Recommended Citation
Zhang R, Griner D, Garrett JW, Qi Z, and Chen GH. Training certified detectives to track down the intrinsic shortcuts in COVID-19 chest x-ray data sets. Sci Rep 2023; 13(1):12690.
Document Type
Article
Publication Date
8-4-2023
Publication Title
Sci Rep
Abstract
Deep learning faces a significant challenge wherein the trained models often underperform when used with external test data sets. This issue has been attributed to spurious correlations between irrelevant features in the input data and corresponding labels. This study uses the classification of COVID-19 from chest x-ray radiographs as an example to demonstrate that the image contrast and sharpness, which are characteristics of a chest radiograph dependent on data acquisition systems and imaging parameters, can be intrinsic shortcuts that impair the model's generalizability. The study proposes training certified shortcut detective models that meet a set of qualification criteria which can then identify these intrinsic shortcuts in a curated data set.
Medical Subject Headings
Humans; COVID-19; Radiography, Thoracic; Deep Learning; X-Rays; Radiographic Image Interpretation, Computer-Assisted
PubMed ID
37542078
Volume
13
Issue
1
First Page
12690
Last Page
12690
