Activity Number:
|
193
- Modeling, Design Strategies and Assessments of Biomarkers
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #313250
|
|
Title:
|
Estimation and Construction of Confidence Intervals for the Cutoff-Points of Continuous Biomarkers in Trichotomous Settings
|
Author(s):
|
Brian Mosier* and Leonidas Bantis
|
Companies:
|
University of Kansas Medical Center and University of Kansas Medical Center
|
Keywords:
|
Biomarker;
ROC Surface;
Sensitivity;
Specificity;
Youden index
|
Abstract:
|
Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal forms of cancer with a 5-year survival rate between 5-7%. Thus, its early detection could be key to reducing PDAC mortality. Novel biomarkers are currently being studied for the early detection of PDAC. The ROC based Youden index is a popular method for choosing the optimal cutoff values of biomarkers and was recently generalized to the three-class case. We provide new parametric and non-parametric approaches that outperform the Youden based cutoffs in terms of variance and thus narrower confidence intervals are achieved. We further generalize our approaches to address k-class problems. Our approach is illustrated using real data from pancreatic cancer patients from a study conducted at the MD Anderson Cancer Center.
|
Authors who are presenting talks have a * after their name.