Breast Cancer Detection with Reduced Feature Set

Abstract

This paper explores feature reduction properties of independent component analysis (ICA) on breast cancer decision support system. Wisconsin diagnostic breast cancer (WDBC) dataset is reduced to one-dimensional feature vector computing an independent component (IC). The original data with 30 features and reduced one feature (IC) are used to evaluate diagnostic accuracy of the classifiers such as k-nearest neighbor (k-NN), artificial neural network (ANN), radial basis function neural network (RBFNN), and support vector machine (SVM). The comparison of the proposed classification using the IC with original feature set is also tested on different validation (5/10-fold cross-validations) and partitioning (20%–40%) methods. These classifiers are evaluated how to effectively categorize tumors as benign and malignant in terms of specificity, sensitivity, accuracy, F-score, Youden's index, discriminant power, and the receiver operating characteristic (ROC) curve with its criterion values including area under curve (AUC) and 95% confidential interval (CI). This represents an improvement in diagnostic decision support system, while reducing computational complexity.

A Hybrid Computer-aided-diagnosis System for Prediction of Breast Cancer Recurrence (HPBCR) Using Optimized Ensemble Learning

“…The performance measures of the classifiers are listed in Table 1, along with their definitions. Discriminant power (DP) was characterized as “poor” when DP < 1, “limited” when DP < 2, “fair” when DP < 3 and “good” in other cases [72], [73]. Among those measures AUC ROC is a global measure of diagnosis accuracy.…”

Dual Energy X-ray Methods for the Characterization, Quantification and Imaging of Calcification Minerals and Masses in Breast

“…However, the mean glandular dose of the polyenergetic approach was substantially increased, as expected, compared to the monoenergetic approach (approximately 53.8 times). 4 Tungsten anode/Rhodium filter; 5 Tungsten anode/Copper filter; 6 Tungsten anode/Cadmium filter; 7 monoenergetic; 8 polyenergetic.…”

“…Breast cancer is one of the most common causes of cancer death in the United States [1], accounting for approximately 12% of globally diagnosed cancers in 2012 [2]. Diagnosis of cancer in early stage leads to more efficient treatment [3][4][5]. Furthermore, early detection through screening and adequate follow-up of women with positive findings could significantly reduce breast cancer mortality (by 15-25%) [6].…”

Computational assessment of breast tumour differentiation using multimodal data

“…The fractions of representation from each of the classes are used to achieve the boundary shift and asymmetric regularisation to extend to the multi-class scenario [24]. The statistical model was evaluated using different classification models for its specificity, sensitivity, accuracy, F-score, Youden's index and the ROC curve [25]. Using the attributes selected for the model, the areas under the ROC curves are obtained [26].…”