RPS 1403 - Quantitative cardiac imaging and standardisation studies

Author Block: W-S. Yoo¹, S. Hong¹, Y. J. Suh¹, C. Kim²; ¹Seoul/KR, ²Ansan/KR
Purpose: To evaluate the reliability of AI-based automatic coronary artery calcium (CAC) quantification and density measurement on non-ECG-gated low-dose chest CT (LDCT) compared with manual measurements on calcium scoring CT (CSCT), and to investigate the influence of scan parameters.
Methods or Background: This retrospective study included 417 patients from four academic hospitals who underwent both CSCT and LDCT within six months, with CAC scores >0. CT scans were obtained on scanners from four different vendors. CSCT was reconstructed at 3-mm with standard kernel; LDCT was reconstructed in 1-mm and 3-mm pairs with various kernels (identical within each pair). Patients were grouped as high- and low-noise by LDCT noise level. Agatston score, peak density attenuation, and density score (Agatston score divided by total calcium area) were measured manually on CSCT by cardiac radiologists and obtained automatically on LDCT using commercial software. Reliability of each variable between manual CSCT and automatic LDCT (LDCTauto) was evaluated using the concordance correlation coefficient with 95% CIs.
Results or Findings: In the low-noise group (n=337), 1-mm LDCTauto demonstrated higher reliability than 3-mm LDCTauto across all variables: Agatston score (0.943 [0.932–0.952] vs. 0.922 [0.908–0.934]), peak density attenuation (0.812 [0.773–0.846] vs. 0.710 [0.664–0.750]), and density score (0.565 [0.496–0.627] vs. 0.556 [0.490–0.615]). However, in the high-noise group (n=80), 1-mm LDCTauto showed lower reliability than 3-mm LDCTauto for Agatston score (0.710 [0.604–0.791] vs. 0.838 [0.763–0.890]) and peak density attenuation (0.157 [0.087–0.226] vs. 0.690 [0.567–0.782]), but slightly higher reliability for density score (0.442 [0.270–0.587] vs. 0.411 [0.219–0.572]).
Conclusion: The reliability of automatic CAC measurements on LDCT is influenced by slice thickness and noise level. With lower-noise kernels, 1-mm LDCT may provide more reliable automatic CAC quantification and density measurement than 3-mm LDCT.
Limitations: Further validation using different software is required.
Funding for this study: None.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The Institutional Review Boards of the participating hospitals approved this study and waived the requirement for informed consent.

Author Block: S. Barus¹, M. T. Hagar², M. Soschynski¹, F. Bamberg¹, C. L. Schlett¹, T. Krauß¹, C. Schuppert¹; ¹Freiburg im Breisgau/DE, ²Charleston, SC/US
Purpose: Photon-counting detector coronary CT angiography (PCD-CCTA) is typically performed at 120/140 kVp (standard dose, SD) to enable full spectral applications, including material decomposition and virtual monoenergetic imaging (VMI). Low-dose (LD) acquisitions at 90/70 kVp are limited to VMI but may suffice to rule out coronary artery disease.
Methods or Background: In this single-center post-hoc study, consecutive low-risk-profile patients (age <60 years, BMI <30, Agatston score <200) underwent clinically indicated CCTA using a dual-source PCD-CT system. Scans were randomly acquired in either full (SD) or reduced spectral mode (LD), using high-pitch or prospective sequence acquisition. Images were reconstructed as 55 keV VMIs using a vascular kernel (Bv48, 0.4 mm). Quantitative image quality included signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) averaged across the ascending aorta, proximal and distal coronaries; CNR of ≥10.0 was defined as diagnostic. Radiation dose was evaluated using the dose-length product (DLP), and multivariate analysis was used to assess indicators of radiation dose.
Results or Findings: A total of 192 patients were included (115 LD, 48.1 ± 8.3 years; 77 SD, 53.0 ± 7.1 years). LD reduced radiation dose by 52.6% (91 [58–179] vs. 192 [137–475] mGy·cm, p<0.01), with slightly lower SNR (15.5 [13.1–17.6] vs. 19.2 [15.5–23.3], p<0.001) and CNR (13.0 [11.1–14.9] vs. 16.2 [13.2–20.6], p<0.001), although both exceeded diagnostic thresholds. In multivariate analysis, LD and high-pitch acquisition predicted lower dose (β = -138.8 and -215.6, p<0.001), while higher BMI and male sex increased dose (β = 7.8 and 47.2, p=0.016 and 0.018).
Conclusion: PCD-CCTA acquired at low tube potentials enables substantial radiation dose reduction while maintaining diagnostic image quality. Our results indicate that in low-risk patients, patient-protective measures should be favored over full spectral capacity.
Limitations: Post-hoc, single-center study.
Funding for this study: No funding was provided for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics Committee of the Medical Center – University of Freiburg (No. 21-1469, approved on September 21, 2021).

Author Block: H. Zhang, M. Lu; Peking/CN
Purpose: The clinical translation of cardiovascular magnetic resonance (CMR) T1 mapping is limited by inconsistent acquisition protocols and reference ranges. We aims to establish the ranges of native T1 and extracellular volume (ECV) in healthy Chinese adults and across various cardiovascular diseases, and to evaluate their diagnostic performance.
Methods or Background: This prospective study included 1,237 human subjects with 15 cardiovascular conditions and 12 swine for histopathologic validation. T1 mapping were acquired using Modified Look-Locker Inversion Recovery (MOLLI) and Shortened MOLLI (ShMOLLI) sequences on a 3 Tesla Siemens scanner. Histological collagen volume fraction (CVF) in swine myocardium was quantified and correlated with T1 mapping. Global native T1 and ECV were compared between myocardial disease groups and healthy controls. The differences between MOLLI and ShMOLLI measurements were analyzed using Bland-Altman analysis.
Results or Findings: Histological validation in swine confirmed strong correlations between T1 mapping parameters and CVF. In clinic, the references ranges derived from healthy controls were 1179±29 ms (MOLLI) and 1147±33 ms (ShMOLLI) for native T1, and 25.1±2.1% (MOLLI) and 25.5±1.9% (ShMOLLI) for ECV. Gender stratification revealed higher native T1 in females compared to males within healthy population. MOLLI yielded systematically higher native T1 values compared to ShMOLLI across the entire cohort and in all conditions. ECV showed inter-sequence consistency. Across the spectrum of cardiovascular diseases, significant deviations in native T1 and ECV were observed. T1 mapping had exceptional diagnostic performance for cardiac amyloidosis and Fabry disease, and strong diagnostic capabilities for most primary cardiomyopathies, ischemic heart disease, and myocarditis.
Conclusion: This large-scale study established comprehensive reference ranges for native T1 and ECV using MOLLI and ShMOLLI at 3 Tesla, highlighting their diagnostic utility across diverse cardiovascular conditions.
Limitations: Patients were enrolled based on a single primary cardiovascular diagnosis
Funding for this study: National Natural Science Foundation of China (grant no. 82471973)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The pre-clinical study was approved by Ethics Committee for Animal Study at Fuwai hospital and the Care of Experimental Animals Committee of the Chinese Academy of Medical Sciences and Peking Union Medical College (0102-1-30-ZX(X)22).
The clinical study was approved by the local Institutional Ethics Committee. All human participants gave written informed consent.

Author Block: A. Perazzolo¹, T. C. Chao², D. Curione¹, M. Rebonato¹, M. Pilari¹, G. Butera¹, L. Natale¹, T. Leiner², A. Secinaro¹; ¹Rome/IT, ²Rochester, MN/US
Purpose: Blood oxygen saturation is a key physiological marker of cardiopulmonary function. Differences in magnetic susceptibility of oxygenated versus deoxygenated hemoglobin affect MRI relaxation times. Models such as the Luz–Meiboom (LM) equation link T2 to oxygen saturation, while synthetic hematocrit (HCT) can be inferred from native T1. Nonetheless, flow-related artifacts remain a limitation to quantification. We evaluated the feasibility of non-invasive CMR-based oxygen saturation assessment, validated against simultaneous sampling during MRI-guided catheterization in pediatric patients, and tested a physiological flow-based correction.
Methods or Background: Synthetic HCT was calculated from native T1 (MOLLI), and T2 maps were obtained from a T2-prep bSSFP sequence (echo times 0 and 55 ms). These inputs were applied to the LM model to generate saturation maps. Right ventricular blood-pool ROIs provided image-based values, compared with catheter samples from pulmonary arteries under matched conditions. Additional morpho-functional indices (ventricular volumes, ejection fraction, main pulmonary artery [MPA] flow) were extracted to assess their influences on image-based SbO₂. A linear correction model was derived associating peak velocity of MPA (VmaxMPA) and ΔSbO₂ (difference between catheter and image-based measurement) verified by leave-one-out cross validation.
Results or Findings: Thirteen pediatric patients were retrospectively included (9 restrictive cardiomyopathies, 1 dilated cardiomyopathy, 1 Uhl’s anomaly, 1 congenitally-corrected transposition, 1 idiopathic right atrial dilatation). MPA peak velocity (Vmax) correlated with ΔSbO₂ (R²=0.93). Leave-one-out cross validation demonstrated that applying VmaxMPA correction reduced root mean square error from 15.92% to 2.77%, mitigating flow-related artifacts.
Conclusion: Non-invasive blood oxygen saturation estimation from standard T1/T2 mapping is feasible when synthetic HCT and the LM equation are combined with a physiological correction for MPA flow. Validation against invasive sampling supports the robustness and clinical applicability of this approach for artifact-resistant quantification.
Limitations: Small sample size and absent external validation.
Funding for this study: Italian Ministry of Health, Current Research funds.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Bambino Gesù Children’s Hospital ethics committee.

Author Block: E. Cantalamessa¹, F. Fiocchi¹, G. Battinelli¹, R. Cuoghi Costantini¹, G. Ligabue²; ¹Modena/IT, ²Sassuolo/IT
Purpose: The use of shortened non-contrast Cardiac Magnetic Resonance (CMR) protocols may reduce examination time and still maintain diagnostic accuracy. This study aimed to assess the effectiveness of such an approach for detecting left ventricular myocardial disease, using mapping and ventricular function as decision-making criteria.
Methods or Background: This single-center retrospective study included 337 patients referred for CMR with heterogeneous clinical indications. Each examination was blindly evaluated using a short protocol including cine sequences, edema-sensitive sequences and T1/T2 mapping. For each one of them, functional parameters (LVEDV, LVEDV/BSA, LVEF, LVSV) and mapping values (T1 MAX, T2 MAX) were collected. T1 MAX and T2 MAX were defined as the highest mean values assigned to a myocardial segment in native T1 and T2 mapping, respectively. The diagnosis obtained with the standard contrast-enhanced protocol served as the reference standard. ROC analysis was used to assess the performance of individual parameters and combined models (logistic regression, decision tree) in predicting LGE positivity.
Results or Findings: T1 Max demonstrated an AUC of 0.716 (sensitivity 67.2%; specificity 66.8%), which improved with the addition of LVEDV/BSA (AUC 0.721). The decision tree achieved an AUC of 0.751. A T1 Max cut-off >1113 ms identified a high-risk subgroup (positivity 70%), whereas T1 Max <1025 ms defined a low-risk subgroup (positivity 13%). In intermediate cases, adding LVEDV/BSA >93 ml/m² improved discriminatory capacity.
Conclusion: In a heterogeneous population, a short non-contrast CMR protocol based on T1 mapping thresholds and left ventricular function may identify patients at higher risk of myocardial injury. Implementing a shortened protocol could significantly reduce acquisition times, supporting more efficient and personalized diagnostic pathways in everyday clinical practice.
Limitations: Limitations include retrospective single-center nature of the study, heterogeneous patient population, and the need for experienced radiologist presence.
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

Author Block: S. Forouzannia¹, S. R. Rafiei Alavi¹, S. M. Forouzannia², M. Umair³, G. Francese⁴; ¹Tehran/IR, ²San Francisco, CA/US, ³New York, NY/US, ⁴Genoa/IT
Purpose: Cardiac amyloidosis (CA) is the leading cause of mortality in systemic amyloidosis, highlighting the need for accurate risk assessment to guide patient management. While the diagnostic value of cardiac MR (CMR) parametric mapping is well established, its prognostic utility remains inconsistent across studies. To perform a systematic review and meta-analysis to evaluate the prognostic value of CMR parametric mapping in patients with CA.
Methods or Background: An extensive search was conducted in Medline, Scopus, Embase and Web of Science databases. Eligible studies were observational studies that reported HRs for predicting adverse outcomes in patients with CA using CMR parametric mapping.
Results or Findings: 22 studies with 3398 patients were included in this systematic review. Higher extracellular volume (ECV) values were associated with increased mortality, both as a dichotomous (HR: 2.90; 95% CI: 1.68 to 5.01) and continuous variable (HR for 1% increase: 1.08; 95% CI: 1.06 to 1.10; HR for 3% increase: 1.17; 95% CI: 1.11 to 1.22 and HR for 10% increase: 2.11; 95% CI: 1.70 to 2.62). Higher native T1 mapping values were associated with mortality as a dichotomous variable (HR: 1.33; 95% CI: 0.79 to 2.24). Native T2 mapping showed inconsistent associations with prognosis across studies.
Conclusion: Higher ECV and native T1 values are associated with worse prognosis in CA, supporting their role in risk stratification. Further studies with standardised CMR protocols are needed to enhance the prognostic utility of parametric mapping in clinical practice.
Limitations: The heterogeneity in ECV and T1 cut-off values across studies presents a challenge to standardisation, thereby limiting direct comparisons. Another limitation is the insufficient data available on native T2 mapping due to the limited number of studies investigating its values in the setting of CA.
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:

Author Block: C. Filliol¹, C. De Bourguignon², A. Ghaouar¹, M. Bourmaaz¹, A. Braillon¹, A. Deliniere¹, A. Jobbe-Duval¹, A. Fouilhoux², S. A. Si-Mohamed¹; ¹Bron/FR, ²Lyon/FR
Purpose: To analyze differences in myocardial extracellular volume (ECV) according to sex, left ventricular hypertrophy (LVH) and fibrosis in Fabry disease.
Methods or Background: Between February 2017 and August 2023, 64 patients with Fabry disease (36 women [56%], mean age 45 ± 15 years) from our reference center who underwent baseline cardiac magnetic resonance imaging (CMR) on a 1.5 Tesla MRI system (Philips XXX) were retrospectively included. ECV was calculated using the hematocrit level obtained on the same day after semi-automatic segmentation on short-axis T1 mapping images, using a clinical workstation (ISPortal, CM Suite, Philips). The presence of fibrosis and left ventricular hypertrophy (LVH) was assessed in consensus by 2 experienced observers. Following data were registered: age, sex, alpha-galactosidase A levels, troponin levels, left ventricular ejection fraction (LVEF). Mann-Whitney and t- test were used according to their normality.
Results or Findings: Age, troponine and LVEF were not significantly different according to sex, Alpha-galactosidase level was significantly lower in men (0.46±0.15 vs 3.07±1.8, p<0.001). LVH status was significantly higher in men (11(44%) vs 5(12.8%), p<0.001) and fibrosis status was significantly higher in men (9(36%) vs 11(28.2%), p<0.001). 13(20%) patients presented both LVH and fibrosis (4 women, 9 men). 46(72%) patients were indemn of LVH and fibrosis (29 women, 17 men). In men, global ECV was significantly higher in presence of LVH (27.7%±3.9 vs 24.4%±2.4, p=0.02), as well as in presence of fibrosis (28.2%±4.2 vs 24.6%±2.3, p=0.04). No statistical difference was found in women.
Conclusion: Myocardial ECV is significantly elevated in men with LVH and fibrosis, highlighting sex-specific differences in cardiac tissue alterations. These findings suggest that assessing ECV, particularly in male patients, may offer valuable insights into myocardial involvement and disease progression in Fabry disease
Limitations: Retrospective and monocentric study.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: HCL ethical commitee N°23257

Author Block: A. Colombo, C. Gnasso, D. Vignale, M. Liberotti, A. Palmisano, A. Esposito; Milan/IT
Purpose: To develop an automatic software for myocardial extracellular volume (ECV) analysis and to evaluate its association with a composite outcome of death and heart failure at 12 months after transcatheter aortic valve implantation (TAVI) in patients with aortic stenosis (AS).
Methods or Background: In this retrospective single-centre study, we considered patients who underwent pre-TAVI CT (10/2020-12/2023) and employed a fully automated pipeline for volumetric quantification of ECV and fibrotic burden (FB), a.k.a. percentage of left ventricular voxels with ECV values above a set threshold. The pipeline consisted of four steps: 1-segmentation of cardiac structures and 2-their post-processing, 3-co-registration of late post-contrast and pre-contrast scans, 4-actual ECV quantification. Agreement between automated (ECVauto) and manual (ECVmanu) measurements by two radiologists was assessed with Bland-Altman analysis, while prognostic value was evaluated using Kaplan-Meier curves with log-rank cutoffs and multivariable Cox models adjusted for clinical and echocardiographic variables. Analyses were performed in Python (v3.9.13), with p=0.05 significance.
Results or Findings: Among the 438 included patients (median age 82 years, F221/M217), median EuroSCORE was 5.2, mean gradient 44 mmHg, and ejection fraction 59%. The 12 months composite outcome occurred in 74 patients (16.9%), including 57 deaths (10.7%). The pipeline processed all CTs in ~4.5min/patient. Bias and limits of agreement between ECVauto and ECVmanu were 0.82% (-11.9%/13.6%). ECVauto and ECVmanu were higher in patients with events (32.5vs29.3%, p=0.001; 30.1vs28.0%, p=0.020, respectively). Elevated ECVauto independently predicted worse outcomes (cutoff 32.9%, p<0.001; HR 1.04 per 1%, p=0.028), as did high FB (cutoff 48.5%, p<0.001; HR 1.02, p=0.031).
Conclusion: The pipeline enabled rapid, automated ECV mapping, independently predicting 12-month outcomes. Results further support CT-based fibrosis assessment.
Limitations: Limitations include retrospective, single-center design, improving the segmentation step to exclude artifacts, and possible inclusion of undiagnosed amyloidosis.
Funding for this study: This study was partially granted by: European Union - Next Generation EU, Mission 4 Component 1, CUP D53D23021100001”, Bando PRIN PNRR 2022 (P2022JBKN2), Italian Ministry of University and Research (MUR).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This monocentric observational study was approved by the institutional review board (CT-based myocardial characterization study: CTMyoC 112/INT/2019 ). The study is a retrospective analysis of prospectively collected data and was conducted according to the Declaration of Helsinki. Written informed consent was obtained for all participants.

Author Block: C. Gnasso, D. Vignale, A. Palmisano, D. Serra, D. Margonato, A. Esposito; Milan/IT
Purpose: Previous studies have demonstrated that CT for transcatheter aortic valve implantation (TAVI) planning can quantify interstitial fibrosis as extracellular volume fraction (ECV), associated with a worse prognosis. Also the echocardiographic flow-gradient pattern (high-gradient [HG]; paradoxical low-flow [LF] low-gradient [LG], and classical LFLG AS) is associated with prognosis, with classical LFLG having the worst. Thus, we aimed to investigate the incremental prognostic value of ECV over flow-gradient pattern.
Methods or Background: Single-center prospective observational study. Consecutive patients undergoing TAVI (October 2020-November 2023) were enrolled and categorized as HG, paradoxical or classical LFLG AS. ECV was calculated from a 5-minute delayed scan. After a follow-up of at least 1 year, a composite endpoint of death or heart failure hospitalization was collected.
Results or Findings: In the final population of 460 patients (82 [77-85] years), 83% had HG AS, 8.0% paradoxical, and 8.9% classical LFLG AS. After a median follow-up of 468 days, patients with classical LFLG had a higher rate of events (39.0%) than paradoxical LFLG (29.7%), and HG (15.6%) (p=0.004) and a higher ECV (32.0% [27.3-34.9] vs 28.0% [25.2-31.2] vs 28.0% [26.2-33.0], respectively; p=0.001). On Kaplan-Meier analysis, the combination of flow-gradient pattern and ECV distinguished distinct risk profiles: patients with classical LG-LF AS and a high ECV had the worst prognosis (p<0.001). The improved risk stratification capability of ECV was confirmed with the likelihood ratio test (p=0.019) and the net reclassification index (0.3).
Conclusion: ECV derived from pre-procedural TAVI planning has an incremental prognostic value over echocardiographic flow-gradient pattern.
Limitations: Relatively small sample size; ECV assessed with a single manual ROI with a single-energy CT method; absence of a standard of reference to detect cardiac amyloidosis (to reduce this bias, we excluded patients with an ECV >40%).
Funding for this study: This study was partially granted by: European Union - Next Generation EU, Mission 4 Component 1, CUP D53D23021100001”, Bando PRIN PNRR 2022 (P2022JBKN2), Italian Ministry of University and Research (MUR).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the local Ethical Committee (CT-based myocardial characterization study: CTMyoC 112/INT/2019 and amendments) and conducted according to the Declaration of Helsinki.