Split and Merge-Based Breast Cancer Segmentation and Classification

Split and Merge-Based Breast Cancer Segmentation and Classification

Ichrak Khoulqi (University of Sultan Moulay Slimane Beni-Mellal, Morocco) and Najlae Idrissi (University of Sultan Moulay Slimane Beni-Mellal, Morocco)
Copyright: © 2020 |Pages: 14
DOI: 10.4018/978-1-7998-1021-6.ch012

Abstract

Breast cancer is the most frequent cancer in morocco with 36.1%. It is the second leading cause of death for women all over the world. The effective way to diagnose and treat breast cancer is the early detection because it increases the success of treatment and the chances of survival. Digitized mammographic images are one of the frequently used diagnosis tools to detect and classify the breast cancer at the early stage. To improve the diagnosis accuracy, computer-aided diagnosis (CAD) systems are beneficial for detection. Generally, a CAD system consists of four stages: pretreatment, segmentation, features extraction, and classification. In this chapter, the authors present some work in the development of a CAD system in order to segment a breast tumor (microcalcifications) on mammographic images and classify it by choosing the algorithm that gives a good rate using a technique of a vote.
Chapter Preview
Top

Introduction

Breast cancer is a malignant tumour that originates in breast cells. The word “malignant” means the tumor can spread (metastasize) to other parts of the body (Institut National du Cancer, 2016) . Breast cells sometimes undergo changes that make their growth or behaviour abnormal. These changes can lead to benign breast conditions, such as atypical hyperplasia and cysts. In some cases, changes in breast cells can cause breast cancer. In the context of prevention it is necessary to use a radiography tool that allows to better visualize the different parts of the breast and one of the best tools is Mammography which is a technique of radiography, particularly adapted to the breasts of the woman. It is intended to detect abnormalities as soon as possible before they cause clinical symptoms (IMENE CHEIKHROUHOU,2012). The mammography image is the result of attenuation of a beam of Xrays passing through the different mammary tissues. The attenuation of this beam depends essentially on the composition of the tissues through which it passes. Indeed, the grease is considered a transparent radio zone since it has a very light physical density. As a result, it appears very dark on a mammogram. The opaque radio zones appear clear and correspond to the fibroglandular tissue and calcium which is the essential component of the mammary lesions (IMENE CHEIKHROUHOU,2012).Mammography is usually taken in different directions called incidences. A good incidence is to visualize the maximum breast tissue by spreading it as much as possible on the X-ray plate. Depending on the part of the breast examined, different implications are used. The most frequently used incidences are the incidence of the face also called Cranio Caudale (CC), the oblique external incidence called Medio Lateral Oblique (MLO) and the incidence of profile.

The mammography is an essential examination for the diagnosis of breast diseases in the presence of a symptom: palpable nodule, skin changes, discharge, inflammation.

Complete Chapter List

Search this Book:
Reset