Harnessing clinical annotations to improve deep learning performance in prostate segmentation

Karthik V. Sarma, Alex G. Raman, Nikhil J. Dhinagar, Alan M. Priester, Stephanie Harmon, Thomas Sanford, Sherif Mehralivand, Baris Turkbey, Leonard S. Marks, Steven S. Raman, William Speier & Corey W. Arnold
Purpose Developing large-scale datasets with research-quality annotations is challenging due to the high cost of refining clinically generated markup into high precision annotations. We evaluated the direct use of a large dataset with only clinically generated annotations in development of high-performance segmentation models for small research-quality challenge datasets. Materials and methods We used a large retrospective dataset from our institution comprised of 1,620 clinically generated segmentations, and two challenge datasets (PROMISE12: 50 patients, ProstateX-2: 99...

Registration Year

  • 2021
    1

Resource Types

  • Dataset
    1

Affiliations

  • National Institutes of Health
    1
  • SUNY Upstate Medical University
    1
  • University of California Los Angeles
    1