Skip to content

Repository files navigation

Task-based-sampling

MIDRC TDP 3d, https://www.midrc.org/technology-development-projects/tdp3

Development Team: Natalie Baughan, PhD; Heather Whitney, PhD; Kyle Myers, PhD; Maryellen Giger, PhD

Task based sampling begins with the identification of cases relevant for a specific task and target population demographic characteristics (such as age range, disease status, and imaging modality). Then, optimized quota sampling is conducted by randomly sampling cases until the maximum category margin (Baughan et al. 2022) is less than a pre-specified value. N. Baughan et al., “Task-Based Sampling of the MIDRC Sequestered Data Commons for Algorithm Performance Evaluation,” presented at Annual Meeting of the American Association of Physicists in Medicine, 2022, E257–E258).

This code has been developed and tested using MacOS 14.1.1, Matlab R2023b, and Microsoft Excel for Mac.

About

Code to conduct task-based sampling.

Resources

Stars

0 stars

Watchers

4 watching

Forks

Packages

Contributors

Languages