Staging of cancer is said to be unsigned in natural history. In this advanced growth of medical research, finding out the tumor growth is demanding in the midst of the researchers. The aim of this study is to provide a pathway for predicting tumor stages with the help of processing tools. Since breast cancer tumor is said to in the worlds second place of disease, prevention of higher growth sound be maintained. With sticking out to reduce costs and high predicted upshot ultrasound images, were chosen. Ultrasound screening images are targeted to produce an ending over currently predicting techniques. Breast cancer tumor growth findings lead to the public to get aware of frequent possible medical checkup. Overcoming the margins of existing techniques and methods, a new approach is stretched out for prediction phase. Neural networks and image segmentation based concepts are composed and a final wrapping up will lead advanced finding of efficiency. This study starts from the collection of ultrasound images, followed by the methods that define for predicting the cancer tumor, and the limitations to be met.