Enhanced Multi-Resolution Feature Extraction for the Early Detection of Breast Cancer

Abstract

In recent years,breast cancer has been one among leading cause for cancer deaths among the female population in the world. Digital mammograms are the most common reliable tool for the elimination and detection of breast cancer. Mammography is an important tool for visualization and detects breast cancer using low-level x-rays. X-ray images of the breast must be accurately evaluated to fetching the earliest stages of cancer growth. In this paper, we proposed a Multi Resolution based Random Forest (MRRF) Technique by prior detection of Mammogram Image Based on K-means Cluster as well as Random Forest Classification. The proposed approach is depending on a multi resolution basedfeature extraction which can be used to extract features at multi-level of decomposed images. The four stages of the proposed strategy include preprocessing to redesign an image, K-means clustering for segmentation, for the feature extraction, discrete wavelet transform as well as the random-forest classification techniques are used for prediction of regular or malignant image.The proposed methods have been implemented in python and the results show that our proposed work is possible and effective for an accurate classification as well as identification of breast cancer.