Year : 2019, Volume : 10, Issue : 5
First page : ( 786) Last page : ( 791)
Print ISSN : 2322-0414. Published online : 2019  01.
Article DOI : 10.5958/0976-5506.2019.01108.2

Computer Aided Classification of Breast Lesions in Digital Mammograms

Sangeethapriya K.1, Dhivya Josephin Arockia1, Thamizhvani T. R.1, Hemalatha R. J.1

Department of Biomedical Engineering, Vels Institute of Science Technology and Advanced tudies, Chennai, India

Corresponding Author: K. Sangeethapriya Department of Biomedical Engineering, Vels Institute of Science Technology and Advanced Studies, Chennai, India Email: sangeetha.se@velsuniv.ac.in

ABSTRACT

Objective: Mammography technique is mostly used for detecting the presence of abnormal breast lesions among women. Differentiating these abnormalities is a most difficult task faced by the radiologists. By using this proposed technique the rate of unnecessary biopsies can be limited. This paper deals with an effective way of detecting the breast lesions using curvelet transform. This proposed paper follows a stepwise procedure such as (a) Preprocessing (b) Region of Interest Segmentation. (c) Applying Curvelet Transform (d) Feature Extraction & finally (e) Classification of features using different kernels of SVM. It is inferred from the observed results that the SVM(Linear) classifier showed a good accuracy rate of 80%.

Top

Keywords

Mammogram, MIAS Database, Cancer Detection, Benign, Malignant, Curvelet transform.

Top