Performance analysis of breast cancer detection using hybrid classification

Abstract

The architectural distorted regions in mammogram images are detected and segmented using computer aided hybrid classification approach in newlinethis research work. The main importance of this research work is to provide a newlinecomputer aided methodology for screening the distorted regions in mammogram newlineimages. In present approach, the classification accuracy of the conventional newlinemethods is not suitable for further diagnosis process such as malignant and newlinebenign. Hence, the main objective of this research is to develop an efficient newlinearchitectural region detection method using soft computing method with high newlineclassification accuracy for further diagnosis purpose. This proposed method has newlinetwo stages of the proposed flow as architectural distorted detected mammogram newlineimage and segmentation of architectural distorted regions in mammogram newlineimages. The first stage of this proposed method uses Random Forest (RF) newlineclassification method which classifies the source mammogram image into either newlinenormal or abnormal. In second stage of the proposed method, the abnormal newlineimage is further classified into either Benign or Malignant using Adaptive Neuro newlineFuzzy Inference System (ANFIS) classification approach. newlineThe proposed methodology for architectural distorted region detection newlineis tested on the publicly available mammogram datasets Mammographic Image newlineAnalysis Society (MIAS) and Digital Database for Screening Mammography newline(DDSM) respectively. In this research, the mammogram images from MIAS newlinedataset are grouped into normal case (156 images), benign case (122 images) newlineand malignant case (98 images). The mammogram images from DDSM dataset newlineare grouped into normal case (144 images), benign case (112 images) and newlinemalignant case (145 images).The overall average detection rate of the proposed newlinesystem on the mammogram images in MIAS dataset is about 98.7%. The overall newlineaverage detection rate of the proposed system on the mammogram images in newlineDDSM dataset is about 98.3%. The extensive simulations are carried out on the newlinemammogram images which are obtained from these data