2017 IEEE International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2017)

280 / 2017-01-31 17:41:22

Computer-aided Diagnosis System for Breast Cancer using RF Classifier

Keywords— Breast cancer, Mammogram, Random Forest, Enhancement, Texture Feature, Feature selection, Correlation.

Breast Cancer is one of the leading causes of cancer deaths among women. The best solution for this is early detection and treatment of breast cancer. Artificial Neural Network is intelligent and most widely used tools in breast cancer diagnosis. This main objective of this research is a diagnosis of breast cancer with a machine learning method based on random forest classifier. The digital mammogram images are taken MIAS database. The dataset consists of 280 mammogram images. Preprocessing is generally needed to enhance the poor quality of the image. The region of Interest is the suspicious area which is segmented, and then features are extracted by texture analysis. Feature selection technique is used for the detection of High-dimensional features and would be classified according to their class each other.
GLCM is used as the texture attribute which is then used to extract the suspicious area. CFS which is correlation based technique is used to select the best possible feature amongst all extracted features. Mean, standard deviation, smoothness, angular second moment (ASM), entropy, and correlation are the best feature that guarantees the improvement of classification with less feature dimension. RF (Random Forest) is used as a classifier. The result shows that the proposed method was achieved the accuracy 97.32%, sensitivity 97.45%, specificity 98.13% and ROC 97.28%.