By Topic

Classification of breast tumors in mammograms using a neural network: utilization of selected features

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

6 Author(s)
Du-Yih Tsai ; Dept. of Electr. Eng., Gifu Nat. Coll. of Technol., Japan ; Fujita, H. ; Horita, K. ; Endo, T.
more authors

An artificial neural network approach to classification of possible tumors into benign and malignant ones in mammograms was developed earlier by the authors (1993), and an average of 75% recognition rate was shown. In this paper the authors propose a different method for improving the recognition rate. After removing the nonuniform background trend superimposed on mammographic images, the authors calculate two selected features, the standard deviation and entropy of an image, representing the texture of the image. The two texture features are then used as input to the neural network. A set of 30 breast images consisting of 20 benign and 20 malignant tumor patterns are used for investigation. The result is achieved with 100% accuracy, an improvement of 25% over the authors' previous result. The performance of the proposed method is considered to be comparable to several-year experienced radiologists. This encouraging result indicates that this method may be useful for classification of benign and malignant tumors in mammograms.

Published in:

Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on  (Volume:1 )

Date of Conference:

25-29 Oct. 1993