Diagnostic Performance of Different AI Tools for Chest Radiography: A CT-Anchored Per-Abnormality Analysis in a Tertiary Referral Center.
- 1Department of Radiology, Faculty of Medicine, Hacettepe University, Ankara 06100, Turkey.
- 2Department of Radiology, School of Medicine and Public Health, University of Wisconsin-Madison, Madison, WI 53705, USA.
- 3Department of Chest Diseases, Faculty of Medicine, Hacettepe University, Ankara 06100, Turkey.
Abstract
Background/Objectives: Artificial intelligence (AI) increasingly supports chest radiograph interpretation, but per-abnormality comparisons of commercial tools using CT-anchored reference standards remain limited. We compared two commercial AI tools using a CT-anchored, CXR-targeted reference framework. Methods: We retrospectively reviewed chest radiographs obtained between 30 October 2023 and 10 October 2024. Each radiograph was paired with CT performed within 14 days or within 2 days for rapidly progressive findings. Seven abnormalities across five analysis categories were evaluated: fracture, nodule, pleural effusion, pneumothorax, and airspace disease (edema, atelectasis, opacity/consolidation). Radiologists first confirmed abnormalities on CT then assessed their visibility on the paired radiograph. Outputs from vendors A and B were recorded. Sensitivity was calculated for CT-confirmed, radiographically visible abnormalities; specificity, precision, and accuracy were calculated within the same detectability framework. Vendors were compared using McNemar's test. Results: The cohort included 1707 radiographs from 1680 patients (mean age, 60.3 ± 16.5 years; 50.5% male). Of 3573 CT-confirmed abnormalities, 1981 (55.4%) were radiographically visible and formed the primary reference-positive set. Vendor A had higher sensitivity for fractures, nodules, and pleural effusions (all p < 0.001), whereas vendor B had higher specificity for these findings (all p < 0.05). For pneumothorax, vendor B showed numerically higher sensitivity but lower specificity and accuracy; case numbers were limited, and precision was low for both tools. No significant differences were observed for harmonized airspace disease. Overall specificity was high, while sensitivity and precision varied across abnormalities. Conclusions: Both tools showed abnormality-specific strengths and limitations, supporting tailored implementation and real-world clinical validation.
Related
- Diagnostic performance with sensitivity/specificity optimized computer-aided detection to detect common abnormalities in chest radiography
- Diagnostic performance of circulating microRNAs for the early diagnosis of different types of acute cerebrovascular disease
- Clinical Diagnosis Versus Plain Radiography for Diabetic Foot Osteomyelitis: A Prospective Study of Diagnostic Agreement in a Resource‐Limited Settings
- Beyond overall accuracy: subgroup performance and threshold calibration of AI-assisted chest radiography for equity-oriented pulmonary tuberculosis screening.
- Wealth-anchored category validation: A diagnostic for categorical orderings in fragile-state household surveys
BibTeX
@article{duzgun2026diagnostic,
title = {{Diagnostic Performance of Different AI Tools for Chest Radiography: A CT-Anchored Per-Abnormality Analysis in a Tertiary Referral Center.}},
author = {Selin Ardalı Düzgün and İlke Taşçı and Atahan Tamtürk and Nuri Saraç and Melih Karadağ and Mustafa Ege Şeker and Deniz Köksal and Sevinç Sarınç and Meltem Gülsün Akpınar and Figen Başaran Demirkazık and Gamze Durhan},
journal = {Diagnostics (Basel, Switzerland)},
year = {2026},
month = sep,
doi = {10.3390/diagnostics16193139},
url = {https://europepmc.org/article/MED/42882295},
note = {CC BY},
}
RIS
TY - JOUR
TI - Diagnostic Performance of Different AI Tools for Chest Radiography: A CT-Anchored Per-Abnormality Analysis in a Tertiary Referral Center.
AU - Selin Ardalı Düzgün
AU - İlke Taşçı
AU - Atahan Tamtürk
AU - Nuri Saraç
AU - Melih Karadağ
AU - Mustafa Ege Şeker
AU - Deniz Köksal
AU - Sevinç Sarınç
AU - Meltem Gülsün Akpınar
AU - Figen Başaran Demirkazık
AU - Gamze Durhan
PY - 2026
DA - 2026/09/27
JO - Diagnostics (Basel, Switzerland)
DO - 10.3390/diagnostics16193139
UR - https://europepmc.org/article/MED/42882295
AB - Background/Objectives: Artificial intelligence (AI) increasingly supports chest radiograph interpretation, but per-abnormality comparisons of commercial tools using CT-anchored reference standards remain limited. We compared two commercial AI tools using a CT-anchored, CXR-targeted reference framework. Methods: We retrospectively reviewed chest radiographs obtained between 30 October 2023 and 10 October 2024. Each radiograph was paired with CT performed within 14 days or within 2 days for rapidly progressive findings. Seven abnormalities across five analysis categories were evaluated: fracture, nodule, pleural effusion, pneumothorax, and airspace disease (edema, atelectasis, opacity/consolidation). Radiologists first confirmed abnormalities on CT then assessed their visibility on the paired radiograph. Outputs from vendors A and B were recorded. Sensitivity was calculated for CT-confirmed, radiographically visible abnormalities; specificity, precision, and accuracy were calculated within the same detectability framework. Vendors were compared using McNemar's test. Results: The cohort included 1707 radiographs from 1680 patients (mean age, 60.3 ± 16.5 years; 50.5% male). Of 3573 CT-confirmed abnormalities, 1981 (55.4%) were radiographically visible and formed the primary reference-positive set. Vendor A had higher sensitivity for fractures, nodules, and pleural effusions (all p < 0.001), whereas vendor B had higher specificity for these findings (all p < 0.05). For pneumothorax, vendor B showed numerically higher sensitivity but lower specificity and accuracy; case numbers were limited, and precision was low for both tools. No significant differences were observed for harmonized airspace disease. Overall specificity was high, while sensitivity and precision varied across abnormalities. Conclusions: Both tools showed abnormality-specific strengths and limitations, supporting tailored implementation and real-world clinical validation.
SN - 2075-4418
ER -