Deep learning‐based lung volume estimation with dynamic chest radiography
Journal of Applied Clinical Medical Physics · 2026
DOI: 10.1002/acm2.70487 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied Clinical Medical Physics, 2026. Verfasst von Nozomi Ishihara und 4 weiteren Autor:innen. Beteiligte Einrichtung: College of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan. Hauptthemen laut MEDLINE: Deep Learning, Radiography, Thoracic, Tomography, X-Ray Computed, Lung. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Dynamic chest radiography (DCR) is a recently developed low‐dose pulmonary functional imaging method that can be performed in a general X‐ray room. DCR provides sequential images during respiration, and the measured changes in lung area are a promising diagnostic indicator of lung function.
Purpose
To investigate lung volume estimation using deep learning from DCR images during respiration and evaluate its accuracy in comparison with previously proposed estimation methods.
Methods
Two convolutional neural networks (CNNs), VGG19 and DenseNet121, were trained using DCR image datasets from 257 patients, with reference lung volumes derived from corresponding computed tomography (CT) images. The performance of the models was evaluated using mean absolute error (MAE) and mean absolute percentage error (MAPE), and compared against that of a conventional linear regression model. Correlation between the estimated and reference lung volumes was assessed using Pearson's correlation coefficient ( r ) and the degrees‐of‐freedom‐adjusted coefficient of determination ( Rf 2 ). Forced vital capacity (FVC) was also estimated by subtracting the lung volume at maximum exhalation from that at maximum inhalation.
Results
The VGG19 and DenseNet121 models demonstrated superior performance in estimating whole lung volume (combined right and left lung) compared to the linear regression method. Specifically, MAE was 373/376 mL, MAPE was 8.1%/7.9%, r was 0.88/0.90, and Rf 2 was 0.76/0.80 for VGG19/DenseNet121, respectively. In contrast, the linear regression model yielded an MAE of 568 mL, MAPE of 12.4%, r of 0.84, and Rf 2 of 0.69. Although the Rf 2 values for DCR‐derived FVC using VGG19 and DenseNet121 indicated moderate correlation, the MAE and MAPE were relatively high at 1.3/1.4 L and 41.1%/47.0%, respectively.
Conclusion
The proposed deep learning‐based approach for lung volume estimation from DCR images outperformed the conventional linear regression method. Further improvements in CNN model architecture and the incorporation of guided forced respiratory maneuvers may enhance the potential for image‐based pulmonary function testing.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- College of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan.
- Department of Respiratory Medicine, Kanazawa University, Kanazawa, Ishikawa, Japan.
- Department of Thoracic Surgery, Kanazawa University Hospital, Kanazawa, Ishikawa, Japan.
ORCID-Profile: Rie Tanaka
Zitieren
APA
Ishihara, N., Tanaka, R., Kikuno, H., Ohkura, N., & Matsumoto, I. (2026). Deep learning‐based lung volume estimation with dynamic chest radiography. Journal of Applied Clinical Medical Physics. https://doi.org/10.1002/acm2.70487
Vancouver
Ishihara N, Tanaka R, Kikuno H, Ohkura N, Matsumoto I. Deep learning‐based lung volume estimation with dynamic chest radiography. Journal of Applied Clinical Medical Physics. 2026. doi:10.1002/acm2.70487
BibTeX
@article{Ishihara2026,
author = {Ishihara, Nozomi and Tanaka, Rie and Kikuno, Haruto and Ohkura, Noriyuki and Matsumoto, Isao},
title = {Deep learning‐based lung volume estimation with dynamic chest radiography},
journal = {Journal of Applied Clinical Medical Physics},
year = {2026},
doi = {10.1002/acm2.70487}
}
EUVIMED