Analysing Breast Cancer using Convolution Neural Network

Abstract

Technological development of Soft Computing and Artificial Intelligence contributed a lot towards disease diagnosis in medical science. For providing solutions to the biological inspired problems in medical domain like Breast Cancer (BC) soft computing methods can give the flexible information. As per 2020 about 2 million women were detected with Breast Cancer (BC) creating most common malignancy among women worldwide. This rises both incidence and mortality has occurred during the past three decades due to evolving risk factors, improved cancer registries, and earlier diagnosis. There is a large number of risk factors for BC, some of which can be changed and others cannot. Eighty percent of people diagnosed with BC nowadays are over the age of fifty. Molecular subtype and developmental stage are both important in determining the likelihood of survival. When it comes to clinical presentation, behaviour and shape, invasive BC span a broad spectrum of tumours. In this paper, Convolution Neural Network (CNN) used to recognize the BC tumor because it is another sort of neural network that can discover key information in both image and time series data. By applying CNN on the 2023 RSNA (Radiological Society for North America) Screening Mammography Breast Cancer data we analysed how best CNN algorithm is for identifying breast cancer with accuracy and also tried to analyse at what age Breast Cancer is mostly occurred in women.