Automated Detection Of Breast Cancer From Mammography Images

Abstract

Breast cancer stands as the most prevalent cancer among women and ranks as the second leading cause of cancer-related deaths in this group. Currently, we lack effective methods for both preventing and curing breast cancer, largely because its exact causes remain unclear. However, early detection plays a vital role in diagnosing and managing the disease, significantly improving the chances of a full recovery. Mammography has emerged as the most reliable tool for spotting breast cancer in its earliest stages, and it remains the go-to method for screening and diagnosing this condition. Additionally, this technique can help identify other issues and may indicate whether a tumor is normal, benign, or malignant. This thesis aims to clarify how to extract features from mammogram images to pinpoint areas affected by cancer, which is a critical step in the detection and verification process. Various algorithms were employed to locate cancerous regions within the mammogram images, directly processing the original grayscale images through a series of image processing techniques. The first step in this identification process is image segmentation, which involves distinguishing the foreground areas of the image from the background.