Improving risk grouping rules for prostate cancer patients using self-organising maps

Authors:
D. Schwartz;K. A. Smith;L. Churilov;M. Dally;R. Weber
Affiliations:
School of Business Systems, Monash University, Victoria 3800, Australia and Department of Industrial Engineering, Faculty of Physical Sciences and Mathematics, University of Chile, Republica 701, ...;School of Business Systems, Monash University, Victoria 3800, Australia;School of Business Systems, Monash University, Victoria 3800, Australia;William Buckland Radiotherapy Department, The Alfred Hospital, Victoria 3000, Australia;Department of Industrial Engineering, Faculty of Physical Sciences and Mathematics, University of Chile, Republica 701, Santiago, Chile
Venue:
Design and application of hybrid intelligent systems
Year:
2003

Citing 5
Cited 1

Self-organization and associative memory: 3rd edition

Self-organization and associative memory: 3rd edition
Data clustering: a review

ACM Computing Surveys (CSUR)
Visual Explorations in Finance

Visual Explorations in Finance
Mining Data from a Knowledge Management Perspective: An Application to Outcome Prediction in Patients with Resectable Hepatocellular Carcinoma

AIME '01 Proceedings of the 8th Conference on AI in Medicine in Europe: Artificial Intelligence Medicine
Machine learning for survival analysis: a case study on recurrence of prostate cancer

Artificial Intelligence in Medicine

Data Mining with Combined Use of Optimization Techniques and Self-Organizing Maps for Improving Risk Grouping Rules: Application to Prostate Cancer Patients

Journal of Management Information Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Accurate classification of prostate cancer patients into risk groups is important to assist in the identification of appropriate treatment paths. Current rules used to classify patients into low, intermediate and high risk groups have been developed by clinical experts using an evidence based approach. The available data from the Alfred Hospital is quite limited however, since records exist for only 258 patients that have been treated and followed up over a five-year period. Consequently, a data-driven rule generation approach is seen as inappropriate for such a limited sample size. Instead, we start with the existing rules and aim to improve accuracy by identifying inconsistencies utilising self-organising maps as a data visualisation tool. The improved classification rules have been able to increase both the quality of prediction and the homogeneity within the risk groups.