Machine Learning Based Classification of Histopathological Image for Breast Cancer
Vaishnawi Priyadarshni, Sanjay Kumar Sharma · 2024
One of most prevalent types of cancer and the main reason of fatality for women is breast cancer. Mammography images of the breast are used by radiologists to search for indications of potential tumor development, such as breast masses, tissue lumps that may be the result of cancer cells, and micro-calcifications, which are tiny calcium deposits that collect around aberrant tissue. Machine learning algorithms, which learn from historical data and can anticipate the category of fresh input, are used to classify benign and malignant tumors. Pre-processing, feature extraction, selection, and classification are the four steps in which the breast detection system is implemented in this paper. This paper introduces the Random Forest classifier which employs feature selection and transfer learning to identify and classify breast cancer in histopathological images. The suggested approach classifies benign and malignant cells by feeding features extracted from pictures into a fully connected layer using VGG-16 and Densenet 121. This classification is an excellent attempt that successfully detects using feature extraction and selection.