ACR Meeting Abstracts

ACR Meeting Abstracts

  • Meetings
    • ACR Convergence 2024
    • ACR Convergence 2023
    • 2023 ACR/ARP PRSYM
    • ACR Convergence 2022
    • ACR Convergence 2021
    • ACR Convergence 2020
    • 2020 ACR/ARP PRSYM
    • 2019 ACR/ARP Annual Meeting
    • 2018-2009 Meetings
    • Download Abstracts
  • Keyword Index
  • Advanced Search
  • Your Favorites
    • Favorites
    • Login
    • View and print all favorites
    • Clear all your favorites
  • ACR Meetings

Abstract Number: 0152

Deep Learning-Derived Chest Radiograph Scores in COVID-19 in Rheumatic Disease Patients versus General Population Comparators

Naomi Patel1, Kristin D'Silva1, Matthew Li1, Tiffany Hsu2, Michael Di Iorio3, Xiaoqing Fu1, Claire Cook4, Lauren Prisco5, Lily Martin6, Kathleen Vanni6, Alessandra Zaccardelli6, Yuqing Zhang7, Jayashree Kalpathy-Cramer1, Jeffrey Sparks6 and Zachary Wallace8, 1Massachusetts General Hospital, Boston, MA, 2Brigham and Women's Hospital, Jamaica Plain, MA, 3Brigham and Women's Hospital, Brookline, MA, 4Division of Rheumatology, Allergy, and Immunology, Mass General Hospital, Boston, MA, 5Brigham and Women's Hospital, Pound Ridge, MA, 6Brigham and Women's Hospital, Boston, MA, 7Massachusetts General Hospital, Quincy, MA, 8Massachusetts General Hospital, Newton, MA

Meeting: ACR Convergence 2021

Keywords: autoimmune diseases, COVID-19, Imaging, pulmonary

  • Tweet
  • Click to email a link to a friend (Opens in new window) Email
  • Click to print (Opens in new window) Print
Session Information

Date: Saturday, November 6, 2021

Title: Imaging of Rheumatic Diseases Poster (0149–0182)

Session Type: Poster Session A

Session Time: 8:30AM-10:30AM

Background/Purpose: Patients with rheumatic diseases and COVID-19 may have higher risk of mechanical ventilation than those without rheumatic diseases. We compared COVID-19 lung disease between rheumatic disease patients and general population comparators using a deep learning algorithm that extracts a quantitative measure of radiographic lung disease severity.

Methods: We performed a comparative cohort study of patients with rheumatic disease who had confirmed COVID-19 and at least one chest radiograph within ±2 weeks of COVID-19 diagnosis between January 31, 2020 and January 31, 2021 in a large healthcare system and comparators (matched up to 10:1 by age, sex, date of diagnosis, and chest radiograph location). Covariates and COVID-19 outcomes (mechanical ventilation and death) were extracted from a centralized data warehouse. A previously validated convolutional Siamese neural network algorithm was used to calculate the Pulmonary X-Ray Severity (PXS) score, a quantitative measure of COVID-19 lung severity. Higher scores indicate more severe pulmonary disease (Figure 1). We compared the PXS scores between rheumatic diseases cases and comparators using quantile regression at the 50th and 85th percentiles. We then evaluated the association of severe PXS score ( >9) with risk of mechanical ventilation and death among patients with rheumatic diseases using logistic regression.

Results: We identified 70 rheumatic disease patients and 463 general population comparators. Rheumatic disease patients had more comorbidities such as hypertension (54% vs. 32%, p< 0.001), diabetes (31% vs. 15%, p< 0.001), and interstitial lung disease (6% vs. 1%, p< 0.01) (Table 1). The most common rheumatic diseases were rheumatoid arthritis (37%), other inflammatory arthritis (17%), and systemic lupus erythematosus (14%). The maximum PXS score within 2 weeks of COVID-19 diagnosis was similar in the rheumatic disease patients and comparators at the 50th percentile (3.17 vs. 3.13, respectively, p=0.10) but significantly higher among rheumatic disease patients at the 85th percentile (9.54 vs. 8.02, p=0.03) (Table 2). Rheumatic disease patients were more likely to have a maximum PXS score >9 (20% vs. 11%, p=0.02), indicating severe pulmonary disease. Among rheumatic disease patients, those with maximum PXS score >9 more frequently required mechanical ventilation (85% vs. 5%, p< 0.0001) and died (14% vs. 2%, p=0.08) compared to those with PXS score ≤9.

Conclusion: Using a previously validated deep learning algorithm, patients with rheumatic disease and COVID-19 had more severe radiographic lung disease than matched comparators. Among rheumatic disease patients, more severe radiographic disease was associated with higher risk of mechanical ventilation. Future studies are needed to determine if PXS scores may help to risk-stratify patients with rheumatic disease and COVID-19.

Figure 1. Examples of chest radiographs from patients with COVID_19 and their associated Pulmonary X-ray Severity (PXS) scores.

Table 1. Clinical characteristics of COVID_19 patients with rheumatic disease and comparators without rheumatic disease who had chest radiographs performed.

Table 2. PXS scores from chest radiographs in rheumatic disease patients and matched comparators, using PXS scores from within 2 weeks before or after COVID_19 diagnosis.


Disclosures: N. Patel, None; K. D'Silva, None; M. Li, None; T. Hsu, None; M. Di Iorio, None; X. Fu, None; C. Cook, None; L. Prisco, None; L. Martin, None; K. Vanni, None; A. Zaccardelli, None; Y. Zhang, None; J. Kalpathy-Cramer, GE Healthcare, 5, AWS, 12, Non-financial support, Genentech Foundation, 5; J. Sparks, Bristol-Myers Squibb, 2, 5, Amgen, 5, Gilead, 2, Inova, 2, Janssen, 2, Optum, 2, Pfizer, 2; Z. Wallace, Bristol-Myers Squibb, 5, Principia/Sanofi, 5, Viela Bio, 2, MedPace, 2.

To cite this abstract in AMA style:

Patel N, D'Silva K, Li M, Hsu T, Di Iorio M, Fu X, Cook C, Prisco L, Martin L, Vanni K, Zaccardelli A, Zhang Y, Kalpathy-Cramer J, Sparks J, Wallace Z. Deep Learning-Derived Chest Radiograph Scores in COVID-19 in Rheumatic Disease Patients versus General Population Comparators [abstract]. Arthritis Rheumatol. 2021; 73 (suppl 9). https://acrabstracts.org/abstract/deep-learning-derived-chest-radiograph-scores-in-covid-19-in-rheumatic-disease-patients-versus-general-population-comparators/. Accessed .
  • Tweet
  • Click to email a link to a friend (Opens in new window) Email
  • Click to print (Opens in new window) Print

« Back to ACR Convergence 2021

ACR Meeting Abstracts - https://acrabstracts.org/abstract/deep-learning-derived-chest-radiograph-scores-in-covid-19-in-rheumatic-disease-patients-versus-general-population-comparators/

Advanced Search

Your Favorites

You can save and print a list of your favorite abstracts during your browser session by clicking the “Favorite” button at the bottom of any abstract. View your favorites »

All abstracts accepted to ACR Convergence are under media embargo once the ACR has notified presenters of their abstract’s acceptance. They may be presented at other meetings or published as manuscripts after this time but should not be discussed in non-scholarly venues or outlets. The following embargo policies are strictly enforced by the ACR.

Accepted abstracts are made available to the public online in advance of the meeting and are published in a special online supplement of our scientific journal, Arthritis & Rheumatology. Information contained in those abstracts may not be released until the abstracts appear online. In an exception to the media embargo, academic institutions, private organizations, and companies with products whose value may be influenced by information contained in an abstract may issue a press release to coincide with the availability of an ACR abstract on the ACR website. However, the ACR continues to require that information that goes beyond that contained in the abstract (e.g., discussion of the abstract done as part of editorial news coverage) is under media embargo until 10:00 AM ET on November 14, 2024. Journalists with access to embargoed information cannot release articles or editorial news coverage before this time. Editorial news coverage is considered original articles/videos developed by employed journalists to report facts, commentary, and subject matter expert quotes in a narrative form using a variety of sources (e.g., research, announcements, press releases, events, etc.).

Violation of this policy may result in the abstract being withdrawn from the meeting and other measures deemed appropriate. Authors are responsible for notifying colleagues, institutions, communications firms, and all other stakeholders related to the development or promotion of the abstract about this policy. If you have questions about the ACR abstract embargo policy, please contact ACR abstracts staff at [email protected].

Wiley

  • Online Journal
  • Privacy Policy
  • Permissions Policies
  • Cookie Preferences

© Copyright 2025 American College of Rheumatology