RPS 1415 - Hot Topic: photon counting CT

Author Block: R. Li¹, N. He¹, Y. Cui², Z. Xu¹, H. Jiang³, F. Yan¹, S. Haller⁴; ¹Shanghai/CN, ²Nanjing/CN, ³Shagnhai/CN, ⁴Geneve/CH
Purpose: To evaluate the preliminary diagnostic performance and utility of UHR-PCCTA and standard resolution (SR) PCCTA compared to immediate post-deployment DSA following FD treatment, with emphasis on comprehensive characterization of post-FD follow-up.
Methods or Background: This prospective study consecutively enrolled 47 participants treated with FD and subsequently underwent PCCTA with UHR and SR reconstructions between August 2023 and August 2024. Two readers independently analyzed images blindly, including aneurysm size, occlusion status, neck coverage, wall apposition, stent deformation, branch vessel with Likert-scale (1-5). The inter-rater reliability for both UHR and SR-PCCTA was assessed using the intraclass correlation coefficient (ICC) and weighted Cohen’s kappa. The diagnostic performance of PCCTA within 1 month after FD was compared with that of immediate post-deployment DSA using a confusion matrix.
Results or Findings: The study included 47 participants with 57 intracranial aneurysms treated with 51 FD stents. UHR-PCCTA exhibited higher diagnostic confidence (median=5, IQRs=5-5) compared to SR-PCCTA (median=4, IQRs=4-5, P<0.05). UHR-PCCTA demonstrated a trend toward higher inter-rater agreement (κ range:0.814-1.0; ICC range:0.902-0.960) relative to SR-PCCTA (κ range:0.668-1.0; ICC range:0.873-0.966) among both aneurysm size and stent morphology. Compared with immediate post-deployment DSA, UHR-PCCTA showed excellent diagnostic accuracy for aneurysm neck coverage (93.3%), wall apposition (100.0%), stent deformation (93.8%), and branch vessel inclusion (93.3%), along with very-good accuracy for aneurysm occlusion (88.2%).
Conclusion: UHR-PCCTA demonstrates high reliability and diagnostic accuracy in the assessment of intracranial aneurysms following FD treatment. Its comprehensive imaging evaluation capability has the potential to inform clinical decision-making.
Limitations: (1) relatively small sample size owing to the preliminary clinical nature; (2) potential changes during the ≤1-month interval between PCCTA and DSA, may affect accuracy assessments; (3) Whether the sub-stent hypodensity detected by UHR-PCCTA suggested neointima requires histopathological confirmation.
Funding for this study: No
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This prospective single-center study was approved by the local review board (KY2023-186) and written informed consent was obtained from all participants.

Author Block: F. M. E. Pinckaers, T. Flohr, B. A. J. M. Wagemans, I. Huijberts, J. E. Wildberger, A. Postma; Maastricht/NL
Purpose: To assess quality improvements in ultra-high resolution (UHR) photon counting detector (PCD) CT angiography (CTA) of the cerebral vasculature when moving to image acquisition at lower kVp and image reconstruction with sharper kernels.
Methods or Background: Over a period of 25 months, all cerebral CTA imaging on PCD-CT was evaluated. Records were excluded in case of protocol deviations, severe motion artifacts or cerebral circulatory arrest. Using UHR resolution (0.2 mm), three protocols were evaluated: (1) 140 kVp/medium sharp kernel (Hv40); (2) 90 kVp/Hv40; (3) 90 kVp/very sharp kernel (Hv72). Vessel attenuation, signal-to-noise ratios (SNR) and contrast-to-noise ratios (CNR) were derived at four locations in the anterior circulation. Vessel sharpness was quantified using the edge rise distance and edge rise slope. Subjective assessments of image noise, vessel attenuation and vessel sharpness were performed by two experienced readers on a 5-point Likert scale.
Results or Findings: Out of 154 consecutive records, 141 (92%) were included. Vessel attenuation, SNR and CNR improved with image acquisition at 90 kVp compared to 140 kVp in the protocols using a medium sharp reconstruction kernel. SNR and CNR decreased when applying (very) sharp reconstruction kernels at 90 kVp due to an increase in noise. However, while an evident decrease in SNR and CNR was observed in smaller calibre vessels when using a medium sharp kernel at 90 kVp, these parameters remained relatively stable when using (very) sharp kernels. Vessel sharpness was markedly improved in the 90 kVp/(very) sharp kernel protocol. Subjective assessments of image quality also favoured the 90 kVp/(very) sharp kernel protocol.
Conclusion: Image quality of UHR PCD-CTA of the cerebral vasculature is improved with image acquisition at 90 kVp and image reconstruction with (very) sharp kernels.
Limitations: Single-centre, retrospective study.
Funding for this study: None.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The medical ethical committee of the Maastricht University Medical Centre waived the need for individual patient consent.

Author Block: F. Xu, Y. Wang, D. Mu; Shanghai/CN
Purpose: Carotid plaque composition is a major determinant of plaque vulnerability and thromboembolic stroke. This study investigated the performance of dual-layer spectral CT angiography (DLCTA) parameters to characterize carotid atherosclerotic plaques components, with high-resolution MRI (HR-MRI) or histopathology of postoperative specimens as reference standard.
Methods or Background: Fifteen patients with moderate-to-severe carotid bifurcation stenosis were prospectively enrolled. All patients underwent both DLCTA and HR-MRI, with findings correlated with histopathology in surgical cases. Plaque components including intraplaque hemorrhage (IPH), lipid-rich necrotic core (LRNC), fibrous tissue, and calcifications were identified on MRI or histopathology, and mapped to CTA using slice-to-slice coregistration. Quantitative parameters recorded included conventional CT (con-CT) values, virtual monoenergetic images (VMI, 40-140keV, 10-keV intervals), effective atomic number (Z-eff), and electron density (ED) . Diagnostic performance was evaluated by ROC analysis.
Results or Findings: A total of 42 slices and 164 ROIs were analyzed. Calcifications were clearly distinguished by conventional CT values, while non-calcified components (LRNC, IPH, and fibrous tissue) showed significant differences on VMI40keV and Z-eff (LRNC: 25.5 ± 16.1 HU, 7.19 ± 0.08; IPH: 122.3 ± 40.5 HU, 7.66 ± 0.43; fibrous tissue: 66.4 ± 20.8 HU, 7.40 ± 0.09; all p < 0.001). VMI40keV and Z-eff demonstrated the highest accuracy for differentiating LRNC from other non-calcified components (AUC 0.99, 95% CI 0.97–1.00 and 0.96, 0.83–1.00, respectively). For distinguishing IPH from fibrous tissue, AUC was 0.82 for VMI40keV.
Conclusion: DLCTA, particularly VMI40keV and Z-eff, enables noninvasive differentiation of carotid plaque components and may offer a practical alternative to contrast-enhanced MRI for identifying vulnerable carotid plaques, with potential implications for stroke risk stratification
Limitations: Study limited by small sample size and limited histopathologic confirmation.
Funding for this study: Not applicable.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: Z. Li¹, X. Fan¹, M. Lyu², F. Feng¹; ¹Beijing/CN, ²Shanghai/CN
Purpose: To evaluate the feasibility and performance of photon-counting CT (PCCT) with black-blood technique for carotid plaque characterization, by comparing with vessel wall MRI (VW-MRI) and biphasic CT angiography (CTA).
Methods or Background: From January 2024 to May 2025, a total of twenty-five patients scheduled for carotid revascularization underwent CTA with arterial and delayed-phase using PCCT and eight patients voluntarily received vessel wall MRI (VW-MRI) within 2 weeks. Black-blood CT (BBCT) images was generated for each CTA phases. Totally 50 bilateral vessels were included for qualitative analysis between BBCT and biphasic CTA. And thirteen plaques were evaluated for quantitative analysis and plaque type classification. Qualitative scores (including calcified/non-calcified plaque visibility, diagnostic confidence, overall image quality) were independently assessed by two radiologists blindly using a 5-point scale between biphasic CTA and BBCT. Quantitative parameters (stenosis rate, plaque area, calcium area, non-calcium area) were compared between BBCT and VW-MRI using intraclass correlation coefficients (ICCs). Carotid Plaque-RADS evaluation was further applied in BBCT, biphasic CTA and VW-MRI.
Results or Findings: Compared with arterial and delayed images, BBCT achieved significantly higher scores in non-calcified plaque visibility, diagnostic confidence, and overall image quality (all p < 0.001). BBCT showed excellent agreement with VW-MRI, especially for non-calcified area (ICC = 0.96). Confusion matrix analysis illustrated that BBCT identified more high-risk plaques compared to biphasic CTA (such as intraplaque hemorrhage), and it exhibited strong concordance with VW-MRI.
Conclusion: BBCT enabled accurate carotid plaque quantification with excellent agreement to VW-MRI and outperforms biphasic CTA in image quality, diagnostic confidence, and high-risk plaque detection, suggesting its potential utility in stroke risk stratification when combined with Carotid Plaque-RADS.
Limitations: The number of carotid plaques with matched VW-MRI is small due to only single-center inclusion.
Funding for this study: None.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Medical Ethics Committee of the Peking Union Medical College Hospital (No.I-24PJ0346). Written consent was obtained from each subject.

Author Block: A. M. Hunkemöller, K. Höffler, K. Katsirntaki, F. Wacker, B. Wiegmann, H-O. Shin; Hannover/DE
Purpose: Photon-counting CT (PCCT) inherently provides spectral data, enabling reconstruction of virtual non-contrast (VNC) and perfused blood volume (PBV) images. However, quantitative PBV assessment remains challenging due to variability in lung inflation and iodine concentration. This study systematically evaluated normalization strategies to improve PBV reproducibility across different inflation states, contrast concentrations, and acquisition conditions in an ex-vivo lung model.
Methods or Background: Five explanted porcine lungs were perfused and ventilated on the Organ Care System (OCS™) Lung platform under three standardized inflation states (inspiration, mid-inspiration, expiration) and two contrast concentrations. PCCT datasets were reconstructed into VNC and PBV images. After manual segmentation, three normalization approaches were applied: (1) normalization to an intravascular reference region (PBV_rel), (2) further normalization to total lung volume (PBV_lung), and (3) adjustment for lung tissue fraction derived from VNC attenuation (PBV_VNC). The dependence of PBV on inflation state was analyzed using repeated-measures ANOVA.
Results or Findings: Unnormalized PBV demonstrated significant dependence on lung inflation (p = 0.0143; SD = 5.00 HU). Normalization to an intravascular reference (PBV_rel) effectively mitigated this effect (p = 0.9388; SD = 0.13). PBV_lung retained minor residual variation (p = 0.0292; SD = 0.11), while PBV_VNC yielded the most consistent results across inflation states (p = 0.0825; SD = 0.08), indicating superior robustness and reduced variability.
Conclusion: PBV quantification in PCCT is strongly affected by lung inflation and iodine concentration. Normalization substantially enhances measurement stability, with VNC-based tissue fraction adjustment (PBV_VNC) providing the most reliable and inflation-independent standardization for quantitative pulmonary perfusion assessment.
Limitations: This ex-vivo study employed a small sample size and controlled ventilation conditions, which may limit direct translation to in-vivo applications.
Funding for this study: No funding.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: S. Mohammadzadeh, I. Kiani, H. Ghorani, M. Saeed, I. Razeghian; Tehran/IR
Purpose: To evaluate and compare coronary stent imaging parameters using photon-counting computed tomography (PC-CT) and conventional energy-integrating detector computed tomography (EID-CT).
Methods or Background: Protocol pre-registration was performed a priori at PROSPERO (CRD42025642922). We searched PubMed, Web of Science, Embase, and Cochrane Library for studies until December, 2024. Risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies-Comparison (QUADAS-C). The imaging modalities were compared with Likert scores of image quality, image noise, and stent diameter using random effects pooling method.
Results or Findings: A total of 10 studies were included with 110 coronary stents. The PC-CT had a significantly higher pooled Likert score of 1.29 (95% CI = 0.41 to 2.17, p-value = 0.004) than EID-CT, indicating its higher image quality. Also, it demonstrated a pooled 18.5% (95% CI = 3.5% to 36.6%, p-value = 0.015), reduction in image noise in comparison to EID–CT. As for stent diameter assessment, the PC-CT revealed a pooled 17.4% (95% CI = 1.9% to 32.8%, p-value = 0.027) lower diameter length than real stent diameter evaluated by calipers. On the other hand, EID-CT revealed a greater discrepancy with 25.8% (95% CI = 16.1% to 35.5%, p-value <0.001) lower diameter length in comparison to real stent diameter indicating higher inaccuracy of the EID-CT in measuring stent sizes relative to PC-CT.
Conclusion: PC-CT is an emerging method for the coronary stents imaging, providing better image quality. PC-CT showed lower image noise and high stent diameter measurement accuracy compared to conventional EID-CT. Our findings support the incorporation of PC-CT into future clinical research to further exploration on its diagnostic efficacy.
Limitations: The primary problem arises from the varied methodologies and acquisition parameters marked by diverse scanner types, reconstruction parameters, phantom designs, and stent characteristics.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: Y. Zhou, Y. Guo, Z. Wang, L. Lei, X. Guo; Zhengzhou/CN
Purpose: This study aims to evaluate the diagnostic value of contrast-enhanced photon-counting CT (PCCT) in differentiating between bland thrombus and tumour thrombus in the portal venous system, with the goal of improving clinical diagnostic accuracy.
Methods or Background: This prospective study enrolled 100 patients with portal venous system thrombosis who underwent spectral contrast-enhanced CT using PCCT. 50 patients were confirmed with bland portal vein thrombosis and 50 with tumour thrombus. ROIs were drawn at the thrombus sites, and spectral post-processing (SPP) datasets were reconstructed to obtain conventional mixed-energy images (CI), virtual monoenergetic images (VMI, 40-150 keV ), iodine density maps (ID), and spectral curves. A multiparametric spectral model combining iodine density and VMI (ID-VMI40-150 keV) was established. For each ROI, conventional CT values (CIHU), iodine density values (ID value), and the slope of the spectral curve (λHU) were measured. Univariate logistic regression was performed to assess the predictive value of the spectral parameters in distinguishing between bland thrombus and tumour thrombus. ROC curves were generated, and the AUC and cutoff values were calculated to compare the diagnostic performance of each spectral parameter.
Results or Findings: Compared with unenhanced CT values (AUC: 0.767), arterial phase CT values (AUC: 0.899), venous phase CT values (AUC: 0.960; ), arterial iodine density (AUC: 0.891), and virtual monoenergetic images at 40 keV (AUC: 0.909), 50 keV (AUC: 0.909; cutoff: 0.744), and 60 keV (AUC: 0.913), the spectral model ID-VMI60 keV demonstrated the highest diagnostic performance for differentiating between bland thrombus and tumour thrombus in the portal vein (AUC: 0.983).
Conclusion: The spectral model offers superior diagnostic accuracy compared to conventional CT values and individual spectral parameters in distinguishing between bland thrombus and tumour thrombus in the portal venous system.
Limitations: Not Applicable.
Funding for this study: The Key Scientific Research Project of Colleges and Universities in Henan Province (20B320047)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study has been reviewed by the ethics committee

Author Block: M. T. Hagar¹, S. Hyska², M. Vecsey-Nagy³, J. I. Griggers¹, J. Osoria-Velasquez⁴, T. S. Emrich⁵, A. Varga-Szemes⁴; ¹Freiburg Im Breisgau/DE, ²Munich/DE, ³Budapest/HU, ⁴Charleston, SC/US, ⁵Mainz/DE
Purpose: Below-knee CT angiography (CTA) is limited by artifacts from calcified plaques in peripheral artery disease (PAD). Combined spectral ultrahigh-resolution (UHR) photon-counting CT (PCD-CT) may improve performance, but optimal conditions remain unclear.
Methods or Background: In this IRB-approved post-hoc study, consecutive patients with PAD undergoing clinically indicated PCD-CTA were included. Scans were performed on a dual-source PCD system in combined spectral and UHR mode (collimation 120 × 0.2 mm, 120 kVp). Axial images were reconstructed as down-sampled (DS, 0.8 mm, Bv40), virtual monoenergetic images (VMIs, 45–90 keV, 0.4 mm, Bv60), iodine maps (IM, 0.4 mm, Bv60), and UHR (0.2 mm, Bv60). Additional extremity-specific reconstructions with isotropic voxel spacing were generated (FOV 205 × 205 mm; UHRfocused: 0.2 mm, spectral: 0.4 mm). For small-caliber vessels (2.0 mm) with calcified plaques, perpendicular line profiles were analyzed to derive lumen and plaque widths using full-width-at-half-maximum (FWHM_lumen and FWHM_plaque). Digital subtraction angiography (DSA) served as reference, and diagnostic performance was assessed.
Results or Findings: A total of 59 patients (mean age 64.6 ± 13.5 years; 40 men, 68%) with 111 lower extremities were analyzed. Both UHR_focused and iodine map (IM_focused) reconstructions offered superior lumen visualization (FWHM_lumen: UHR 1.68 ± 0.76; IM 1.70 ± 0.76) and minimized blooming (FWHMplaque: UHR 1.01 ± 0.28; IM 0.98 ± 0.27; all p<0.01). In contrast, DS reconstructions were limited by blooming artifacts (FWHMlumen: 0.60 ± 0.78; FWHMplaque: 2.11 ± 0.60). UHR_focused and IM_focsued achieved highest sensitivity (93% (95% CI: 77–99%)), with UHRfocused additionally reaching the highest per-segment accuracy of 94% (95% CI: 83–99%, n=50).
Conclusion: UHR PCD-CT with per-extremity reconstruction using isotropic voxel spacing enhances image quality, improves lumen and plaque delineation, and yields highest diagnostic accuracy in below-knee CTA.
Limitations: The limited sample size requires confirmatory research.
Funding for this study: No
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: ID: Pro99133632, approval date: 4/25/2025