RPS 1704 - Advanced chest imaging

Author Block: T. Viana Miranda Lima, N. Saltybaeva, T. Gassenmaier, L. Ebner, J. E. Roos; Lucerne/CH
Purpose: The COVID-19 pandemic has significantly impacted healthcare practices worldwide, including the use of CT in chest examinations. This study is aimed to analyse the trends in CT chest examinations before, during, and after the COVID-19 pandemic.
Methods or Background: Data was retrospectively collected from 10 public hospitals over a period pre-pandemic years (2019 until February 2020), through the pandemic (March 2020- April 2023) (as defined by the World Health Organisation) and post-pandemic (May 2023- December 2024). Patient exposure information from more than 240 000 CT examinations was collected and analysed using a commercial dose management system. Statistical analysis was performed with descriptive and inferential statistics employed to compare the number of CT chest examinations and patient radiation exposure across different periods.
Results or Findings: The results indicate a increase in CT chest examinations during the pandemic, which reached a plateau and remained stable post-pandemic (p-value < 0.001). Importantly, the average radiation exposure per patient has decreased with the evolution of technology. There was also a shift between protocols used during the pandemic with the move towards more dedicated procedures.
Conclusion: These findings highlight the sustained demand for CT chest examinations and the effective management of patient radiation exposure during and after the pandemic. The study underscores those possible practices obtained during the pandemic became the normal after the end of the pandemic.
Limitations: This study has few limitations: firstly, the data are derived from DICOM data, which, while ensuring reproducibility and objectivity, may not capture all clinical nuances for example a patient being imaged on the wrong protocol. Secondly, the analysis of radiation exposure was conducted primarily to demonstrate overall trends in dose levels rather than to thoroughly investigate or attribute the main factors underlying the observed changes.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics was waived for this project by the North- and Central Switzerland Ethics Commission (Req-2025-00111), and all patient data was anonymised.

Author Block: F. Villone, L. Calandriello, G. Cicchetti, R. D'Abronzo, F. Scrocca, F. Del Prete, G. Sgalla, L. Natale, A. R. Larici; Roma/IT
Purpose: Prognostic assessment in interstitial lung disease (ILD) is challenging, due to unpredictable behaviour and scarcity of validated progression markers. Clinico-functional progression criteria are limited by subjectivity and variability, while radiological progression relies on visual assessment, affected by inter-observer variability. Artificial intelligence(AI) on high-resolution Computed Tomography(HRCT) has emerged as a reliable tool for ILD assessment.
This study evaluated deep-learning HRCT analysis with commercially available software in assessing progression and mortality in a real-life prospective ILD patients cohort.
Methods or Background: HRCT scans from ILD patients (idiopathic pulmonary fibrosis [IPF] and non-IPF ILDs) at baseline and after 12±4 months were analyzed using AVIEW software (Coreline Soft). Correlation among baseline AI variables and baseline forced vital capacity (FVC) categories(≥80%; 79-50%; ≤49%) and between AI changes and FVC decline categories(≥10%; 5-9%; ≤4%) were assessed. Multivariate logistic regression tested correlation of AI variables with all-cause mortality and functional progression. Integration of quantitative data into clinico-functional models to predict mortality was evaluated with C-index.
Results or Findings: 148 HRCT scans from 74 ILD patients were assessed (mean age 73.3±7.8 yrs; male 61.3%;IPF 68.8%). Thirty-nine patients(52.7%) progressed and twenty-four(32.4%) died. At baseline, most AI variables inversely correlate with the FVC categories; changes in the quantitative measurements correlates with the categorial FVC decline, showing responsiveness. At multivariate analysis, 19 variables at baseline were independently predictive of mortality; at follow-up, changes in 5 variables predicted mortality, with the best predictor being Δ% of consolidation(HR 4.01, 95%CI 1.69-10.13, p=0.003). Incorporating AI variables into clinico-functional models improved mortality prediction at baseline (C-index 0.707 vs 0.57; p=0.012) and follow-up(0.866 vs 0.763; p=0.005).
Conclusion: Deep-learning-based HRCT measurements and longitudinal changes predict progression and mortality in fibrotic ILD. Composite AI clinico-functional models may allow more precise prognostic assessment .
Limitations: Single-center study
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study protocol was approved by the local institutional commmittee on research ethics.

Author Block: H. Zhang¹, L. Xu²; ¹Joondalup/CN, ²Dazhou/CN
Purpose: To delineate the microradiological features of ultra-early-stage lung squamous cell carcinoma (UESLSCC) and investigate hypotheses on growth dynamics, vascular supply, and early morphological evolution between central- and peripheral-type lesions.
Methods or Background: A retrospective study included 38 patients with pathologically confirmed UESLSCC identified via opportunistic screening or health examinations. Pre-diagnostic thin-section CT was reviewed, with microradiological post-processing (MPR, CPR, MIP, VB) to enhance visualization. Lesions were classified by anatomical location (central- vs peripheral-type). Recognition intervals, morphological subtypes, microradiological signs (tail sign, microvascular convergence), and relationships with bronchovascular structures were evaluated by experienced thoracic radiologists. The study was approved by the institutional ethics committee, and informed consent was waived due to its retrospective design.
Results or Findings: Central-type lesions had significantly shorter recognition intervals than peripheral-type (P < 0.05), plausibly due to earlier respiratory symptoms, earlier bronchoscopy, and/or larger lesion volumes. We hypothesize that central lesions—supported by richer bronchial arterial flow and larger airway caliber—proliferate and present earlier, whereas peripheral lesions—reliant on pulmonary circulation in narrower distal airways—grow more slowly. Morphologically, wall-type and luminal-type patterns predominated in central lesions, while non-specific alveolar-type and pleural-type patterns were more common peripherally, suggesting an alveolar origin. Peripheral lesions frequently arose near bronchial bifurcations and showed tail signs, indicating bidirectional growth along adjacent bronchioles. Microradiological post-processing significantly improved detection and characterization of ultra-early lesions, especially subtle bronchovascular relationships.
Conclusion: Ultra-early–stage lung squamous cell carcinoma predominantly originates from the bronchial wall. Microradiological imaging is pivotal for detecting and characterizing these lesions, revealing distinct vascular supply patterns and growth dynamics between central and peripheral types. These findings propose new pathophysiological hypotheses and provide a foundation for early screening, near-pathological preoperative diagnosis, and tailored management of UESLSCC.
Limitations: Single-center, retrospective design, a small opportunistically screened cohort.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: H. Zhang, Y. Meng; Xuzhou/CN
Purpose: To evaluate the accuracy, radiation dose, and clinical feasibility of a high-pitch low-dose chest CT (Sa36LDCT) protocol with calcium-aware reconstruction for coronary artery calcium scoring (CACS).
Methods or Background: This prospective study included 90 patients underwent both standard CAC scans and Sa36LDCT. CACS was quantified as Agatston score, calcium volume, and equivalent mass. Risk stratification was assessed using Agatston categories. Agreement between protocols was evaluated using ICC, Bland–Altman analysis, and weighted kappa. Subgroup analyses were performed according to heart rate (≤75 vs >75 bpm). Radiation dose and subjective image quality were compared.
Results or Findings: CAC metrics showed no significant difference between protocols, with excellent agreement (ICC=0.983-0.996). Subgroup analysis revealed that heart rate didn't significantly influence CAC quantification or risk classification with Agatston score ICC values of 0.98 for ≤75 bpm and 0.99 for >75 bpm, demonstrating accuracy in both subgroups. Radiation dose reduction was consistently observed across both subgroups, with Sa36LDCT reducing effective radiation dose by 62% (0.74 vs 0.28 mSv, P<0.001). Risk stratification showed near-perfect consistency across heart rate subgroups.
Conclusion: High-pitch Sa36LDCT provides accurate CACS equivalent to standard CACS with radiation dose reduction and preserved image quality. The protocol's robustness across heart rate subgroups supports its clinical use as a safe and practical one-stop approach for combined pulmonary and cardiovascular screening.
Limitations: First, it was conducted at a single center with a relatively limited sample size. Second, calcium-aware Sa36 reconstruction kernel is manufacturer-specific, and our results may not be directly applicable to scanners from other vendors or reconstruction algorithms with different technical implementations. Third, we did not specifically investigate the influence of BMI on calcium quantification.
Funding for this study: We acknowledge financial support from the Jiangsu Traditional Chinese Medicine Science and Technology Development Plan Project (MS2021100), and the Key Research and Development Program of Xuzhou Science and Technology Bureau (KC20159).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This prospective study was approved by the Institutional Review Board of our institution (XYFY2025-KL037)

Author Block: E. Verelst¹, S. Bayat², S. Verbanck¹, G. Van Gompel¹, J. De Mey¹, N. Buls¹; ¹Brussels/BE, ²Grenoble/FR
Purpose: To investigate regional ventilation in healthy and pathological in vivo rabbit lungs using Xenon (Xe)-enhanced spectral CT-based ventilation maps.
Methods or Background: Six anesthetized rabbits [2.6 ± 0.2 (SD) kg] were mechanically ventilated. Dynamic multi-breath imaging was performed during Xe washout using a clinical DECT (Revolution, GE Healthcare) before and after the administration 40 mg/mL Methacholine (MCh) to induce bronchoconstriction (post-MCh). Xe concentration (Xeconc) was monitored using a mass spectrometer (LR6000, Logan Research). The dynamic DECT data was decomposed into Xe densities images. Ventilation maps were generated depicting specific ventilation (sV), quantified as the time constant (min-1) of a fitted mono-exponential model, following image registration. Agreement between global sV derived from ventilation maps (sV_image) and mass spectrometry data (sV_ms) was assessed. Ventilation defect fraction (VDF) was computed as voxels with sV below 25% of the mean and compared between baseline and post-MCh. Regional ventilation was evaluated by comparing the mean sV within the VD-area (sV_VDF) compared to the non-VD area (sV_non-VDF).
Results or Findings: The ventilation maps revealed differences in ventilation distribution between baseline and post-MCh conditions. VDF increased from 22.6% ± 10.6% at baseline, to 37.2% ± 9.8% post-MCh, p = 0.003. At baseline and post-MCh, sV_VDF (6.6 ± 9.92 min-1, 5.27 ± 8.38 min-1) was reduced compared to sV_non-VDF (0.29 ± 0.09 min-1, 0.22 ± 0.03 min-1), p < 0.001. Lastly, a strong agreement was found between global sV_image and sV_ms at baseline (3.0 min-1 ± [-0.77 min-1, 6.7 min-1]) and post-MCh (3.8 min-1 ± [-1.15 min-1, 8.8 min-1]).
Conclusion: This study shows the potential of dynamic Xe-enhanced spectral CT-imaging for quantitative studies of ventilation defects, supporting its utility for regional assessments of pulmonary function.
Limitations: This study was an in vivo animal study.
Funding for this study: Flemish Research Foundation (FWO) personal grant, number: 1SH1Z24N.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical approval obtained by the Ethical committee on Animal T esting of the Vrije Universiteit
Brussel (18-272-15).

Author Block: O. Weinheimer¹, S. Riepenhausen², R. Tenardi-Wenge², M. A. Mall³, H-U. Kauczor¹, H. Omran², M. O. Wielpütz⁴, J. Raidt²; ¹Heidelberg/DE, ²Muenster/DE, ³Berlin/DE, ⁴Greifswald/DE
Purpose: Primary Ciliary Dyskinesia (PCD) is a rare genetic disorder where tiny hair-like structures called cilia don’t work properly. Various quantitative CT (QCT) techniques have been proven to deliver useful and objective biomarkers describing lung parenchyma and airways. This study explores which QCT biomarkers best characterize lung and airway abnormalities in PCD.
Methods or Background: A total of 148 inspiratory CT scans from 106 genetically confirmed PCD subjects (55 female, mean age 29.2 ± 17.7 years), collected from multiple medical centers, were automatically analyzed using YACTA software. 59 CT scans from 35 individuals with situs inversus were mirrored to enable standard YACTA evaluation. Quantitative CT parameters of the airways and lung parenchyma were assessed on a six-lobe basis.
Results or Findings: Mirroring the CT scans allowed for the inclusion of all situs inversus subjects in the analysis. Airway wall percentage (AWP) was significantly higher in the lower lobes compared to the upper lobes (60.54% vs 55.89%, p<2.2e-16). Similarly, the bronchiectasis index (BEI) was elevated in the lower lobes (1.64 vs 0.45, p<2.2e-16). Mean lung density (MLD) and the 75th percentile of the lung histogram were also higher in the lower lobes (–729.39 HU vs –763.45 HU, p=1.95e-11 and –854.42 HU vs –892.74 HU, p=3.79e-10). The values in the right middle lobe and lingula were comparable to those in the lower lobes. AWP, BEI and MLD showed a moderate correlation with subject age (R=-0.57, p=8.91e-14 and R=0.27, p<0.001 and R=-0.47, p=2.85e-09).
Conclusion: QCT showed significant changes in airway and lung parenchyma paramaters predominantly in the lower lobes, differing from the typical pattern seen in cystic fibrosis, where changes are distributed across all lobes. QCT may be a valuable tool for PCD research.
Limitations: Retrospective study, no uniform CT protocol.
Funding for this study: DFG OM6/7, OM6/8, OM6/10, OM6/14, OM6/16, CRU 326 (subprojects OM6/11 (H. Omran), RA3522/1 (J. Raidt)) IZKF Muenster
(Om2/010/20; OM2/014/24), BMFTR - project ReproTrackMS (grant 01GR2303). Part of the authors are Healthcare Professionals in the European Reference Network ERN LUNG.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of the Medical Association of Westphalia-Lippe and the local ethics committee of the University of Münster (Münster, Germany; AZ 2011-270-f-S).

Author Block: E. Muscogiuri, F. De Keyzer, W. De Wever, L. Hardy, G. Aerts, R. Quarck, M. Delcroix, T. Verbelen, L. Godinas; Leuven/BE
Purpose: Chronic Thromboembolic Pulmonary Hypertension (CTEPH) is a rare form of pulmonary hypertension, with pulmonary endarterectomy (PEA) being the treatment for suitable candidates. This study investigates the prognostic value of quantitative imaging biomarkers from Computed Tomography Pulmonary Angiography (CTPA) in patients undergoing PEA.
Methods or Background: We retrospectively enrolled CTEPH patients who underwent CTPA within 6 months pre-PEA, and right heart catheterization (RHC) pre- and post-PEA. CTPA scans were analyzed with a dedicated software (Aview, Coreline Soft, Inc) to extract BV5 and BV10 (representing total blood volume of pulmonary vessels with cross-sectional areas <5mm2 and <10mm2, respectively), also indexing them on total blood (TBV) and lung volume (respectively BV5/TBV, BV10/TBV; ρBV5, ρBV10). Significant associations (p<0.05) with RHC parameters, including their percentage change (Δ%), were evaluated using Spearman’s rank correlation.
Results or Findings: Thirty-three patients (16 men, mean age 60.2 years) were included. Pre-operatively, BV5/TBV and BV10 negatively correlated with cardiac output (CO) and cardiac index (CI). Post-operative analysis revealed a negative correlation between BV5/TBV and the RHC parameters mean pulmonary artery pressure (mPAP), right atrial pressure (RAP), and pulmonary artery wedge pressure (PAWP), with r=-0.39, -0.43, -0.55 respectively. A negative correlation was also observed between ρBV5 and both post-operative pulmonary vascular resistance (PVR) and percentage change (Δ%) in PVR (r=-0.38 for both), highlighting the association between increased peripheral involvement and a reduced improvement of vascular resistance.
Conclusion: CTPA-derived quantitative biomarkers show that a lesser involvement of the peripheral vasculature results in a better post-PEA outcome. The extent of peripheral disease, quantified with CTPA, correlates with greater reductions in pulmonary pressures and vascular resistances, and improved cardiac function post-PEA, supporting the prognostic value of these quantitative imaging metrics.
Limitations: Small study population with possible selection bias.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study has been approved by the institutional IRB (S70797).

Author Block: K. Z. Lim, E. Yang, B. Jong, K. K-P. Lau; Melbourne/AU
Purpose: Deep-learning motion correction algorithm (DLMCA) (CLEAR MOTION, Canon Medical Systems) works by creating a 4D motion vector-map from raw CT data to compensate for coronary artery movement during reconstruction and corrects motion artifacts without increasing radiation dose. Reducing motion/pulsation artifacts in CT pulmonary angiography (CTPA) is essential in evaluating pulmonary arteries (PA) and the adjacent ascending aorta (Ao) and aortic root (Ar). We aimed to investigate the efficacy of DLMCA in CTPA in image quality (IQ) improvement.
Methods or Background: All adult consecutive CTPAs over 3-month period in a specialist cardiac-lung hospital were included. Images of each CTPA were reconstructed with standard deep-learning image-reconstruction (SDLIM) and DLMCA. Two radiologists independently and randomly assessed the image quality using Likert 5-point scale of PA, Ao and Ar. Interobserver agreement with Kappa score was determined. Objective IQ using signal-to-noise (SNR) and contrast-to-noise ratio (CNR) were calculated. Quantitatively, the degrees of motion artifact (the width of double-shadow) at PA and Ao were assessed. Statistical analysis with paired t-test and Wilcoxon signed-rank test was undertaken (p-value <0.05 as significant).
Results or Findings: 100 CTPAs (44% male, mean age of 65.5 years) were included. There was substantial to excellent interobserver agreement with kappa-score of 0.73 for SDLIM and 0.81 for DLMCA. Significant improvement in subjective image clarity with all three structures and significant reduction of double shadow sign of ascending aorta (mean 1.9mm difference, p<0.0001) and pulmonary trunk (mean 1.2mm difference, p<0.0001) were achieved. However, mean SNR and CNR for both aorta and pulmonary trunk were reduced with DLMCA (p<0.0001).
Conclusion: DLMCA has benefit of reducing motion artifact of PA, Ao and Ar in CTPA, and therefore, improving image clarity, despite some degradation of SNR and CNR.
Limitations: Single centre retrospective study.
Funding for this study: None.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: