Author: Robinson, Caleb; Trivedi, Anusua; Blazes, Marian; Ortiz, Anthony; Desbiens, Jocelyn; Gupta, Sunil; Dodhia, Rahul; Bhatraju, Pavan K.; Liles, W. Conrad; Lee, Aaron; Kalpathy-Cramer, Jayashree; Ferres, Juan M. Lavista
Title: Deep learning models for COVID-19 chest x-ray classification: Preventing shortcut learning using feature disentanglement Cord-id: sr4tr5rb Document date: 2021_2_13
ID: sr4tr5rb
Snippet: In response to the COVID-19 global pandemic, recent research has proposed creating deep learning based models that use chest radiographs (CXRs) in a variety of clinical tasks to help manage the crisis. However, the size of existing datasets of CXRs from COVID-19+ patients are relatively small, and researchers often pool CXR data from multiple sources, for example, using different x-ray machines in various patient populations under different clinical scenarios. Deep learning models trained on suc
Document: In response to the COVID-19 global pandemic, recent research has proposed creating deep learning based models that use chest radiographs (CXRs) in a variety of clinical tasks to help manage the crisis. However, the size of existing datasets of CXRs from COVID-19+ patients are relatively small, and researchers often pool CXR data from multiple sources, for example, using different x-ray machines in various patient populations under different clinical scenarios. Deep learning models trained on such datasets have been shown to overfit to erroneous features instead of learning pulmonary characteristics – a phenomenon known as shortcut learning. We propose adding feature disentanglement to the training process, forcing the models to identify pulmonary features from the images while penalizing them for learning features that can discriminate between the original datasets that the images come from. We find that models trained in this way indeed have better generalization performance on unseen data; in the best case we found that it improved AUC by 0.13 on held out data. We further find that this outperforms masking out non-lung parts of the CXRs and performing histogram equalization, both of which are recently proposed methods for removing biases in CXR datasets.
Search related documents:
Co phrase search for related documents- Try single phrases listed below for: 1
Co phrase search for related documents, hyperlinks ordered by date