Page 101 - AIH-2-3

Artificial Intelligence in Health

ORIGINAL RESEARCH ARTICLE
Enhancing COVID-19 severity assessment with

artificial intelligence-based bone suppression
technique in chest radiography

1
2
Asumi Yamazaki , Masashi Seki , and Takayuki Ishida *
1
1 Division of Health Sciences, Osaka University Graduate School of Medicine, Suita, Osaka, Japan
2 Department of Radiology, Kitasato University Hospital, Sagamihara, Kanagawa, Japan

Abstract
Chest radiography (CXR) is widely used for initial respiratory assessment, but its
lesion detection capability is typically inferior to that of computed tomography.
Several studies have reported that artificial intelligence (AI)-based bone suppression
techniques can enhance the accuracy of lesion detection and disease classification.
Previously, we developed an AI-based bone suppression system based on dual-
energy subtraction principles. However, the subtraction process limited its versatility
and introduced significant artifacts. To overcome these challenges, we improved
the system to generate bone-suppressed images directly, eliminating the need for
subtraction. This study demonstrates the utility of the updated bone suppression
*Corresponding author: system as a pre-processing tool for regression analyses in assessing coronavirus
Takayuki Ishida disease 2019 severity. Four regression models – DenseNet, ResNet18, ResNet50, and
(tishida@sahs.med.osaka-u.ac.jp) RegNetY-120 – were employed to predict the severity based on scores annotated
Citation: Yamazaki A, Seki M, by radiologists. Except for DenseNet, all models showed statistically significant
Ishida T. Enhancing COVID-19 improvements in Pearson correlation coefficients (PCCs) when using bone-suppressed
severity assessment with
artificial intelligence-based bone images generated by the updated model. The highest PCC, 0.895, was achieved by
suppression technique in chest the ResNet18 model. The direct image generation process improved the clinical
radiography. Artif Intell Health. practicality of the bone suppression system while reducing artifacts. Furthermore,
2025;2(3):95-106.
doi: 10.36922/aih.5608 the significant improvement in linearity suggests that AI-driven bone suppression
enhances the visibility of abnormalities and improves the accuracy for pulmonary
Received: October 28, 2024
condition assessments. These advancements could expand the application of bone
1st revised: January 10, 2025 suppression techniques in various regression analyses, including disease severity,
2nd revised: February 11, 2025 progression, and recurrence risk. Nonetheless, further validation using larger and
more diverse datasets, as well as a broader range of prediction models, is necessary.
Accepted: February 21, 2025
Published online: March 5, 2025
Keywords: Bone suppression; Artificial intelligence; COVID-19; Regression model;
Copyright: © 2025 Author(s). Chest radiography
This is an Open-Access article
distributed under the terms of the
Creative Commons Attribution
License, permitting distribution,
and reproduction in any medium, 1. Introduction
provided the original work is
properly cited. Coronavirus disease 2019 (COVID-19) emerged in late 2019, rapidly developing into a
1
Publisher’s Note: AccScience global pandemic that overwhelmed healthcare systems due to its high contagiousness,
Publishing remains neutral with unprecedented morbidity, and mortality rates. Since then, with the spread of various
2,3
regard to jurisdictional claims in 4
published maps and institutional mutant strains, over seven million deaths have been reported worldwide. While severe
affiliations. cases of COVID-19 can lead to death, many patients remain asymptomatic or experience

Volume 2 Issue 3 (2025) 95 doi: 10.36922/aih.5608