Effect of BIRADS shape descriptors on breast cancer analysis
by B. Surendiran; P. Ramanathan; A. Vadivel
International Journal of Medical Engineering and Informatics (IJMEI), Vol. 7, No. 1, 2015

Abstract: This paper focuses on an approach for characterising the spatial structure of mammogram masses using various multimodal features. Experiments have been conducted on benchmark digital database for screening mammography (DDSM) database with 300 mammograms. According to Breast Imaging Reporting and Data System (BIRADS) spatial structure of mammogram masses can be discriminated using its shape, size and density properties. Various new geometrical shape, margin and texture features and DDSM descriptors are used to characterise the morphology of masses as either benign or malignant. The effect of DDSM descriptors on classification accuracy is analysed. The experimental results shows that all 20 features including DDSM descriptors produce better accuracy of 93.3& compared to 86.7& without using DDSM descriptors. The overall Wilk's lambda value is reduced when all 20 features are used for classification. The proposed features set using shape, texture and DDSM descriptors provide better (c/v) ratio compared to wavelet, Gabor and other features reported in the literature.

Online publication date: Tue, 09-Dec-2014

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Medical Engineering and Informatics (IJMEI):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com