Breast Cancer Segmentation in Mammogram Using Artificial Intelligence and Image Processing: A Systematic Review

Abstract

Background: Breast cancer is the second leading cause of death in females worldwide. Mammograms are useful in early cancer diagnosis as well when the patient can sense symptoms or they become observable. Inspection of mammograms in search of breast tumors is a difficult task that radiologists must carry out frequently. Objective: This paper provides a summary of possible strategies used in automated systems for a mammogram, especially focusing on segmentation techniques used for cancer localization in mammograms. Methods: This article is intended to present a brief overview for nonexperts and beginners in this field. It starts with an overview of the mammograms, public and private available datasets, image processing techniques used for a mammogram and cancer classification followed by cancer segmentation using the machine and deep learning techniques Conclusion: The approaches used in these stages are summarized, and their advantages and disadvantages with possible future research directions are discussed. In the future, we will train a model of medical images that can be used for transfer learning in mammograms.