European heart journal. Imaging methods and practice最新文献

Congenital heart disease (CHD), the most common congenital malformation worldwide, affects approximately 8-10 per 1000 live births. Despite major advances in diagnosis and treatment, significant disparities remain between high- and low-resource settings, with a substantial impact on survival and long-term outcomes. The Aesculapius Project was conceived to address these inequalities by employing advanced imaging and three-dimensional (3D) printing as educational tools to enhance paediatric cardiac surgery training. Initiated by the humanitarian foundations Bambini Cardiopatici nel Mondo and European Heart for Children, the program offers free training to early-career surgeons from resource-limited countries. Patient-specific 3D cardiac models are generated from computed tomography (CT) and magnetic resonance imaging (MRI) data and used in structured courses that combine lectures, supervised hands-on surgical simulations, and remote mentorship. To date, 81 physicians from 16 countries have participated in the program, performing 10-20 simulated procedures on complex CHD models under expert supervision. Technical performance scores improved by an average of 30%, and participants emphasized the educational value of repeated supervised practice and video-assisted review. The Aesculapius Project demonstrates that 3D anatomical models offer an ethical and effective platform for surgical education, representing a concrete step toward expanding access to high-quality paediatric cardiac surgery training in resource-limited settings.

Aims: Myocardial strain is a powerful, non-invasive diagnostic and prognostic marker in patients with heart disease. However, its applicability is hindered by uncertain repeatability, particularly for segmental values. This study measures the repeatability of myocardial strain across eight imaging methods.

Methods and results: In this prospective study, 20 healthy volunteers (aged 34 $\pm$ 8, 14 men) were recruited and scanned twice with eight strain imaging protocols: cardiac magnetic resonance (CMR) at 1.5T and 3T with cine, tagging, and displacement encoding with stimulated echoes (DENSE) sequences, and 2D and 3D echocardiography (Echo). Global and segmental strains were quantified from each scan. Inter-scan repeatability was assessed with the coefficient of variation (CoV), intraclass correlation coefficient, and Bland-Altman analysis.

Results: Inter-scan repeatability of global strains ranged from excellent to fair (CoV ≤ 20%) depending on protocol. Using CMR feature tracking at 1.5T, relative global longitudinal strain (GLS) changes exceeding 11.2% are unlikely to be caused by measurement variability alone; this figure is 5.5% for 2D echocardiography. Segmental strain values frequently had poor repeatability (CoV > 20%), particularly for longitudinal and radial strains.

Conclusion: Imaging protocols including CMR and Echo can measure global strain parameters with fair repeatability, but segmental strain values are unreliable. Future work should aim to improve the repeatability of segmental strain values, particularly longitudinal strain.

Aims: Heart transplantation is a critical life-saving procedure for patients with end-stage heart failure. However, predicting postoperative mortality remains challenging. The aim of this study is to examine the effectiveness of machine learning (ML) models for predicting 1-year mortality among heart transplant recipients in Saudi Arabia.

Methods and results: A retrospective observational study was conducted using data from King Faisal Specialist Hospital & Research Centre, a large tertiary hospital in Saudi Arabia, that included all heart transplant cases from January 2007 to December 2022. We evaluate and compare the accuracy of support vector machine (SVM) and logistic regression (LR) models in predicting 1-year mortality. We also identify key predictive variables influencing mortality rates among recipients. SVM and LR models were developed and compared using accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve as performance metrics. The study analysed data from 419 patients, revealing that ischaemia time, devices like left ventricle assist device, extracorporeal membrane oxygenation, and body mass index (BMI) were significant mortality predictors. The LR model showed a testing accuracy of 96.43%, with weight and BMI having the strongest influence on mortality prediction. The SVM model had a testing accuracy of 95.24%, demonstrating consistent performance across dataset.

Conclusion: The findings indicate that ML models, particularly SVM and LR, are effective in predicting 1-year mortality post-heart transplantation as well as identifying significant predictors of mortality. This research contributes to the global knowledge in heart transplant and highlights the importance of new technologies in tailoring healthcare strategies for the Saudi population.

Aims: Non-invasive estimation of left ventricular pressure (LVP) is crucial for managing cardiovascular diseases such as heart failure and myocardial infarction (MI). Current clinical practices rely on invasive catheterization, limiting its feasibility for routine or longitudinal monitoring. This study evaluates the accuracy of a novel LVP reconstruction algorithm in preclinical (rat) experiments.

Methods and results: Using a standard coronary occlusion/reperfusion model (n = 39 rats), we validated our algorithm across three physiological states: baseline, myocardial ischaemia, and MI. LVP waveforms were reconstructed using only carotid pressure waveforms and echocardiographic measurements. Algorithm performance was assessed by comparing reconstructed LVP waveforms to invasively measured LVP, using key haemodynamic metrics such as left ventricular end-diastolic pressure (LVEDP) and the subendocardial viability ratio (SEVR). Agreements between waveforms were assessed using intraclass correlation coefficients (ICC), normalized Euclidean distance (NED), and differences in harmonic modulus. The algorithm accurately estimated LVEDP across all physiological states (mean absolute error: 1.5 mmHg), with strong correlation to invasively measured LVEDPs (r = 0.91). Predicted SEVR also showed strong agreement with measured values (r = 0.96). The algorithm captured the expected LVEDP elevation and SEVR reduction during myocardial ischaemia, and the metric's partial recovery after reperfusion. Waveform-level agreement demonstrated near-perfect alignment, with high ICC (98.5%), low NED (0.062), and minimal harmonic modulus differences (0.043) for all tested cases.

Conclusion: This study demonstrates that LVP can be accurately reconstructed using the proposed algorithm in rats. Our algorithm reliably captured key LVP metrics and waveform features across varying physiological states, supporting its potential for cardiac monitoring.

Interventional cardiac magnetic resonance imaging (iCMR) offers distinct advantages for guiding complex cardiac procedures, including 3D visualization, soft tissue characterization, and avoidance of ionizing radiation. Transseptal puncture (TSP), essential for left heart access, poses specific challenges under MR-guidance. The development of MR-compatible TSP sets comprising non-ferromagnetic sheaths, dilators, and needles, represents a major step toward safe execution of TSP in the MRI environment. This report provides practical, step-by-step guidance for MR-guided TSP, focusing on imaging strategies and integration of advanced 2D and 3D navigation tools. Real-time cine imaging in dedicated planes enables precise localization of the fossa ovalis, confirmation of septal tenting, and avoidance of adjacent structures. Complementary use of a vendor-neutral MR-compatible 3D navigation system allows dynamic catheter tracking within a segmented static 3D anatomical shell, enhancing spatial orientation and procedural accuracy. Feasibility was demonstrated in a porcine model, where an MR-compatible sheath and trackable dilator were successfully navigated to the interatrial septum and TSP was achieved, enabling left atrial (LA) access. Subsequent mapping confirmed catheter positioning within the LA. Remaining challenges include limited guidewire visibility, low image temporal resolution compared with fluoroscopy, and the investigational status of current MR-compatible TSP sets. These factors must be addressed before clinical translation. In conclusion, MR-guided TSP using dedicated imaging planes and MR-compatible devices is technically feasible and may facilitate future radiation-free left heart interventions. Continued device refinement, including improved passive instrument visibility and active tracking technologies, faster real-time cine imaging, and regulatory approval are critical for safe and widespread clinical adoption.

Aims: Left atrial volume (LAV) is a recognized prognostic marker of cardiovascular events. However, normal LAV thresholds on cardiac computed tomography angiography (CCTA) remain poorly defined, and the optimal assessment method is unclear. Our aim was to determine normal reference values of maximal systolic LAV (LAVmax) and ventricular dimensions normalized by age, sex, and body surface area (BSA) from CCTA by using dedicated semi-automated analysis software.

Methods and results: This single-centre retrospective study included 250 healthy subjects with no cardiac history or significant CCTA abnormalities, stratified by sex and age. LAVmax was measured using both 3D and area-length methods, and all other cardiac chambers were analysed with dedicated 3D semi-automated software (Vitrea^®). LAVmax increased significantly with age in both sexes but showed no sex difference when indexed to BSA. The 3D volumetric method was more reproducible (r = 0.86, P < 0.001) and yielded larger values compared with area-length method. Reference values for LAVmax and all cardiac chambers were provided for both sexes and each age group. Age, in concert with sex, was associated with significant differences in RV end-diastolic volume and LV ejection fraction (P-values 0.027 and 0.03).

Conclusion: Indexed LAVmax was not significantly different across sexes but increased with age. LAVmax can be reliably reported from CCTA datasets, with 3D volumetric method providing the largest and most reproducible values. Normal values for all cardiac chambers according to age categories and sex were also provided. These normative values enhance the clinical utility of routine CCTA beyond coronary imaging.

Aims: Left ventricular (LV) enlargement is a common training-induced adaptation in athletes, particularly in endurance sports. Previous research indicates that indexing LV volumes and mass to absolute peak oxygen uptake (VO₂_peak) better reflects physiological adaptation than traditional indexing to body surface area (BSA). Therefore, we investigated whether indexing LV end-diastolic volume (LVEDV) and mass to VO_2peak could eliminate differences in LV size among athletes from different sport categories (endurance, mixed, power, and technical).

Methods and results: This analysis included 70 athletes from the multicenter COSMO-S in Germany and 15 elite endurance athletes from Norway. All participants (29 ± 8 years, 52 male) underwent echocardiography and cardiopulmonary exercise testing. In regression analyses, VO_2peak (L/min) accounted for a significantly greater proportion of the variance in LVEDV than BSA (R² 0.64 vs. 0.19, P < 0.001), while this difference was not significant for LV mass (R² 0.54 vs. 0.36, P = 0.06). When indexed to BSA, both LVEDV and LV mass revealed significant differences across sports (both P ≤ 0.019), that disappeared when indexed to VO₂_peak (all P ≥ 0.40). In a cohort of 12 dilated cardiomyopathy (DCM) patients serving as a pathological reference group, indexing LVEDV and LV mass to VO_2peak better differentiated DCM patients from athletes than indexing to BSA.

Conclusion: Indexing LV size to VO₂_peak may provide a more physiological interpretation of cardiac adaptations in athletes and reduce sport-specific differences due to better consideration of training-induced adaptations. These findings should be replicated in larger cohorts and tested for the ability to detect subtle pathologies.

With the growing global adoption of transcatheter tricuspid valve intervention (TTVI) and the increasing number of available devices, a comprehensive understanding of right heart dysfunction has become essential. Right heart dysfunction is frequently observed both before and after TTVI and is associated with adverse clinical outcomes. Therefore, a thorough understanding of right heart anatomy and physiology is critical for accurately assessing its pathological states. This review synthesizes current knowledge by integrating findings from major landmark studies on TTVI, with a focus on the available assessment tools for predicting patient outcomes. The anatomy section systematically reviews each component of the right heart-the right atrium, right ventricle, tricuspid valve, and pulmonary valve -while the physiology section emphasizes microstructural characteristics and the pressure-volume relationships. In addition, recommendations from the Tricuspid Valve Academic Research Consortium and the imaging parameters used in recent studies are discussed. Finally, future directions for imaging-based assessment of right heart function in the context of TTVI are highlighted.

Aims: Ejection fraction (EF) and end-systolic volume (ESV) are prognostic markers in cardiovascular disease. While MRI provides accurate assessments, its cost limits widespread use. Non-contrast cardiac CT (NCCT), used for coronary artery disease screening, may offer additional functional information. To evaluate the accuracy of AI-derived ventricular volumes and EF from NCCT compared with contrast cardiac CT (CCT) and MRI.

Methods and results: This single center study included 205 patients who underwent cardiac CT for valve planning, divided into retrospective and prospective cohorts. A validated AI algorithm was applied to low-dose NCCT images at end-diastole and end-systole. Right (RV) and left ventricles (LV) volumes and their EFs were compared with CCT and MRI. In the prospective cohort (49 women, 53 men; mean age 73.9 ± 10.3 years), NCCT correlated strongly with CCT for LVEDV (152 mL; -14.2% relative difference; r = 0.91) and LVESV (96 mL; +32.6%; r = 0.84), with similar correlations for RVEDV (163 mL; -8.4%; r = 0.82) and RVESV (121.4 mL; +33.1%; r = 0.85). NCCT predicted LVEF <40% with 98% negative predictive value and 87% accuracy. LVEDV correlated strongly with MRI (n = 16) for CCT (240 mL; +4.2%; r = 0.99) and NCCT (197 mL; -14.3%; r = 0.97), as did LVESV for CCT (115 mL; -5%; r = 0.99) and NCCT (134 mL; +11%; r = 0.97).

Conclusion: AI-derived ventricular volumes from NCCT show moderate to strong correlations, but EF is underestimated. The derived EF can be a screening tool to rule out significant ventricular dysfunction.