Breast Cancer Detection Using Image Processing Techniques and Classification Algorithms

EasyChair Preprint 2101

Abstract

Breast cancer is the top cancer in women worldwide. Early detection of this disease and its classification into cases can improve the prognosis and even save lives by promoting timely clinical management to patients. An accurate diagnosis of breast cancer and its classification into benign, malignant and normal cases is a challenge in cancer research. Because of the ability to enable the computer to learn from past samples to detect and classify patterns, in machine learning, classification algorithms are widely used for cancer identification. However, many of them are focused on binary classification (cancer and no-cancer; benign and malignant). In this work, we present a Computer-Aided Diagnosis (CAD) approach for diagnosis and classification of patients into three conditions (malignant, benign and normal) from pixel mammogram images. For the classification task, we explore and compare three outstanding classifiers: Support Vector Machine (SVM), k-Nearest Neighbors (K-NN), and Random Forest (RF) to analyze their accuracy in decision making. In addition, we discuss the effects of pre-processed mammogram images before entering the classifier, which results in higher effective classification.

Keyphrases: Breast Cancer Detection, Machine Learning Classifiers, image processing

@booklet{EasyChair:2101,
  author    = {V Harvind Viswanath and Lorena Guachi-Guachi and Saravana Prakash Thirumuruganandham},
  title     = {Breast Cancer Detection Using Image Processing Techniques and Classification Algorithms},
  howpublished = {EasyChair Preprint 2101},
  year      = {EasyChair, 2019}}