Title: A subset polynomial neural networks approach for breast cancer diagnosis

Authors: T.J. O'Neill, Jack Penm, Jonathan Penm

Addresses: School of Finance and Applied Statistics, College of Business and Economics, Australian National University, Canberra ACT 0200, Australia. ' School of Finance and Applied Statistics, College of Business and Economics, Australian National University, Canberra ACT 0200, Australia. ' School of Finance and Applied Statistics, College of Business and Economics, Australian National University, Canberra ACT 0200, Australia

Abstract: Breast cancer is a very common and serious cancer for women that is diagnosed in one of every eight Australian women before the age of 85. The conventional method of breast cancer diagnosis is mammography. However, mammography has been reported to have poor diagnostic capability. In this paper we have used subset polynomial neural network techniques in conjunction with fine needle aspiration cytology to undertake this difficult task of predicting breast cancer. The successful findings indicate that adoption of NNs is likely to lead to increased survival of women with breast cancer, improved electronic healthcare, and enhanced quality of life.

Keywords: breast cancer; classification; subset polynomial neural networks; healthcare improvements; cancer diagnosis; Australia; fine needle aspiration cytology; electronic healthcare; e-health; quality of life.

DOI: 10.1504/IJEH.2007.014549

International Journal of Electronic Healthcare, 2007 Vol.3 No.3, pp.293 - 302

Published online: 16 Jul 2007 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article