Prediction, Detection and Recurrence of Breast Cancer Using Machine Learning Based on Image and Gene Datasets

Abstract

AbstractHealth care gives the life expectancy to each individual. Good health services procure the health of individuals. Presently, in 2021, the Health Care Index of India is 66.25 which is very low as compared to the expected HCI which should be 118.68. There are two type of diseases which affect the health of a person (a) communicable and (b) non-communicable. Cancer is a non-communicable disease which is a leading cause of death. This disease is caused by the uncontrolled division of abnormal cell in a part of body. There are various types of cancer like lung, breast, pancreas, ovary, colon and rectum. Among them, breast cancer is the second major cause of death in female. According to GLOBOCAN report 2018, breast cancer incidents around the world are 2,088,849, and in India, it is 162468. As per the National Breast Cancer Organization, breast cancer stages are of four types. In this review, we have mentioned about the noble technique of multivariant dataset and semi-supervised technique combined with hybrid feature selection method for the detection, prognosis and recurrence of breast cancer. There are two types of data through which breast cancer can be detected: one is genes and another one is images. These types can be analyzed by the collection of datasets for getting the best result for preventing the breast cancer. Genes in our DNA are the basic cause of mutation which causes breast cancer, and with the help of images, we get the scan of tumor. An image can be benign or malignant. Various researchers have worked on single dataset only which also have a drawback of false-positive results, but no researcher has used the multivariant technique to predict breast cancer, so there is a need to use the multivariant dataset together which will be a combination of images and genes dataset with feature selection method of machine learning to increase the level of accuracy and to predict the breast cancer. This review paper is divided into two parts, first part consists of image dataset for the diagnosis, prediction and recurrence of breast cancer, and second part consists of genes dataset with different machine learning techniques used by different researchers. Aim of this review paper is to analyze which machine learning technique is suitable when we combine two different datasets together with single feature selection method on both datasets and which will improve the accuracy rate for the detection, prediction and recurrence of breast cancer in females.KeywordsBreast cancerSemi-supervisedMachine learningHuman genomicsHealth careImage processing