Machine Learning Approach for Breast Cancer Prediction
S. Vasundhara1, B.V. Kiranmayee2, Chalumuru Suresh3
1S. Vasundhara, Department of Computer Science Engineering, VNR Vignana Jyothi Institute of Engineering & Technology, India.
2B.V. Kiranmayee, Department of Computer Science Engineering, VNR Vignana Jyothi Institute of Engineering & Technology, India.
3Chalumuru Suresh, Department of Computer Science Engineering, VNR Vignana Jyothi Institute of Engineering & Technology, India.
Manuscript received on 01 April 2019 | Revised Manuscript received on 07 May 2019 | Manuscript published on 30 May 2019 | PP: 2619-2625 | Volume-8 Issue-1, May 2019 | Retrieval Number: A1276058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Breast cancer is horrendous disease after skin cancer which is most common in woman and it is a foremost cause for the upsurge in mortality rate. Screening mammography is the operative procedure for detecting masses and abnormalities allied to breast cancer. Digital mammograms are utmost operative source that helps in early detection of cancer in women with no symptoms and diagnose cancer in women with symptoms like pain in lump, nipple discharge which diminutions deaths and upsurges chances of survival. Usually clinician cannot spare more time on a patient to weigh the complaints and suggest a possible diagnosis by considering past records. During this stage, there is more chance to medical errors and wrong diagnosis. By using machine learning in diagnosing breast cancer improves accuracy by reducing misclassifications and saves time in diagnosing. The proposed work is instinctive classification of mammogram images as Benign, Malignant and Normal using various machine learning algorithms. Classification is an identification technique used to classify consolidated data into different categories. Initially pre-processing of mammograms is performed by using Gabor wavelet filtering technique, Adaptive global threshold and morphological operations like opening and closing techniques for cancer analysis to shrink false positives. Finally classification of the pre-processed images is performed and mammograms are classified into benign, malignant and normal with the use of 3 classifiers Support Vector Machines, Convolutional Neural Network and Random Forest. The performance of the trained classifiers is evaluated using metrics to attest which model is efficient in classifying mammograms and predicting breast cancer.
Index Terms: Breast Cancer, Classification, Machine Learning, Mammograms.
Scope of the Article: Machine Learning