Skip to Main Content
Skip Nav Destination
ASME Press Select Proceedings
Intelligent Engineering Systems Through Artificial Neural Networks, Volume 17Available to Purchase
Editor
C. H. Dagli
C. H. Dagli
Search for other works by this author on:
ISBN-10:
0791802655
No. of Pages:
650
Publisher:
ASME Press
Publication date:
2007

This paper evaluated the contribution of both magnetic resonance mammography (MRM) image and non-image (NI) features (breast cancer risk factors) to benign ∕ malignant diagnostic accuracy using a new data set. This paper focus was using a new Kernealized-Partial Least Squares (K-PLS) and EP-SVM paradigms for this evaluation. An important initial finding is that (only) the non-imag , non-linear mapping region of the feature to diagnostic solution space contains significant diagnostic information, with a resultant overall Az of 0.77. The MRI image features of mass size, mass shape and mass margin yielded an Az of 0.96, which is excellent diagnostic performance. The EP-SVM Az diagnostic results for the same basic data set and statistical cross validation methods were slightly lower because “outliers” were included in these analyses.

Abstract
Introduction
Data Description
Kernel Defined Feature Space
Evolutionary Programming
Results
Conclusions
Acknowledgements
References
This content is only available via PDF.
You do not currently have access to this chapter.

or Create an Account

Close Modal
Close Modal