Abstract Details
Activity Number:
|
606
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 7, 2014 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #313069
|
View Presentation
|
Title:
|
Statistical Methodology for Multiclass Classifications Applications to Dementia
|
Author(s):
|
Hong Li*+ and Sue Leurgans
|
Companies:
|
Rush UMC and Rush UMC
|
Keywords:
|
Alzheimer's Disease ;
Multiclass Classification ;
Diagnostic Accuracy ;
Optimal Cut-off Points ;
Volume under the Surface ;
Biomarkers
|
Abstract:
|
Alzheimer's disease (AD) is common, devastating, and a heavy economic burden. Accelerated efforts to identify presymptomatic stages of AD and biomarkers to best classify disease are urgent needs. No current biomarkers can perfectly discriminate individuals into multiple disease categories of AD. Although many biomarkers for diagnosis and various features are being studied, we lack sophisticated statistical methods which can fully utilize biomarkers to classify AD accurately, thereby facilitating evaluation of putative markers both alone and in combination. We propose a modified forward selection procedure with a generalized additive model for variable selection and a method to construct optimal cutoff points to distinguish disease stages using Neyman-Pearson Lemma. We apply these to data from the Religious Orders Study to examine the feasibility of combining biomarkers easily collected. The methods proposed show higher diagnostic accuracy measured by volume under the surface and the correct classification rate comparing with the existing methods. These techniques therefore facilitate evaluations of biomarkers for conditions with intermediate, rather than binary classifications.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2014 program
|
2014 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Professional Development program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.