Title: Artificial neural network as classification method for prostate cancer incidence

Authors: Khaled S. Alqahtani; Shankar Srinivasan; Dinesh P. Mital

Addresses: Scientific Research Center, Umm Al-Qura University, Makkah Al-Mukarramah, KSA ' Department of Health Informatics, Rutgers Biomedical and Health Sciences, Newark, NJ 07107-3001, USA ' Department of Health Informatics, Rutgers Biomedical and Health Sciences, Newark, NJ 07107-3001, USA

Abstract: This study uses artificial neural network (ANN) classification method to predict the occurrence of prostate cancer. The surveillance, epidemiology, and end results (SEER) program data, for the years 2004 to 2011, was used to train and test the performance of ANN. In addition, ANN model's result was compared with the most used classification methods. The result of classification methods' tests showed that the artificial neural network had success rates of 99.36% compared to logistic regression (84.95%), decision tree classifier (90.34%) and support vector machine (88.52%).

Keywords: prostate cancer; prostate-specific antigen; artificial neural networks; ANNs; logistic regression; decision tree classifier; support vector machines; SVM; cancer prediction.

DOI: 10.1504/IJMEI.2017.080925

International Journal of Medical Engineering and Informatics, 2017 Vol.9 No.1, pp.61 - 72

Available online: 09 Nov 2016 *

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