|
Activity Number:
|
528
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #304222 |
|
Title:
|
An Algorithm for Selecting Cases to Evaluate a Digital Mammography System
|
|
Author(s):
|
David C. Hoaglin*+ and Lynn Fletcher-Heath
|
|
Companies:
|
Abt Bio-Pharma Solutions, Inc. and Carestream Health, Inc.
|
|
Address:
|
181 Spring Street, Lexington, MA, 02421,
|
|
Keywords:
|
Enriched reader study
|
|
Abstract:
|
The U.S. Food and Drug Administration recommends that a comparison stress test, against a previously approved system, be applied to a new digital mammography imaging system. A challenging set of mammograms may be assembled by specifying a total number of cases and a target distribution of cases over the categories of each characteristic. In the pool of available cases, each case belongs to a cell in the cross-classification of the characteristics, and their joint distribution may make it difficult to satisfy the targets on all of the individual characteristics. Given a set of weights on each characteristic that represent the relative difficulty of completing the categories, we suggest a measure of the distance between the marginal distributions and the corresponding targets. We then describe an algorithm that selects cases by minimizing the distance measure.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |
