Skip to main content

Reproducibility of Lung Densitometry with Photon-counting Detector and Energy-integrating Detector CT: A Multi-Institutional Phantom Study

In brief

Photon-counting CT cuts lung-density measurement error to 3.7 HU in phantom tests

Across four centers, photon-counting CT measured lung phantom attenuation more accurately than energy-integrating CT, with a mean error of 3.7 HU versus 6.4 and 12.4 HU. Its measurements also varied less across body sizes, sites and dose levels, while matched protocols produced similar results across scanners. These phantom findings need confirmation in patients before their clinical impact is clear.

Journal
Radiology. Cardiothoracic imaging (Q1)
Published
1 December 2026
Study design
Unclassified
Evidence level
Level 5, Expert Opinion (CEBM 5)
Authors
Saman Sotoudeh-Paima, Jim O'Doherty, Stephen M Humphries, Jarron Atha, Marrissa J McIntosh, Jessica C Sieren, et al.
PMID
43018449
DOI
10.1148/ryct.250312

Why clinicians should know about it

  • Picked for Medical Physics (top studies of the week, 15 November 2026): Multicenter phantom study evaluates CT densitometry accuracy and reproducibility

Abstract

Purpose To identify accurate imaging protocols for energy-integrating detector (EID) CT and photon-counting detector (PCD) CT, to establish comparable protocols across systems, and to evaluate the reproducibility of lung densitometry across body sizes, sites, and dose levels. Materials and Methods A chest phantom and a multidiameter phantom were scanned across four academic medical centers with PCD CT and EID CT systems across three doses. PCD CT included standard-resolution and ultrahigh-resolution (UHR) acquisitions with low-energy threshold (T3D) and virtual monoenergetic image (VMI) reconstructions, whereas EID CT used filtered back projection. Soft and sharp body and quantitative kernels were evaluated. Optimal protocols minimized mean absolute error (MAE) from reference CT attenuation values. Comparable protocols were identified using the histogram similarity index (HSI) within the lung sample (chest phantom), and reproducibility was assessed using air-insert density measurements. Statistical analyses included a Gaussian linear mixed-effects model and nonparametric tests for reproducibility. Results PCD CT outperformed both EID CT systems (P < .05) across 7392 measurements, achieving a lower MAE (3.7 HU ± 3.7 [SD]) than EID CT1 (6.4 HU ± 4.0; P < .05) and EID CT2 (12.4 HU ± 9.7; P < .05). The optimal PCD CT protocol used the UHR mode, high-dose, standard-resolution VMI, and soft quantitative reconstructions. Protocol matching across scanners was feasible, yielding an HSI of 95.6% ± 1.2 under optimal settings. Compared with EID CT, PCD CT exhibited greater reproducibility across body sizes (<2.2 HU), across sites (<2.5 HU), across dose levels (<3.9 HU), and within dose levels (<2.5 HU). Conclusion PCD CT provided more accurate and reproducible lung densitometry while maintaining comparability with EID CT protocols. Keywords: Phantom Studies, CT, CT-Photon Counting, CT-Quantitative, Thorax, Lung, Chronic Obstructive Pulmonary Disease, Comparative Studies, Technology Assessment, Physics Supplemental material is available for this article. © RSNA, 2026.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.