Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Domain Shift Analysis in Chest Radiographs Classification in a Veterans Healthcare Administration Population

Journal Article · · Journal of Imaging Informatics in Medicine
 [1];  [1];  [2];  [3];  [3];  [1];  [4];  [4];  [4];  [4];  [5];  [1];  [5];  [5];  [5];  [6];  [4];  [2];  [4];  [7] more »;  [8] « less
  1. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
  2. Thermo-Fisher Scientific, Waltham, MA (United States)
  3. Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
  4. Department of Veterans Affairs, West Haven, CT (United States)
  5. VA Boston Healthcare System, Boston, MA (United States). Million Veteran Program Boston Coordinating Center
  6. Stanford Univ., CA (United States)
  7. NVIDIA Corporation, Santa Clara, CA (United States)
  8. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States); Vanderbilt University Medical Center, Nashville, TN (United States)
+ Show Author Affiliations
This study aims to assess the impact of domain shift on chest X-ray classification accuracy and to analyze the influence of ground truth label quality and demographic factors such as age group, sex, and study year. We used a DenseNet121 model pre-trained MIMIC-CXR dataset for deep learning-based multi-label classification using ground truth labels from radiology reports extracted using the CheXpert and CheXbert Labeler. We compared the performance of the 14 chest X-ray labels on the MIMIC-CXR and Veterans Healthcare Administration chest X-ray dataset (VA-CXR). The validation of ground truth and the assessment of multi-label classification performance across various NLP extraction tools revealed that the VA-CXR dataset exhibited lower disagreement rates than the MIMIC-CXR datasets. Additionally, there were notable differences in AUC scores between models utilizing CheXpert and CheXbert. When evaluating multi-label classification performance across different datasets, minimal domain shift was observed in the unseen VA dataset, except for the label “Enlarged Cardiomediastinum.” The subgroup with the most significant variations in multi-label classification performance was study year. These findings underscore the importance of considering domain shift in chest X-ray classification tasks, paying particular attention to the temporality of the exam. Our study reveals the significant impact of domain shift and demographic factors on chest X-ray classification, emphasizing the need for improved transfer learning and robust model development. Addressing these challenges is crucial for advancing medical imaging research and improving patient care.
Research Organization:
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
Sponsoring Organization:
USDOE
Grant/Contract Number:
AC05-00OR22725
OSTI ID:
2573574
Journal Information:
Journal of Imaging Informatics in Medicine, Journal Name: Journal of Imaging Informatics in Medicine; ISSN 2948-2933
Publisher:
Springer Science and Business Media LLCCopyright Statement
Country of Publication:
United States
Language:
English

References (17)

Drawbacks of Artificial Intelligence and Their Potential Solutions in the Healthcare Sector journal February 2023
AI recognition of patient race in medical imaging: a modelling study journal June 2022
Expert-level detection of pathologies from unannotated chest X-ray images via self-supervised learning journal September 2022
Foundation models for generalist medical artificial intelligence journal April 2023
MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports journal December 2019
The Current and Future State of AI Interpretation of Medical Images journal May 2023
Densely Connected Convolutional Networks conference July 2017
What Makes Transfer Learning Work for Medical Images: Feature Reuse & Other Factors conference June 2022
Chest radiographs and machine learning – Past, present and future journal June 2021
Effect of image resolution on automated classification of chest X-rays journal August 2023
Radiologist shortage leaves patient care at risk, warns royal college journal October 2017
CheXclusion: Fairness gaps in deep chest X-ray classifiers conference November 2020
CheXpert: A Large Chest Radiograph Dataset with Uncertainty Labels and Expert Comparison journal July 2019
Combining Automatic Labelers and Expert Annotations for Accurate Radiology Report Labeling Using BERT conference January 2020
An Extensive Survey of Machine Learning Based Approaches on Automated Pathology Detection in Chest X-Rays conference January 2021
Chest X-ray in Emergency Radiology: What Artificial Intelligence Applications Are Available? journal January 2023
Cross Dataset Analysis of Domain Shift in CXR Lung Region Detection journal March 2023

Similar Records

Effect of image resolution on automated classification of chest X-rays
Journal Article · Thu Aug 03 20:00:00 EDT 2023 · Journal of Medical Imaging · OSTI ID:1996738
Validating deep learning inference during chest X-ray classification for COVID-19 screening
Journal Article · Sun Aug 08 20:00:00 EDT 2021 · Scientific Reports · OSTI ID:1812409

Related Subjects

chest X-ray image classification
domain shift
multi-label classification