Research Presentation Session: Vascular

RPS 1415 - Exploring imaging and interventional treatment of atherosclerotic lesions: fresh perspectives

March 1, 12:30 - 13:30 CET

7 min
Semi-automatic measurement of carotid intima-media thickness in atherosclerosis
Jiatong Xu, Shanghai / China
Author Block: L. Zhang, J. Xu, M. Chen; Shanghai/CN
Purpose: This study aims to demonstrate if semi-automatic measurement can improve the accuracy of carotid intima-media thickness (CIMT) .
Methods or Background: This study analysed the results of carotid ultrasound examinations of 1637 patients. During the complete cardiac cycle, three sonographers used semi-automatic measurement of the systolic and diastolic and manual measurement to measure CIMT. The relevant data of the patients was also collected, including age, gender, diabetes status, and other cardiovascular risk factors.
Results or Findings: For semi-automatic measurement of CIMT, the intraclass correlation coefficient (ICC) of the three-operator agreement analysis is 0.99 (95% confidence interval: 0.984-0.994. There is a statistical difference between the semi-automatic and manual measurement. There is no statistical difference between the left CIMT and the right CIMT measured by semi-automatic and manual measurement. Compared with the cerebrovascular event group, the bilateral semi-automatic CIMT values between the normal group and the cerebrovascular event group were statistically different.
Conclusion: Using the semi-automatic measurement of CIMT, the consistency between operators is better, and the repeatability is higher. It can replace the manual measurement of CIMT, provide a more accurate and stable inspection method, and better evaluate the degree of atherosclerosis.
Limitations: The study also has certain limitations. In addition to being subject to the inherent limitations of retrospective research, due to technical problems, it needs to use specific machine measurements. Since the samples of this study come from the routine examination of the hospital, a larger community study is needed at a later stage, to further prove the conclusion of this study.
Funding for this study: This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval was obtained from the Ethics Committee of Shanghai Tongren Hospital.
7 min
Accuracy of an artificial intelligence-based algorithm for fully automated detection of aortic wall calcifications in chest CT
Christopher Schuppert, Heidelberg / Germany
Author Block: C. Schuppert1, R. Saffar1, J. Sperl2, J. Oravcová3, M. Kreibich1, J. Weiß1, F. Bamberg1, M. Czerny1, C. L. Schlett1; 1Freiburg im Breisgau/DE, 2Erlangen/DE, 3Bratislava/SK
Purpose: This study aimed to assess the accuracy of an artificial intelligence-based algorithm for fully automated detection of aortic wall calcifications in chest CT with a special focus on the surgically relevant clamping zone in the upper ascending aorta.
Methods or Background: We retrospectively included 100 chest CT scans from 92 patients who were examined on a third-generation dual-source scanner. Subsamples comprised 47 scans with an aortic angiography using iodinated intravenous contrast media and ECG-gating and 53 unenhanced scans. A previously validated deep learning algorithm performed aortic landmark detection and aorta segmentation. The segmentation mask was divided into eight anatomic segments and aortic wall calcifications were detected using a mean-based HU thresholding, yielding a binary score for their presence. Algorithm parameters (calcium cluster size threshold, aortic mask dilatation) were varied to determine optimal performance. A visual rating served as a reference. Standard estimates of diagnostic accuracy and interrater agreement using Cohen’s kappa were calculated.
Results or Findings: Aortic wall calcifications were observed in 74% of the examinations with a prevalence of 27% to 70% by aorta segment. Using different parameter combinations, the algorithm provided results for 97% to 100% of the examinations and 95% to 99% of the combined aortic segments. The best-performing parameter combination for the presence of calcifications in the aortic clamping zone yielded a sensitivity of 93% and a specificity of 82%, with an area under the receiver operating characteristic curve of 0.874. Using these parameters, the interrater agreement ranged from κ 0.66 to 0.92 per segment.
Conclusion: Fully automated detection of aortic wall calcifications in chest CT performs with good accuracy. This includes the critical preoperative assessment of the aortic clamping zone.
Limitations: Intravenous contrast media and ECG-gating were not individually assessed as factors influencing algorithm performance, which is a limitation.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective single-centre study was approved by the local institutional review board.
7 min
Quantitative parameters from dual-layer detector spectral CT (DLS-CT) in distinguishing non-calcified plaques and chronic total occlusion (CTO) components in lower limb arteries
Li Zhou, Xi'an / China
Author Block: L. Zhou, N. Ding, Y. Cui, Y. Han, J. Yang, X. Zhang, P. Cao, X. Huang; Xi'an/CN
Purpose: This study aimed to explore the application value of quantitative parameters from dual-layer detector spectral CT (DLS-CT) in distinguishing non-calcified plaques and chronic total occlusion (CTO) components in lower limb arteries
Methods or Background: Peripheral arterial disease (PAD) patients with DLS-CT examinations were prospectively enrolled from October 2022 to February 2023. Non-calcified plaques with a stenosis degree > 50% and CTO lesions in the above-knee arteries were analysed. Quantitative parameters of conventional images, 40keV, 70keV, virtual non-contrast, iodine density, Z-effective (Zeff), electron density (ED), and slope of the energy spectrum curve (λHU, λHU = (CT40keV -CT70 keV)/-30) of the non-calcified plaques and CTO lesions were evaluated for statistical analysis.
Results or Findings: Ninety-three lesions in 24 patients were evaluated, consisting of 72 non-calcified plaques and 21 CTO lesions. Analysis of the results indicated that CT40keV, iodine density and Z-effective of CTO tended to decrease and λHU tended to increase compared with non-calcified plaques, with significant difference of all (P<0.05). The area under the curve of ROC (AUC) for CT40keV, iodine density, Z-effective, and λHU in distinguishing non-calcified plaques from CTO were 0.71 (95% CI 0.60-0.79), 0.74 (95% CI 0.52-0.725), 0.70 (95% CI 0.59-0.79), and 0.63 (95% CI 0.53-0.73), respectively. The AUC for the combination of these parameters was 0.87 (95% CI 0.727-0.929) for distinguishing non-calcified plaques from CTO.
Conclusion: DLS-CT quantitative parameters have the potential to identify the composition characteristics of non-calcified plaques and CTO, providing new imaging indicators for understanding the pathological changes and progression mechanism of PAD.
Limitations: No limitations were identified.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of the First Affiliated Hospital (ethics approval number: XJTUIAF2021LSY-223).
7 min
Voxel-based calcified plaque analysis: a comparison between a novel silicon-based photon-counting CT prototype, a conventional energy-integrating detector CT and µCT
Emma Verelst, Brussels / Belgium
Author Block: E. Verelst1, G. Van Gompel1, B. Keelson1, J. De Mey1, D. Crotty2, P. D. Deak3, N. Buls1; 1Brussels/BE, 2Cork/IE, 3Ittigen/CH
Purpose: The study aimed to investigate the accuracy of a second-generation silicon-based photon-counting CT (Si-PCCT) prototype in a voxel-wise evaluation of calcified plaques, compared to conventional energy-integrating detector CT (EIDCT) with high-resolution micro-CT as a reference.
Methods or Background: Three human-resected arteries were embedded in a 2% agar-water phantom (d=20 cm). Helical scans were acquired using a second-generation Si-PCCT prototype (GE Healthcare, Milwaukee, USA) and a conventional EIDCT system (GE Revolution, Milwaukee, USA) at similar scan parameters (120 kV, 40 mm collimation, 0.9 pitch, 1s rotation and 17 mGy CTDIvol). Images were reconstructed using a bone kernel, 1024 matrix, 150 mm field-of-view and slice thickness of 0.42 mm (Si-PCCT) and 0.63 mm (EIDCT). Micro-CT images were acquired with a benchtop system (X-cube, Molecubes) using a spiral high-resolution acquisition protocol at 0.075 mm xy-pixel size and 0.2 mm slice thickness. Plaques (n=7) were analysed by a voxel-wise computation of plaque volume (mm3), plaque burden (%) and calcified plaque volume (%), i.e. voxels with CT-values >1000 HU. Micro-CT images were used as a reference. The differences between the two systems were tested with a paired sample t-test.
Results or Findings: Overall, Si-PCCT allowed for more accurate volume measurements, as the mean error towards the reference volume (12.2 mm3) was significantly reduced when compared to EIDCT (27 mm3), p=0.03. The mean error towards the reference plaque burden was also found to be significantly lower for Si-PCCT (6.6%) when compared to EIDCT (11.4%), p<0.001. Si-PCCT also allowed for a significant increase in detection of heavily calcified regions (60.5%), when compared to EIDCT (38.6%), p=0.01.
Conclusion: This study demonstrates an improved spatial resolution of a second-generation Si-PCCT prototype, allowing for an improved voxel-wise analysis on plaque volume, burden and composition, when compared to conventional EIDCT.
Limitations: The study was an ex vivo phantom study.
Funding for this study: This study was funded by the Radiology Department of the University Hospital of Brussels (UZ Brussel).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical approval was obtained for the use of three human arteries (carotid, femoral and iliac).
7 min
Role of dual-energy computed tomography (DECT) in detection of carotid artery monosodium urate deposition in patients with gout
Muhammad Danish Sarfraz, Vancouver / Canada
Author Block: M. D. Sarfraz, L. Treanor, S. Masood, N. Murray, A. Sheikh, S. Nicolaou; Vancouver, BC/CA
Purpose: The study aimed to find out if a dual-energy CT can detect monosodium urate deposition in carotid arteries and whether the presence of monosodium urate crystals has any effect on atherosclerotic disease in terms of plaque volume.
Methods or Background: This is a retrospective study. All patients who underwent any dual-energy neck imaging (like carotid angiogram, neck soft tissues or cervical spine) from January 2015 to December 2022 were included.
All patients’ charts were rigorously reviewed to find gout patients and avoid confounding variables. The same number of healthy controls (without a clinical history of gout or other rheumatic diseases) were included during the same study period. These healthy controls were then age (within 5 years), sex and confounders-matched to gout patients.
DECT datasets were post-processed using gout application and a volumetric analysis of atherosclerotic plaques was performed using a calcium scoring application in matched gout patients with carotid monosodium deposition, gout patients without carotid monosodium deposition and control/non-gout patients.
Results or Findings: Out of a total 2157 patients who underwent dual-energy neck imaging during the study period, 85 were confirmed gout cases. From the 85 gout cases, two were excluded due to presence of streak artefacts from dentures, hence n=83.
Out of the 83, monosodium urate deposition was detected in the carotid arteries of 10 patients (12%).
None of the matched control patients demonstrated monosodium crystal deposition.
Volumetric analysis of atherosclerotic plaques demonstrated larger plaque volumes in gout patients with monosodium urate deposition (n=10) than matched gout patients without monosodium urate deposition (n=10) and matched control/non-gout patients (n=10) (p value of 0.03).
Conclusion: Carotid arteries are not a common site for monosodium urate deposition, but if present can lead to increased atherosclerosis in the involved vessel, resulting in increased incidence of TIA/strokes.
Limitations: No limitations were identified.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval was obtained before starting the project.
7 min
Correlation between carotid NASCET-stenosis and plaque quality as characterised by the new Plaque-RADS score
Gian Luca Chabert, Cagliari / Italy
Author Block: G. L. Chabert1, R. Cau1, A. Schindler2, G. Grassi1, M. Porcu1, T. Saam3, L. Saba1; 1Cagliari/IT, 2Munich/DE, 3Rosenheim/DE
Purpose: The study aimed to investigate the correlation between the degree of carotid plaque stenosis, as measured by the NASCET method, and plaque composition as categorised by the new Plaque-RADS grading system.
Methods or Background: One hundred and twenty subsequent patients (61 males, 59 females; mean age 67.9±18.61 SD) who were admitted for suspected cerebral ischaemic disease (CVD) and underwent CT-angiography of the carotid arteries were included in the study. Both carotid arteries were evaluated for the presence and quality of atherosclerotic plaque via the recently introduced Plaque-RADS score, and the degree of carotid stenosis was quantified according to NASCET.
Results or Findings: In carotid vessels of the right side the average degree of stenosis increased with increasing lesion severity from 0±SD0% for Plaque-RADS 1 (n=41), 8.6±9.8SD% for Plaque-RADS 2 (n=36), 27.0±10.4SD% for Plaque-RADS 3 (n=33), to 46.4±23.8SD% for Plaque-RADS 4 (n=10). The same increase was present for the left side from 0±SD0% for Plaque-RADS 1 (n=41), 9.6±10.3SD% for Plaque-RADS 2 (n=25), 25.7±13.2SD% for Plaque-RADS 3 (n=45), to 62.4±30.2SD% for Plaque-RADS 4 (n=9). A t-test was conducted with a significant difference result of <0.005. Pearson correlation was also used (r=.807 for right stenosis with right plaque-RADS values; r=.763 for the left side), when considering only plaque-RADS values of 3 and 4 (r=.505 for right side; r=.622 for left side).
Conclusion: Plaque-RADS scores positively correlate with degree of carotid stenosis, especially at lower lesion grades. On the contrary, our results also show that already in vessels with low grade stenosis, features of high risk plaques may be encountered and easily described in a standardised fashion using the novel Plaque-RADS score.
Limitations: The study's data collection was confined to a single institution and evaluated with a new, yet to be completely confirmed, scoring method.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: The study is retrospective.
7 min
Imaging of the aorta in dual-source photon-counting CT: impact of low energy virtual monoenergetic imaging on image quality, vascular contrast and diagnostic assessability
Ibrahim Yel, Frankfurt a. Main / Germany
Author Block: I. Yel1, C. Booz1, V. Koch1, L. D. Gruenewald1, L. S. Alizadeh1, S. Martin1, T. J. Vogl1, S. Waldeck2, D. P. Overhoff2; 1Frankfurt/DE, 2Koblenz/DE
Purpose: To evaluate the impact of low energy VMI+ reconstructions on quantitative and qualitative image quality, vascular contrast and diagnostic assessability of the aorta in patients undergoing photon-counting CTA.
Methods or Background: A total of 125 patients (69 male) who had undergone dual-source photon-counting CTA scans of the aorta were retrospectively analysed in this study. Standard 120 kV CT images and low keV VMI+ series from 40 to 100 keV with an interval of 15 keV were reconstructed. Quantitative analyses included evaluation of vascular CT numbers, SNR and CNR. CT number measurements were performed in the ascending aorta, the aortic arch, the thoracic and infrarenal descending aorta. Qualitative analyses were performed by three board-certified radiologists independently using five-point scales to evaluate image quality, vascular contrast and diagnostic assessability.
Results or Findings: Mean attenuation, CNR and SNR values were highest in 40 keV VMI+ reconstructions (HU, 1312 ± 13; CNR, 32 ± 8; SNR, 34 ± 10) followed by 55-keV VMI+ reconstructions (HU, 731 ± 9; CNR, 24 ± 6; SNR, 27 ± 9); all three mean values at these keV levels were significantly higher compared with the remaining VMI+ series and standard 120 kV CT series (HU, 160 ± 8; CNR, 18 ± 5; SNR, 26 ± 6) (p<.0001). The qualitative analysis showed the highest rating scores for 55 keV VMI+ reconstructions followed by 40 keV and 70 keV VMI+ series with a significant difference compared to standard 120 kV CT images regarding image quality, vascular contrast and diagnostic assessability of the aorta (p<.0001).
Conclusion: Low keV VMI+ reconstructions at a level of 40-55 keV significantly improve image quality, vascular contrast and diagnostic assessability of the aorta compared with standard CT series in photon-counting CTA.
Limitations: The single-centre retrospective study design was an identified limitation.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the local IRB.

This session will not be streamed, nor will it be available on-demand!