CFP last date
22 April 2024
Reseach Article

Detection and Classification of Tumors in a Digital Mammogram

Published on February 2013 by Rajashekar K. R
International Conference on Electronic Design and Signal Processing
Foundation of Computer Science USA
ICEDSP - Number 1
February 2013
Authors: Rajashekar K. R
78d030e8-2cd6-4e69-b576-ff39ba053263

Rajashekar K. R . Detection and Classification of Tumors in a Digital Mammogram. International Conference on Electronic Design and Signal Processing. ICEDSP, 1 (February 2013), 24-28.

@article{
author = { Rajashekar K. R },
title = { Detection and Classification of Tumors in a Digital Mammogram },
journal = { International Conference on Electronic Design and Signal Processing },
issue_date = { February 2013 },
volume = { ICEDSP },
number = { 1 },
month = { February },
year = { 2013 },
issn = 0975-8887,
pages = { 24-28 },
numpages = 5,
url = { /specialissues/icedsp/number1/10350-1008/ },
publisher = {Foundation of Computer Science (FCS), NY, USA},
address = {New York, USA}
}
%0 Special Issue Article
%1 International Conference on Electronic Design and Signal Processing
%A Rajashekar K. R
%T Detection and Classification of Tumors in a Digital Mammogram
%J International Conference on Electronic Design and Signal Processing
%@ 0975-8887
%V ICEDSP
%N 1
%P 24-28
%D 2013
%I International Journal of Computer Applications
Abstract

Mammography is an effective way that has demonstrated the ability to detect breast cancer at early stages with high sensitivity and specificity. Due to textural variation in image intensity , diagnosis performance varies from 60% to 80% in manual reading of mammogram . This paper demonstrates a novel approach for classifying mammograms by computer aided design using image processing and data mining techniques. This experiment consists of four stages namely preprocessing, segmentation, extraction of features and classification. In preprocessing the breast image is standardized. Then suspicious regions of cancer are acquired from mammogram by K-means clustering technique. Features are extracted from these region and are given as input to the pretrained decision tree based classifier, which in turn classifies the mammogram into normal, benign and malignant. The system has very high accuracy and has been verified with the ground truth given in the database (mini-MIAS database & DDSM). The false negative rate was as very low compared to the other existing methods.

References
  1. Gonzalez R. C, Woods R. E and Eddins S. ,2002 Digital Image Processing Using MATLAB® -, 2e, Prentice Hall.
  2. Gonzalez R. C and Woods R. E 2002Digital Image Processing, 2e,Prentice Hall.
  3. Witten H. and Frank E, Data Mining Practical Machine Learning Tools and Techniques, Second Edition
  4. Duda R. ,Hart P. , Stork D, 2001 Pattern Classification , 2e, Wiley.
  5. Hand D , Mannila H, Smyth P. , Principles Of Data Mining, Eastern Economy Edition , Prentice Hall
  6. Osmar R. Za¨?ane,Luiza M. Antonie, Alexandru C "Mammography Classification by an Association Rulebased Classifier".
  7. Vijayakumar C. , Damayanti G. Chanda S. , Sreedhar C and Bhargava B, "Wavelet and Co-occurance Matrix Based Artificial Neural Network Tools for the Segmentation of Multiple sclerosis Lesions on MR Images", IEEE Trans. Signal and Image Processing, Vol. 1, 2006, pp. 333-337
  8. Pisano E. D,Yaffe M. J, Kuzmiak C. M, "Digital Mammography",Lippincott Williams & Wilkins ,2004
  9. Vasantha M . Bharathi S. , "Classifications of Mammogram Images using HybridFeatures", Euro Journals Publishing, 2011.
  10. Nalini S. ,Mohapatra A. G, Gurukalyan K. , "Breast Cancer MassDetection in Mammograms using K-means and Fuzzy C-means Clustering", International Journal of computer applications, 2011.
  11. Vibha L, Harshavardhan G M, Pranaw K, P DeepaShenoy, Venugopal K R, L M Patnaik,"Classification of Mammograms Using Decision Trees", 10th International database Engineering and Applications symposiums (IDEAS ' 06), 2006
  12. T. N. C. I. W. site. Available: www. cancer. gov
Index Terms

Computer Science
Information Sciences

Keywords

Breast Cancer Classifier Digital Mammography Feature Extraction Segmentation