Mammographic Mass Detection using Wavelets as Input to Neural Networks

Authors:
Niyazi Kilic;Pelin Gorgel;Osman N. Ucan;Ahmet Sertbas
Affiliations:
Electrical&Electronics Department, The Faculty of Engineering, Istanbul University, Istanbul, Turkey 34320;Computer Science Department, The Faculty of Engineering, Istanbul University, Istanbul, Turkey 34320;Electrical&Electronics Department, The Faculty of Engineering, Istanbul University, Istanbul, Turkey 34320;Computer Science Department, The Faculty of Engineering, Istanbul University, Istanbul, Turkey 34320
Venue:
Journal of Medical Systems
Year:
2010

Citing 7
Cited 1

Ten lectures on wavelets

Ten lectures on wavelets
Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
A review of evidence of health benefit from artificial neural networks in medical intervention

Neural Networks
Analysis of mammogram classification using a wavelet transform decomposition

Pattern Recognition Letters - Special issue: Sibgrapi 2001
Detection of Microcalcifications in Digital Mammograms Images Using Wavelet Transform

CERMA '06 Proceedings of the Electronics, Robotics and Automotive Mechanics Conference - Volume 02
Predicting Metastasis in Breast Cancer: Comparing a Decision Tree with Domain Experts

Journal of Medical Systems
A novel soft cluster neural network for the classification of suspicious areas in digital mammograms

Pattern Recognition

High performance concrete compressive strength forecasting using ensemble models based on discrete wavelet transform

Engineering Applications of Artificial Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

The objective of this paper is to demonstrate the utility of artificial neural networks, in combination with wavelet transforms for the detection of mammogram masses as malign or benign. A total of 45 patients who had breast masses in their mammography were enrolled in the study. The neural network was trained on the wavelet based feature vectors extracted from the mammogram masses for both benign and malign data. Therefore, in this study, Multilayer ANN was trained with the Backpropagation, Conjugate Gradient and Levenberg---Marquardt algorithms and ten-fold cross validation procedure was used. A satisfying sensitivity percentage of 89.2% was achieved with Levenberg---Marquardt algorithm. Since, this algorithm combines the best features of the Gauss---Newton technique and the other steepest-descent algorithms and thus it reaches desired results very fast.