BACKGROUND Although the prognostic value of vasodilator stress cardiovascular magnetic resonance (CMR) is established, it has never been assessed in patients <45 years with suspected coronary artery disease. The aim was to assess the long-term prognostic value of stress CMR to predict mortality in a multicenter cohort of patients <45 years-old. METHODS Between December 2008 and December 2024, a prospective multicenter registry enrolled all consecutive patients undergoing vasodilator stress CMR at three French centers. For the present retrospective analysis, all patients aged <45 years were included. The primary endpoint was all-cause mortality, assessed through the French National Death Registry. Cox regression analyses were performed to determine the prognostic value of inducible ischemia and late gadolinium enhancement (LGE). RESULTS Of the 3,671 patients (mean age 38±6 years, 65% males), 102 (2.8%) died after a median follow-up of 6.8 years (interquartile range: 3-11 years). Annualized mortality rates increased gradually with the number of segments of ischemia and LGE (both p-trend<0.001). The extent of ischemia (hazard ratio [HR]: 1.45; 95% confidence interval [CI]: 1.24-1.70) and LGE (HR: 1.56; 95% CI: 1.39-1.75) were both associated with mortality (both p<0.001). After adjustment for traditional prognostic factors, both the extent of ischemia (adjusted HR: 1.26, per 1-segment increase; 95% CI: 1.06-1.51, p=0.011) and LGE (adjusted HR: 1.50, per 1-segment increase; 95% CI: 1.25-1.80, p<0.001) remained independent predictors of mortality. CONCLUSIONS In patients <45 years undergoing stress CMR, the extent of both inducible ischemia and LGE provide strong and independent prognostic value for predicting all-cause mortality.
Background Multisystem inflammatory syndrome in children (MIS-C) is a sequela of SARS-CoV-2 infection; however, longitudinal cardiac outcomes are unclear. Herein, we test the association between MIS-C biomarkers, inpatient echocardiographic features, and outcomes of interest: presence of myocarditis and myocardial delayed enhancement (MDE) on cardiac magnetic resonance (CMR). Furthermore, we describe the evolution of CMR features in MIS-C. Methods Patients diagnosed with MIS-C from 2020-2021 with CMR were included. Admission and serial biomarkers and echocardiographic features were collected. Primary outcomes measured were diagnosis of myocarditis (using modified Lake Louise criteria) and myocardial delayed enhancement (quantified by full width at half maximum method) on CMR. Secondary outcomes included abnormal LVEF on admission echocardiogram and changes in parametric mapping with serial CMR. Results A total of 66 patients were included (median age 9.7 IQR [7.2-14.7] years). Admission BNP (P=0.016), CRP (P=0.027), and ESR (P=0.002) associated with LV dysfunction on inpatient echocardiogram. CMR1 was obtained a median of 38 [26-61] days after hospital presentation. Myocarditis was present in 44% and MDE in 52% on CMR1, with median MDE burden of 10.8 [8.5-16.8]% in those affected. Inpatient echocardiographic features were not associated with presence of myocarditis or MDE. Admission ESR (P=0.031) associated with MDE. Admission platelets associated with myocarditis (P=0.040) and MDE (P=0.004). CMR2 (n=19), occurring a median of 11 [7-14] months after baseline CMR, demonstrated lower global T2 z-scores (P=0.035) without significant change in global T1 z-scores or ECV. Myocarditis (16% [3/19]) and MDE (28% [5/18]) were less frequent on CMR2. Conclusions For MIS-C patients, initial echocardiographic indices do not associate with myocarditis or MDE by CMR one month later; admission ESR and platelet levels may portend myocarditis or MDE on initial CMR. Reduced T2 z-score over serial CMR suggests resolving edema.
BACKGROUND:Non-contrast T1-weighted cardiovascular MR (CMR) enables assessment of coronary atherosclerotic plaques by exploiting elevated signal intensity within high-risk plaque components. The novel iT2prep-BOOST sequence provides co-registered coronary lumen and vessel wall whole-heart imaging, potentially facilitating plaque assessment. However, direct comparison of quantitative plaque measures between iT2prep-BOOST and coronary computed tomography angiography (CCTA), the non-invasive standard for plaque assessment, remains limited. This study aimed to compare coronary plaque burden and plaque signal intensity measured by iT2prep-BOOST with plaque burden and plaque attenuation characteristics derived from CCTA. METHODS:In this prospective, observational study, patients with stable coronary artery disease confirmed by diagnostic CCTA were recruited and subsequently underwent CMR using the iT2prep-BOOST sequence. Using semi-automated software, per-lesion cross-sectional areas of total, calcified, non-calcified, and low-density non-calcified plaque (≤30 Hounsfield unit) together with measures of plaque burden (percentage vessel atheroma) were quantified on CCTA. The corresponding coronary atherosclerotic lesion was analysed on T1-weighted black-blood iT2prep-BOOST CMR to assess cross-sectional plaque area, burden, and plaque-to-myocardial signal intensity ratio (PMR). RESULTS:A total of 188 lesions in 85 patients were analysed and compared between iT2prep-BOOST and CCTA. Plaque burden estimates derived from iT2prep-BOOST demonstrated moderate correlation with CCTA-derived total plaque burden (r = 0.58) and non-calcified plaque burden (r = 0.54). However, agreement between modalities was limited. For total plaque burden, the mean difference was -4% (95% CI: -7% to -2%) with limits of agreement ranging from -38% to 29%. For non-calcified plaque burden, the mean difference was 3% (95% CI: 0% to 5%) with limits of agreement from -34% to 39%. In univariate regression analyses, PMR was associated with increasing total, non-calcified, and low-density non-calcified plaque areas and decreasing CT attenuation values. CONCLUSION:Coronary plaque burden assessed by iT2prep-BOOST demonstrated modest correlation and limited agreement with CCTA-derived plaque burden, indicating that the two modalities are not directly interchangeable for plaque burden quantification. Nevertheless, the observed associations between PMR and adverse CCTA plaque characteristics support the potential role of T1-weighted CMR as a complementary, non-invasive tool for coronary plaque characterization.
BACKGROUND:Left ventricular remodeling predicts cardiovascular mortality in chronic kidney disease (CKD). Prior studies have linked renal functional markers to cardiac structural changes, but the direct relationship between cardiac MRI parameters and renal interstitial fibrosis remains uncharacterized. We aimed to investigate the association between cardiac remodeling and biopsy-assessed renal interstitial fibrosis in patients with CKD. MATERIALS & METHODS:This prospective single-center cohort included CKD patients with native kidneys who underwent same-week cardiac MRI and renal biopsy. Renal interstitial fibrosis was quantified as a percentage of the affected cortex area. Five cardiac parameters were selected: LVMi, LVEDVi, GLS, GCS, and myocardial T1. Associations with renal fibrosis were assessed using Spearman correlations, and multivariable linear regression adjusted for age and estimated glomerular filtration rate (eGFR). Sensitivity analyses were performed after excluding outliers. RESULTS:Forty-seven patients were included (median age, 49 years; 55% male). LVMi correlated with renal fibrosis (ρ = 0.49; p < 0.001). GLS (ρ = 0.33; p = 0.03) and GCS (ρ = 0.30; p = 0.04) showed consistent trends that did not survive statistical correction. Renal fibrosis remained associated with LVMi after adjustment for age and eGFR (β = 0.33; p < 0.001). Adding LVEDVi attenuated the fibrosis coefficient by 22% (β = 0.26; p < 0.001). The renal fibrosis - left ventricular mass association persisted after excluding patients with high LVMi or high fibrosis grades (ρ = 0.45, p = 0.002; ρ = 0.35, p = 0.026, respectively). CONCLUSION:In CKD patients, LV mass was associated with biopsy-assessed renal fibrosis, independently of eGFR, suggesting that structural renal injury provides information beyond kidney function alone. This association was only partially attenuated after adjustment for LV volume, indicating that volume-related hemodynamic factors may not fully account for the link between renal fibrosis and LV remodeling.
BACKGROUND:Heart failure with preserved ejection fraction (HFpEF) is a hemodynamically heterogeneous syndrome including stage C HFpEF (overt) and exercise-induced HFpEF. While impaired compliance reserve is well recognized, contractile reserve remains underexplored. In addition, distinguishing exercise-induced HFpEF from non-cardiac dyspnea (NCD) remains challenging. Exercise cardiovascular magnetic resonance (Ex-CMR) enables quantification of reserve metrics, yet standardizing exercise protocol in dyspneic patients with limited capacity is challenging. This study aimed to establish an analytic framework for deriving non-invasive compliance and contractile reserve metrics independent of fixed exercise thresholds using Ex-CMR, and to evaluate their utility for pathophysiological differentiation among NCD, exercise-induced HFpEF, and stage C HFpEF. METHODS:We proposed a non-invasive work-volume (W-V) loop model from Ex-CMR using left ventricular end-diastolic (LVEDV) and end-systolic volumes (LVESV) at rest and stress, along with maximum workload during supine cycle ergometer exercise, to construct a trapezoidal loop. The loop's base angles θ₁ and θ₂ represent effort-adjusted compliance and contractile reserve, respectively. In a retrospective analysis of a prospective multi-center Ex-CMR study, these markers were calculated in healthy controls, NCD, exercise-induced HFpEF, and stage C HFpEF groups. Patient cohorts were defined based on invasive hemodynamic exercise-testing thresholds and non-invasive data. ANOVA and post hoc testing were performed. Reproducibility was assessed. RESULTS:Among 120 participants (40 healthy controls, 27 NCD, 20 exercise-induced HFpEF, 33 stage C HFpEF), effort-adjusted compliance reserve was impaired in both HFpEF subgroups compared with healthy controls and NCD (p<0.0001). Effort-adjusted contractile reserve was significantly higher in exercise-induced HFpEF than in healthy controls, NCD, and stage C HFpEF (p=0.018, 0.006, and <0.0001, respectively), despite comparable absolute ΔLVESV to NCD and lower ΔLVESV than healthy controls. In contrast, stage C HFpEF demonstrated depressed contractile reserve compared with other groups (p<0.0001). Reproducibility of reserve markers was good to excellent. CONCLUSIONS:Ex-CMR-derived W-V loop geometry reveals distinctive features across HFpEF subgroups and provides pathophysiological insight from an imaging perspective, demonstrating universal impairment of compliance reserve in HFpEF and a newly identified hypercontractile profile characteristic of exercise-induced HFpEF. Elevated θ₂ in exercise-induced HFpEF may identify a window for intervention before contractile reserve becomes compromised.
AIMS:Cardiac MRI is central to evaluating ventricular function and hemodynamics in pediatric congenital heart disease (CHD). Conventional two-dimensional phase-contrast (2D-PC) imaging is challenged by fixed planes, operator dependence, and multiple breath-holds, where four-dimensional flow (4DF) MRI overcomes these challenges. This study compares accelerated whole heart 4D Flow (WH-4DF) MRI with 2D-PC and volumetric measurements in routine clinical follow up of pediatric CHD and highlights its applicability. METHODS AND RESULTS:For this prospective study seventy-one consecutive pediatric patients (median age 14 ± 2.4 years; 49 male) with surgically corrected CHD underwent both 2D-PC and accelerated WH-4DF MRI. Planning and acquisition times and overall success rate were recorded. Flow, velocity and volumetric measurements were compared using Bland-Altman analysis, orthogonal regression, and intraclass correlation coefficients. WH-4DF showed excellent agreement with 2D-PC and short-axis volumetry for aortic, pulmonary, and ventricular stroke volumes (mean differences <5%). Mean planning and acquisition time for WH-4DF was 10.3minutes (SD 1.1), where planning and acquisition time of 2D-PC was 11.2minutes (SD 6.25). Aliasing was observed in 11% of the WH-4DF acquisitions but did not compromise interpretability. CONCLUSIONS:Accelerated WH-4DF MRI is clinically robust and broadly applicable across CHD, including complex postoperative anatomies. It provides comparable numbers to 2D-PC, reproducible, and time-efficient flow quantification with superior coverage and reduced operator dependence. A single WH-4DF scan can replace multiple 2D-PC acquisitions with comparable results and may enhance assessment of regurgitant flow. These features underscore its potential role in routine clinical practice and in advancing our understanding of disease processes.
BACKGROUND:Cardiovascular magnetic resonance (CMR) is the reference standard for assessing cardiac function, yet its widespread clinical use remains challenged by time-intensive workflows requiring ECG gating, repetitive breath-holding, and expert planning. The free-running framework (FRF) addresses these barriers by enabling respiratory and cardiac motion resolved (5D) whole-heart imaging without ECG, breath-holds, or expert scan plane planning. Its fast interrupted steady-state (FISS) variant further enables high quality clinical imaging with gadolinium-based or ferumoxytol contrast agents by providing balanced steady-state free precession (bSSFP)-like contrast with intrinsic fat suppression. The combined 5D FISS-FRF approach, therefore, enables 3D, motion-resolved, free-breathing whole-heart imaging with efficient fat suppression. Early single-center studies have demonstrated feasibility and high concordance with conventional 2D cine imaging, but its generalizability and clinical utility in a multi-center, multi-vendor setting is not yet established. METHODS/DESIGN:FAST-CMR (Assessment and validation of the established free-running framework for cardiac function by magnetic resonance imaging) is a prospective, observational, intra-individually controlled, pragmatic multi-center study enrolling 300 patients with cardiac disease across 21 international sites from 6 continents using 1.5T MR systems of both Siemens Healthineers and Philips Healthcare. Patients will undergo conventional 2D cine bSSFP CMR in both short-axis and long-axis orientations and a post-contrast 5D FISS-FRF acquisition of six minutes in length. Stratification will include equal representation across three disease cohorts: patients with congenital heart disease (CHD), patients unable to breath-hold, and patients who are able to breath-hold. 5D FISS-FRF raw image acquisition data will undergo centralized reconstruction at the coordinating center (CHUV) before blinded analysis at the core laboratory (Mayo Clinic). The primary endpoint is the precision of the mean paired difference in left ventricular ejection fraction (LVEF) between 5D FISS-FRF and conventional 2D cine CMR, assessed on a within-subject basis. Agreement will be further characterized using confidence intervals for the mean difference and Bland-Altman analysis. Secondary endpoints include detection of regional wall-motion abnormalities, image quality, scan efficiency, patient comfort, LV mass, left and right atrial volumes, and feasibility of automated post-processing and artificial intelligence (AI)-based reconstruction. DISCUSSION/CONCLUSION:FAST-CMR will provide the first prospective international multi-center, multi-vendor evaluation of 5D FISS-FRF. By combining standardized acquisition, centralized reconstruction, and harmonized analysis across 21 international sites and two vendor systems, the study will assess the generalizability and robustness of ventricular functional measurements obtained with 5D FISS-FRF. Using a pre-specified precision-based framework for global ventricular function and complementary agreement analyses, FAST-CMR will evaluate whether 5D FISS-FRF can provide measurements that are comparable to conventional 2D cine across diverse clinical settings, while potentially offering improvements in workflow, scan efficiency and patient experience. These results will inform future studies assessing clinical interchangeability with conventional 2D cine imaging and implementation in routine CMR practice.
BACKGROUND:Anderson-Fabry Disease (AFD) is a lysosomal storage disorder characterized by glycosphingolipid accumulation, leading to multi-organ failure. Cardiac involvement, marked by left ventricular (LV) hypertrophy and fibrosis, is a major cause of mortality. Myocardial inflammation is a frequent finding in these patients, but its role across disease stages remains unclear. This pilot study investigates the association between myocardial inflammation, assessed via cardiac magnetic resonance (CMR) T2 mapping (T2m), and cardiac damage progression, measured by late gadolinium enhancement (LGE) expansion. METHODS:This retrospective, single-center study, includes 131 paired CMR scans (1.5T, Philips Ingenia) collected between June 2018 and December 2023 from 41 AFD patients. Cine imaging, T2m, and LGE sequences were acquired in all. Scans were categorised based on quantitative LGE changes into three groups: LGE- (no LGE), LGE= (stable LGE compared to the previous scan), and LGE+ (new/increased LGE compared to the previous scan). T2m was measured globally, in the basal slice (B T2m), infero-lateral basal segment (ILB T2m), and in LGE-matching regions. RESULTS:The sample consisted of 61% women (average of 3±1 scans/patient); at baseline, mean age was 44±18 years, the LV mass was 58±24g/m2. The average time between CMR scans was 14±4 months. In LGE-matching regions, T2m was significantly higher in LGE+ scans compared to LGE= (62ms vs 56ms, p=0.03), and ILB T2m was higher in LGE+ than LGE- (56ms vs 50ms, p=0.007). No other significant differences were observed. CONCLUSION:In AFD patients, we demonstrated an association between myocardial inflammation and LGE expansion, with T2m potentially serving as an early marker of disease activity. Region-specific T2m measurement is particularly useful in this disease Further research is needed to understand whether T2m changes precedes structural myocardial damage and assess T2m's prognostic value in AFD.
BACKGROUND:Age-related structural and functional remodeling of the heart and vessels increases cardiovascular disease (CVD) risk, yet comprehensive assessments using multimodal imaging and genetic characterization remains limited. We aimed to quantify cardiovascular aging using multimodal biomarkers and evaluate its genetic architecture, lifestyle determinants, and prognostic relevance. METHODS:From the UK biobank, cardiovascular magnetic resonance (CMR), electrocardiogram, arterial stiffness, and carotid ultrasound biomarkers were integrated to establish a cardiovascular aging measure using machine learning (ML) models. Seven ML models were evaluated to predict cardiovascular age, and cardiovascular age gap (CardioAG) was calculated using the best-performing model. Associations between CardioAG and incident CVD outcomes were assessed, alongside modality-specific analyses to evaluate the incremental value of multimodal integration. Whole-genome sequencing (WGS) analyses were conducted to identify genetic variants associated with CardioAG, and linear regression models were applied to examine relationships between CardioAG and key lifestyle factors. RESULTS:Among 22,452 participants with complete multimodal data, 13,694 individuals free of baseline CVDs (median age 62.6 years [IQR 56.5-68.3], median follow-up 4.8 years [IQR 3.7-6.3]) were selected for model development. With 57 cardiovascular aging biomarkers, the CatBoost model performed best on the test dataset (Pearson r = 0.75; mean absolute error = 3.88 years). CardioAG was independently associated with hypertension, stroke, atrial fibrillation, coronary artery disease, and composite major adverse cardiovascular event (MACE, hazard ratio = 1.08, 95% CI, 1.06-1.10). Incremental and ablation analyses demonstrated complementary contributions across imaging and functional modalities, with multimodal integration enhancing predictive accuracy and providing independent prognostic value for MACE. WGS analysis identified novel common genetic variants associated with interindividual variability in CardioAG, including RN7SKP155, SVIL, and CBFA2T3, highlighting genetic contributions to vascular remodeling, electrophysiological and hemodynamic regulation, and cardiovascular functional reserve. Dietary factors, sleep duration, physical activity level, smoking, and alcohol intake were found to be significantly associated with CardioAG. CONCLUSIONS:This study developed a unified multimodal cardiovascular aging metric that integrates cardiac structure and function, vascular remodeling and stiffness, and electrophysiological features into a single biologically grounded, cumulative aging signal, providing incremental prognostic utility for CVD, and broadening the biological and genetic landscape underlying cardiovascular aging.
BACKGROUND:Conventional cardiovascular magnetic resonance (CMR) in pediatric and congenital heart disease uses 2D, breath-hold (BH), balanced steady state free precession (bSSFP) cine imaging for assessment of function, in addition to cardiac-gated, respiratory-navigated, static 3D bSSFP whole-heart imaging for anatomical assessment. Our aim is to concatenate a stack of 2D free-breathing real-time cines and use Deep Learning (DL) to create an isotropic fully segmented 'pseudo' 3D-cine dataset from these images. METHODS:Four DL models were trained on open-source data that performed: a) Interslice signal-correction; b) Interslice respiratory-correction; c) Super-resolution in the slice direction; and d) Segmentation of right and left atria and ventricles (RA, LA, RV, and LV), thoracic aorta (Ao) and pulmonary arteries (PA). Our method was validated in 20 patients undergoing routine cardiovascular examination, by converting prospectively acquired sagittal stacks of real-time cine images to segmented, isotropic pseudo 3D-cine data. Quantitative metrics (ventricular volumes and vessel diameters) and image quality of the DL pseudo-3D-cines were compared to reference-standard breath-hold cine and whole-heart imaging. RESULTS:All real-time data were successfully transformed into pseudo 3D-cines with a total offline reconstruction and post-processing time of <1min in all cases. There were no significant biases in any left ventricular (LV) or right ventricular (RV) metrics (bias ± standard deviation in ml, LV end diastolic volume (EDV): 0.7 ± 8.8, LV end systolic volume (ESV): -1.7 ± 7.1, RV EDV: 1.9 ± 12.4, RV ESV: -1.5 ± 10.7) with reasonable limits of agreement and correlation. There is also reasonable agreement for all vessel diameters, although there was a small but significant overestimation (p<0.05) of right PA (RPA) and main PA (MPA) diameter (RPA: bias = -1.0mm. MPA: bias = -1.1mm). The DL pseudo-3D-cine data were assessed to be of adequate diagnostic quality unlike the unprocessed 2D real-time data. CONCLUSION:We have demonstrated the potential of creating a pseudo 3D-cine data from concatenated 2D real-time cine images using a series of DL models. Our method has short acquisition and reconstruction times with fully segmented data being available in less than one minute. Our models are trained from fully open-source datasets, allowing our technique to be easily shared with other clinical centers. The agreement with reference-standard imaging suggests that our method could help to significantly speed up CMR in clinical practice.
Background Aortic wall shear stress (WSS) maps allow three-dimensional visualization and quantification of WSS patterns in the aorta and can be used to identify patch-wise differences between cohorts. However, a method addressing the multiplicity of the testing should be considered when investigating patch-wise differences. In this study, we sought to explore the capability of permutation tests in addressing the regional statistical differences in WSS values between patients with mild-to-moderate aortic dilation with tricuspid valves and age- and sex- matched controls. Methods 4D Flow MRI-derived WSS parameters were computed in a cohort of 46 patients with mild-to-moderate ascending aortic dilation ("cases") and a cohort of 51 age- and sex-matched individuals without dilation ("controls"). The 3D WSS maps were mapped to a shared geometry where the ascending aorta was divided into 40×40 surface patches. Synthetic data were created and used to assess the false positive and false negative rates for permutation tests with threshold-free cluster enhancement (TFCE) and, for comparison, for statistical tests common in cardiovascular studies, namely the Student’s t-test (with and without Bonferroni correction and Benjamin-Hochberg procedure), and the Wilcoxon rank sum test. Results Significance maps obtained from synthetic data showed that permutation tests had lower false positive rates compared to the uncorrected Student’s t-test and the Wilcoxon rank sum test, lower false negative rates compared to the Student’s t-test with Bonferroni correction, but higher false negative rates compared to the Student’s t-test with the Benjamin-Hochberg procedure. However, the false negatives were concentrated near the boundary of regions of no difference between the groups. On the real data, several regions, indicated as significant by the non-corrected Student’s t-test, were not significant when statistical testing was corrected for multiple comparisons. Significance maps created with TFCE-based permutation tests reveal that individuals with mild-to-moderate aortic dilation and tricuspid valves have lower peak WSS, higher OSI, and higher WSS in diastole. Conclusion Permutation tests applied to 4D Flow-derived WSS parameters are suitable for the local analysis of WSS differences between cohorts, correcting for the family-wise error rate while accounting for spatial relationships between WSS values.
AIMS:Ischemic heart disease (IHD) patients undergo cardiovascular alterations that can accelerate heart ageing. Estimating biological heart age using advanced cardiac magnetic resonance (CMR) and electrocardiogram (ECG)-derived phenotypes provides a biomarker for heart ageing. We investigated the relationship of IHD and heart ageing using biological age estimation biomarkers, and the contribution of conventional cardiac imaging indices and vascular risk factors (VRFs). METHODS AND RESULTS:Heart age was estimated in prevalent IHD cases (n = 2,142) incorporating CMR radiomics and ECG features. Heart age gap (HAG), representing the disparity between predicted and actual heart age, was calculated. IHD subjects had significantly higher heart age compared to those without the disease (HAG: 1.55 years ± 5.66; p <0.001). The main radiomics and ECG features linked to heart ageing in IHD delineated a phenotype of structural and electrical remodelling not typically seen in the ageing heart. Conventional CMR indices accounted for only a negligible fraction of the association between IHD and HAG. Among the VRFs, adiposity and hypertension were significantly associated with increasing HAG in IHD. CONCLUSION:Individuals with IHD showed higher estimated heart ageing than controls, consistent with greater deviation from normal heart ageing, and this was associated with selected VRFs. This relationship was only minimally explained by conventional CMR indices, suggesting that the heart age model captures additional imaging and electrical features beyond standard CMR measures. HAG may offer an exploratory framework for characterising phenotypic heterogeneity in IHD and contextualising cardiovascular risk; however, prospective validation is required prior to clinical application.
BACKGROUND:Cardiac magnetic resonance feature tracking (CMR-FT) of left atrial (LA) strain is hindered by thin-wall contouring errors, motion heterogeneity, and temporal drift, while manual or landmark-based methods lack reproducibility and scalability. METHODS:We retrospectively collected a multi-center, two-vendor cine MRI dataset. A multi-task learning model was developed to quantify LA strain directly from two-chamber and four-chamber cine images, by coupling a groupwise registration network with a segmentation network through a spatiotemporal cross-attention module and synergistic losses. Performance was benchmarked against common feature tracking algorithms, including optical flow, pairwise registration, and VoxelMorph, via various metrics such as mean-squared error, contour distance, mitral annular tracking accuracy, and drift error. Diagnostic performance to distinguish healthy from diseased subjects was assessed by ROC analysis. RESULTS:546 subjects (142 healthy; age 49±18 years; 343 male) were included for method development and internal/external testing. The proposed method outperformed all other methods in tracking accuracy and reduced the drift effect commonly observed in optical flow and pairwise registration to a level comparable to fixed-reference learning-based registration. Inference required half a second. Automatic strains agreed closely with manual-segmentation-derived values (reservoir r=0.95, conduit r=0.96, booster r=0.92; all p<0.001). In the external dataset, all three strain components were lower in diseased subjects than normal controls (reservoir 24.4±13.2% vs 45.7±12.6%, conduit 14.1±8.8% vs 31.1±10.2%, and booster 10.3±6.4% vs 14.6±5.1%, all p<0.001). Compared with alternative methods, the automatic reservoir strain achieved the highest discriminative power across multiple diseased groups (AUC: 0.81-0.97). CONCLUSION:A fully automatic, multi-task learning framework for LA strain quantification, validated in multi-center two-vendor data, enhances tracking accuracy and speed over prior methods, enabling rapid, scalable atrial function assessment in routine care. Source code is available at SJTU-CMRLab/Dual_Task_LA_Strain_Quantification.
BACKGROUND:Although transcatheter aortic valve replacement (TAVR) can improve cerebral blood flow (CBF) in patients with aortic stenosis (AS), the effect of pre-TAVR flow rate on post-TAVR flow rate increase is unclear. We aimed to evaluate blood flow changes before and after TAVR in the ascending aorta (AAo) and brachiocephalic artery (BCA) using four-dimensional flow cardiovascular magnetic resonance (4D flow CMR) and assess whether the post-TAVR flow response depends on the pre-TAVR flow rate. METHODS:This prospective study included patients who underwent TAVR for AS at our institution between January 2022 and December 2023 and who underwent 4D flow CMR before and after TAVR. Patients were divided into two groups, normal-flow (NF) and low-flow (LF) groups, using a cutoff value of 35mL/beat/m2 for the ratio of pre-TAVR AAo flow rate per beat to body surface area. Flow rates in the AAo and BCA were analyzed. RESULTS:Sixty-one patients were enrolled. The median age was 84 (82-87) years, and twenty-three patients (38%) were men. The net flow rates in the AAo and BCA increased after TAVR, but the changes were not statistically significant (AAo, 3206 (2499-3975) mL/min vs. 3591 (2880-4419) mL/min, p = 0.23; BCA, 435 (302-560) mL/min vs. 512 (377-666) mL/min, p = 0.06). In the analysis of the NF and LF groups, the LF group demonstrated significant increases in net flow after TAVR (AAo, 2598 (1972-3079) vs. 3468 (2900-4131) mL/min, p < 0.001; BCA, 365 (230-529) vs. 522 (377-659) mL/min, p = 0.006). CONCLUSION:4D flow CMR revealed that TAVR significantly increased the AAo and BCA flow rates in patients with LF AS. (274 words / 350 limit).
BACKGROUND:Volumetric, one-stop cardiovascular magnetic resonance parametric mapping techniques have been developed and continue to evolve; however, clinical adoption remains limited by insufficient evidence supporting their feasibility in routine practice and the need to demonstrate agreement with established standard techniques. PURPOSE:To validate three-dimensional free-breathing mSAVA for comprehensive myocardial tissue characterization, including pre-/post-contrast T1, T2, ECV, and synthesized bright-blood (BB-) and dark-blood (DB-) LGE in patients, and to compare its performance with conventional T1 mapping, T2 mapping, and phase-sensitive inversion recovery (PSIR) techniques. METHODS:We deployed mSAVA in a clinical hospital setting and had it operated by radiologists. A total of 40 patients (22 males, age range: 17-67 years) with various cardiac diseases referred for CMR evaluation underwent mSAVA imaging along with standard clinical sequences, including MOLLI for native/post-contrast T1 mapping, mGraSE for T2 mapping, and PSIR. The study was approved by the institutional review board, and written informed consent was obtained from all participants before imaging. We evaluated the clinical feasibility of mSAVA, its agreement with established conventional techniques, and its ability to synthesize LGE. RESULTS:Of 40 enrolled patients, 2 had non-diagnostic mSAVA images and 7 lacked post-contrast scans or had atypical cardiac anatomy. In the remaining 31 patients (17 male; age range, 17-67 years), mSAVA required 6.1±1.7min of free-breathing acquisition to obtain whole-heart joint T1/T2 maps, compared with 1.5min for MOLLI T1 mapping and 1.7min for mGraSE T2 mapping over three slices. mSAVA showed strong correlation for post-contrast T1 and ECV, and moderate correlation for native T1 and T2 with conventional clinical standards, and similarly characterized scar-remote tissue differences. The 3D post-contrast T1/T2 maps also enabled retrospective synthesis of bright-blood and dark-blood LGE. Synthesized BB-LGE demonstrated good accuracy for detecting enhancement compared with PSIR, whereas synthesized DB-LGE showed moderate precision and variable scar depiction. CONCLUSION:This initial clinical study suggests that free-breathing, one-stop mSAVA is a promising technique for comprehensive CMR tissue characterization, particularly in patients unable to perform breath-holds, in scenarios requiring multiple quantitative measurements, or when whole-heart coverage is desired. Further clinical studies are warranted to establish its references and its clinical value.
BACKGROUND:4D flow MRI hemodynamic biomarkers (e.g. peak velocity and wall shear stress [WSS]) have shown promise for improved risk-stratification in patients with thoracic aortic disease (TAD). However, lengthy 4D flow scan times, complex data analysis and need for dedicated expertise limit clinical translation. In this study, we developed a fluid-physics informed deep generative adversarial neural network, CE-MRA-FLOWnet, to predict aortic hemodynamics directly from standard-of-care contrast-enhanced magnetic resonance angiography (CE-MRA) images. METHODS:We retrospectively identified 1392 patients (age: 52±12years, 1011 male, 954 with bicuspid aortic valve, BAV; 438 with tricuspid aortic valve, TAV) who underwent paired clinical 4D flow MRI and CE-MRA between 2011 and 2020 for suspected TAD. The CE-MRA-FLOWnet used CE-MRA data (1127 for network training, 265 for testing) to generate a prediction of aortic hemodynamics. 4D flow-measured aortic systolic 3D blood flow velocity vector fields served as ground truth. Analysis included comparison of AI-derived and ground truth aortic systolic peak velocity (PV) and WSS, the relative area of the AAo (%) exposed to elevated WSS, and aortic valve stenosis severity grading. The predictive value of CE-MRA-FLOWnet hemodynamics metrics was assessed in a subgroup of 133 BAV patients with known multi-year adverse outcomes. RESULTS:CE-MRA-FLOWnet training time was 8100mins; inference time per CE-MRA was 0.88±0.05seconds. AI-derived PV showed strong regional agreement and low biases with ground truth 4D flow (0.00-0.04m/s), and relative differences within 10.6%-12.7%. Aortic WSS also demonstrated minimal bias (-0.01-0.04Pa) and close alignment of the area of elevate WSS between AI and 4D flow (19.6±15.1% vs. 19.3±16.2%, p=0.84). AS severity was accurately classified in 88% of cases, with all grading errors limited to a one-grade difference (Kappa 0.84). ROC analysis showed that CE-MRA-FLOWnet derived hemodynamic metrics outperformed diameter alone for predicting adverse outcomes (PV AUC = 0.73-0.76; WSS = 0.83-0.88; diameter = 0.52-0.62). CONCLUSIONS:CE-MRA-FLOWnet accurately predicts aortic hemodynamics in TAD patients using standard-of-care CE-MRA images. Our results demonstrate potential for clinical integration by providing physicians with near real-time hemodynamic data from widely available clinical CE-MRA images. CLINICAL PERSPECTIVE:Approximately 3% of the population has or is at risk for thoracic aortic disease (TAD) which can result in significant complications including progressive aortic dilation and dissection. Historically, aortic diameter measured on CT angiography or MR angiography (MRA) has been the primary marker of risk and TAD patients undergo frequent surveillance imaging to evaluate aortic size and growth. However, it is well established that aortic diameter is an imperfect risk assessment tool and a large percentage of patients who are outside of guideline-endorsed diameter surgical thresholds can have aortic dissection. Recently, aortic hemodynamics measured with 4D flow MRI have shown promise for detecting patients who have higher aortic growth rates and 4D flow derived wall shear stress (WSS) has been linked to histopathologic damage to the aortic wall. Thus, aortic hemodynamic assessment could supplement aortic diameter to improve TAD risk-stratification. Unfortunately, 4D flow MRI is not widely available and can be difficult acquire, analyze, and interpret which has led to low utilization. In this study, we have developed CE-MRA-FLOWnet, a fluid-physics informed generative neural network that quantifies peak systolic aorta hemodynamics using only standard anatomic contrast-enhanced (CE) MRA as input. We found that CE-MRA-FLOWnet peak systolic velocity and WSS quantification is highly accurate compared to 4D flow MRI. We also found that these outputs can accurately grade aortic stenosis severity and are superior to aortic diameter for predicting adverse aortic outcomes. CE-MRA-FLOWnet could significantly expand access these important risk metrics to many more TAD patients.
BACKGROUND:Reperfusion injury after primary percutaneous coronary intervention is a potential target for improving treatment of myocardial infarction. Reperfusion represents a sudden metabolic challenge of the ischemic myocardial tissue, which is thought to be a driver of reperfusion injury that is difficult to evaluate with current technologies. METHODS:We investigated the use of metabolic magnetic resonance imaging with hyperpolarized [1-13C]pyruvate for investigation of post-reperfusion metabolism in a pig model of acute myocardial infarction. To induce ischemia, 15 pigs had a coronary balloon placed in the left anterior descendent for one hour. Seven animals underwent mechanical preload reduction in effort to investigate the effects of increased salvage. After three hours of reperfusion, the animals were scanned with hyperpolarized [1-13C]pyruvate MRI to image the balance between mitochondrial and glycolytic metabolism. RESULTS:Larger conversion from [1-13C]pyruvate to [1-13C]lactate, representing glycolytic metabolism, was observed in the area-at-risk (P <.001). Similar, but non-significant, hypermetabolism was observed for conversion of [1-13C]pyruvate to [13C]bicarbonate (P =.07), representing mitochondrial metabolism. Conversion from [1-13C]pyruvate to [1-13C]alanine seemed to decrease in the group with preload reduction but increase in the sham group. The observed hypermetabolism correlated with the amount of salvaged myocardium (r =.7, P <.01), but not with the size of the area-at-risk. CONCLUSION:This study shows that the salvaged myocardium becomes hypermetabolic and glycolytic after reperfusion, and that a metabolic MRI may be a valuable tool in attempts to modulate post-reperfusion metabolism in preclinical and clinical research alike.
