bagging-decision trees were utilized. On the classifying of malign and benign nodule patterns, classification performance values are calculated as 94.7 % sensitivity and 0.950 AUROC for benign class; 80.0 % sensitivity and 0.888 AUROC for malign class; 77.8 % sensitivity and 0.935 AUROC for uncertain class by 86.8 % accuracy of the classifier.