Extraction and Grading of Malignant Tumor from Abnormal Mammogram Image using Marker Controlled Watershed Segmenta-tion

Abstract

Breast cancer is one of the most important causes of increased death rates among women worldwide. Mammography is the most effective method for early diagnosis of breast cancer. The main goal of mammography is to detect small, non-palpable cancers at a premature stage. Doctors need a reliable diagnostic procedure to detect the tumor at its early stage from the mammogram since early detection is vital for breast cancer diagnosis. The main objective of this paper is to extract and detect the tumor region from abnormal mammogram image, measure tumor growth and also help with treatment planning and dose calculation prior to radiation therapy. Hence automatic and efficient segmentation of calcification region from the abnormal mammogram image, marker controlled watershed approach has been employed. After segmentation benign or malignant tumor has been identified using shape features. The modified T classification system has been used to detect the grade of cancer based on surface area.