Curvelet and PNN classifier based approach for early detection and classification of breast cancer in digital mammograms

Breast Cancer is the most common incursive cancer which is found in females all through the world. of all the female cancers it comprises of 16% and it accounts for 22.9% of invasive cancer in women. of all the cancer deaths 18.2% are from breast cancer which includes males and females. As the modern science is improving many researches and techniques have been emerged to eradicate this dreadful disease. So there is a need of an automated computer aided diagnosis system and it is proposed here. This paper proposes a method to detect microcalcifications and circumscribed masses and also classify them as benign and malignant. For this a curvelet and texture analysis are used for feature extraction and PNN classifier is used for classification. The proposed method is evaluated using Mini Mammographic Image Analysis Society(Mini-MIAS) dataset and an accuracy of 93% has been obtained.