Detection and Classification of Breast Cancer using Artificial Intelligence Approaches
Rajesh Kumar, Satyendra Singh, Sakshi Agrawal · Zenodo (CERN European Organization for Nuclear Research) · 2022
Breast Cancer is a prevalent cancer in women globally, according to statistics from around the world, it accounts for many new cancer cases and cancer-related deaths today, making it a huge public health issue. Early identification of breast cancer can significantly increase prediction and survival chances by encouraging patients to receive timely clinical therapy. In this experimental work an automatically breast cancer detection model is developed using histopathological image. This application can provide accurate classification of Benign (non-cancerous) and Malignant (cancerous) tumors (i.e., binary classification). The model developed here provided an accuracy of 95.00%. The pre-trained DenseNet201 model, using Transfer Learning, is used to develop this model removing its “top” and using its pre-trained layers generate a feature stack using our target dataset provided, then the “fully-connected layer” is bootstrapped to generate the predictions. The goal of this model is to help the user identify breast cancer early on and be a beneficial tool for all medical professionals and society at large.