Detection of Malignancy in Mammogram by Modified Convolution Neural Network

In recent decades, cases on breast cancer are increasing rapidly and there are many awareness campaigns conducted in order to ensure the timely diagnosis of the same. Mammogram is a medical imaging technique by X-Ray for the breast there are two views of the mammogram they are namely mediolateral oblique view and craniocaudal view the mammogram images for research are provided by various data banks like mammographic image analysis society (MIAS) and many others. Engineers play a vital role in enhancing, processing and classifying medical mammogram images. Support from engineer to radiologist increases the precision in detection of tumor the malignant tumor detection in the digital mammogram involves the stages starting with preprocessing followed by segmentation then feature extraction and finally classification. Target in the field of mammogram tumor research is to identify the positive of the disease in early stage and to improve accuracy of detecting diseased patient and healthy person from digital mammogram the proposed Marker controlled watershed Algorithm with AlexNet (MCWAAN) mechanism shows its better performance in reducing False Positive and increasing accuracy of detection. © 2021 IEEE.