Modified Dense Convolution Neural Network (md-cnn) for Breast Cancer Detection Using Mammography Images

Cancer is a widely dreaded illness globally, posing a significant challenge to enhancing life expectancy in numerous countries. As per the current statistical surveys, breast cancer is one of the foremost cancer types diagnosed around the world, particularly affecting women. Early and accurate diagnosis plays a key role in improvising the outcomes of treatment and increasing survival rate. Regular clinical breast-exams, self-exams, and mammography facilitate identification of the disease at a very budding stage. With the advent of recent technologies, techniques like deep learning (DL) and machine learning (ML) are often utilized for improving effectiveness of the early detection of the disease and can also minimize the occurrences of false positives and missed diagnoses. The aforementioned study proposes a novel model for timely diagnosis of breast cancer using DL techniques and the concept of transfer learning over mammographic images. The suggested model, Modified Dense-Convolution Neural Network (MD-CNN) showcases promising results as compared to the various models developed in state of the art. The model is proficient in classifying mammography images into benign or malignant categories, thus aiding the process of the early detection or timely diagnosis.