Infarct quantification with cardiovascular magnetic resonance using "standard deviation from remote" is unreliable: validation in multi-centre multi-vendor data

Noninvasive Pressure-Volume Loops Predict Major Adverse Cardiac Events in Heart Failure With Reduced Ejection Fraction

Background

Heart failure with reduced ejection fraction (HFrEF) is characterized by ventricular remodeling and impaired myocardial energetics. Left ventricular pressure-volume (PV) loop analysis can be performed noninvasively using cardiovascular magnetic resonance (CMR) imaging to assess cardiac thermodynamic efficiency.

Objectives

The aim of the study was to investigate whether noninvasive PV loop parameters, derived from CMR, could predict major adverse cardiac events (MACE) in HFrEF patients.

Methods

PV loop parameters (stroke work, ventricular efficiency, external power, contractility, and energy per ejected volume) were computed from CMR cine images and brachial blood pressure. The primary end point was MACE (cardiovascular death, heart failure (HF) hospitalization, myocardial infarction, revascularization, ventricular tachycardia/fibrillation, heart transplantation, or left ventricular assist device implantation within 5 years). Associations between PV loop parameters and MACE were evaluated using multivariable Cox regression.

Results

One hundred and sixty-four HFrEF patients (left ventricular ejection fraction ≤40%, age 63 [IQR: 55-70] years, 79% male) who underwent clinical CMR examination between 2004 and 2014 were included. Eighty-eight patients (54%) experienced at least one MACE after an average of 2.8 years. Unadjusted models demonstrated a significant association between MACE and all PV loop parameters (P < 0.05 for all), HF etiology (P < 0.001), left ventricular ejection fraction (P = 0.003), global longitudinal strain (P < 0.001), and N-terminal prohormone of brain natriuretic peptide level (P = 0.001). In the multivariable Cox regression analysis adjusted for age, sex, hypertension, diabetes, and HF etiology, ventricular efficiency was associated with MACE (HR: 1.04 (95% CI: 1.01-1.08) per-% decrease, P = 0.01).

Conclusions

Ventricular efficiency, derived from noninvasive PV loop analysis from standard CMR scans, is associated with MACE in patients with HFrEF.

Impact of late gadolinium enhancement image acquisition resolution on neural network based automatic scar segmentation

BACKGROUND

Automatic myocardial scar segmentation from late gadolinium enhancement (LGE) images using neural networks promises an alternative to time-consuming and observer-dependent semi-automatic approaches. However, alterations in data acquisition, reconstruction as well as post-processing may compromise network performance. The objective of the present work was to systematically assess network performance degradation due to a mismatch of point-spread function between training and testing data.

METHODS

Thirty-six high-resolution (0.7×0.7×2.0 mm3) LGE k-space datasets were acquired post-mortem in porcine models of myocardial infarction. The in-plane point-spread function and hence in-plane resolution Δx was retrospectively degraded using k-space lowpass filtering, while field-of-view and matrix size were kept constant. Manual segmentation of the left ventricle (LV) and healthy remote myocardium was performed to quantify location and area (% of myocardium) of scar by thresholding (≥ SD5 above remote). Three standard U-Nets were trained on training resolutions Δxtrain = 0.7, 1.2 and 1.7 mm to predict endo- and epicardial borders of LV myocardium and scar. The scar prediction of the three networks for varying test resolutions (Δxtest = 0.7 to 1.7 mm) was compared against the reference SD5 thresholding at 0.7 mm. Finally, a fourth network trained on a combination of resolutions (Δxtrain = 0.7 to 1.7 mm) was tested.

RESULTS

The prediction of relative scar areas showed the highest precision when the resolution of the test data was identical to or close to the resolution used during training. The median fractional scar errors and precisions (IQR) from networks trained and tested on the same resolution were 0.0 percentage points (p.p.) (1.24 - 1.45), and - 0.5 - 0.0 p.p. (2.00 - 3.25) for networks trained and tested on the most differing resolutions, respectively. Deploying the network trained on multiple resolutions resulted in reduced resolution dependency with median scar errors and IQRs of 0.0 p.p. (1.24 - 1.69) for all investigated test resolutions.

CONCLUSION

A mismatch of the imaging point-spread function between training and test data can lead to degradation of scar segmentation when using current U-Net architectures as demonstrated on LGE porcine myocardial infarction data. Training networks on multi-resolution data can alleviate the resolution dependency.

Current use of cardiac MRI in animals

Cardiovascular magnetic resonance (CMR) imaging has evolved to become an indispensable tool in human cardiology. It is a non-invasive technique that enables objective assessment of myocardial function, size, and tissue composition. Recent innovations in magnetic resonance imaging scanner technology and parallel imaging techniques have facilitated the generation of parametric mapping to explore tissue characteristics, and the emergence of strain imaging has enabled cardiologists to evaluate cardiac function beyond conventional metrics. As veterinary cardiology continues to utilize CMR beyond the reference standard, clinical application of CMR will further expand our capabilities. This article describes the current use of CMR and adoption of more recent advances such as T1/T2 mapping in veterinary cardiology.

Late-gadolinium enhancement is common in older pediatric heart transplant recipients and is associated with lower ejection fraction

BACKGROUND

Chronic graft failure and cumulative rejection history in pediatric heart transplant recipients (PHTR) are associated with myocardial fibrosis on endomyocardial biopsy (EMB). Cardiovascular magnetic resonance imaging (CMR) is a validated, non-invasive method to detect myocardial fibrosis via the presence of late gadolinium enhancement (LGE). In adult heart transplant recipients, LGE is associated with increased risk of future adverse clinical events including hospitalization and death. We describe the prevalence, pattern, and extent of LGE on CMR in a cohort of PHTR and its associations with recipient and graft characteristics.

METHODS

This was a retrospective study of consecutive PHTR who underwent CMR over a 6-year period at a single center. Two independent reviewers assessed the presence and distribution of left ventricular (LV) LGE using the American Heart Association (AHA) 17-segment model. LGE quantification was performed on studies with visible fibrosis (LGE+). Patient demographics, clinical history, and CMR-derived volumetry and ejection fractions were obtained.

RESULTS

Eighty-one CMR studies were performed on 59 unique PHTR. Mean age at CMR was 14.8 ± 6.2 years; mean time since transplant was 7.3 ± 5.0 years. The CMR indication was routine surveillance (without a clinical concern based on laboratory parameters, echocardiography, or cardiac catheterization) in 63% (51/81) of studies. LGE was present in 36% (29/81) of PHTR. In these LGE + studies, patterns included inferoseptal in 76% of LGE + studies (22/29), lateral wall in 41% (12/29), and diffuse, involving > 4 AHA segments, in 21% (6/29). The mean LV LGE burden as a percentage of myocardial mass was 18.0 ± 9.0%. When reviewing only the initial CMR per PHTR (n = 59), LGE + patients were older (16.7 ± 2.9 vs. 12.8 ± 4.6 years, p = 0.001), with greater time since transplant (8.3 ± 5.4 vs. 5.7 ± 3.9 years, p = 0.041). These patients demonstrated higher LV end-systolic volume index (LVESVI) (34.7 ± 11.7 vs. 28.7 ± 6.1 ml/m2, p = 0.011) and decreased LV ejection fraction (LVEF) (56.2 ± 8.1 vs. 60.6 ± 5.3%, p = 0.015). There were no significant differences in history of moderate/severe rejection (p = 0.196) or cardiac allograft vasculopathy (CAV) (p = 0.709).

CONCLUSIONS

LV LGE was present in approximately one third of PHTR, more commonly in older patients with longer time since transplantation. Grafts with LGE have lower LVEF. CMR-derived LGE may aid in surveillance of chronic graft failure in PHTR.

Quality control-driven deep ensemble for accountable automated segmentation of cardiac magnetic resonance LGE and VNE images

Background

Late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) imaging is the gold standard for non-invasive myocardial tissue characterisation. However, accurate segmentation of the left ventricular (LV) myocardium remains a challenge due to limited training data and lack of quality control. This study addresses these issues by leveraging generative adversarial networks (GAN)-generated virtual native enhancement (VNE) images to expand the training set and incorporating an automated quality control-driven (QCD) framework to improve segmentation reliability.

Methods

A dataset comprising 4,716 LGE images (from 1,363 patients with hypertrophic cardiomyopathy and myocardial infarction) was used for development. To generate additional clinically validated data, LGE data were augmented with a GAN-based generator to produce VNE images. LV was contoured on these images manually by clinical observers. To create diverse candidate segmentations, the QCD framework involved multiple U-Nets, which were combined using statistical rank filters. The framework predicted the Dice Similarity Coefficient (DSC) for each candidate segmentation, with the highest predicted DSC indicating the most accurate and reliable result. The performance of the QCD ensemble framework was evaluated on both LGE and VNE test datasets (309 LGE/VNE images from 103 patients), assessing segmentation accuracy (DSC) and quality prediction (mean absolute error (MAE) and binary classification accuracy).

Results

The QCD framework effectively and rapidly segmented the LV myocardium (<1 s per image) on both LGE and VNE images, demonstrating robust performance on both test datasets with similar mean DSC (LGE: 0.845 ± 0.075 ; VNE: 0.845 ± 0.071 ; p = n s ). Incorporating GAN-generated VNE data into the training process consistently led to enhanced performance for both individual models and the overall framework. The quality control mechanism yielded a high performance (MAE = 0.043 , accuracy = 0.951 ) emphasising the accuracy of the quality control-driven strategy in predicting segmentation quality in clinical settings. Overall, no statistical difference (p = n s ) was found when comparing the LGE and VNE test sets across all experiments.

Conclusions

The QCD ensemble framework, leveraging GAN-generated VNE data and an automated quality control mechanism, significantly improved the accuracy and reliability of LGE segmentation, paving the way for enhanced and accountable diagnostic imaging in routine clinical use.