Developing and Implementing the AJCC Prognostic System for Breast Cancer.

Abstract

In the past staging' systems provided a simple, easily understood ordering of patient' outcomes. For over thirty years breast cancer outcome prediction has been based on the TNM staging system. There are two problems with staging systems generally, and specifically with the TNM system: (1) they are not very accurate, i.e., their predictions are not close to the true outcomes), and (2) their accuracy can not be substantially improved because additional predictive factors can not be included in the system without increasing the system's complexity to the point where it is not longer useful to the clinician. The objective of this research is to replace with current TNM stage system with a new prognostic system that is inherently more accurate than the current system and that can integrate new prognostic factors to further improve prognostic accuracy. There are three components to - accomplishing this objective, which are the goals of this research project: (1) the development of the prognostic model itself, (2) the creation of the prognostic system by training the model with breast cancer outcome data, and (3) the computer-based implementation of the system for clinicians and tumor registries (clinical decision support system).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1999
Accession Number
ADA367434

Entities

People

  • Philip H. Goodman

Organizations

  • University of Nevada, Reno

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Biomedical Research
  • Breast Cancer
  • Computational Science
  • Data Mining
  • Databases
  • Genetics
  • Health Services
  • Information Processing
  • Information Science
  • Information Systems
  • Medical Personnel
  • Network Science
  • Oncology
  • Pain
  • Surveys

Readers

  • Oncology and Biomarker-Based Cancer Detection.
  • Software Engineering.
  • Systems Analysis and Design