Virtual Instrumentation Based Breast Cancer Detection and Classification Using Image-Processing

Breast cancer is an important cause of death in woman. An initial symbol of breast cancer is multiplication of malignant cells in the mammogram. The projected method has been applied in five stages consisting of image preprocessing, extraction of ROI, feature extraction, normalization and classification. In this procedure, we can count the number of defected cells and find their location with image processing. To classify the cancerous region the SVM (SUPPORT VECTOR MACHINE) based classifier technique is used in this paper. The projected method has been estimated with MINI MAMMOGRAM IMAGE ANALYSIS SOCIETY (MIAS) database and validated with 200 images of MIAS database.