Cardiac Remodeling After Hypertensive Pregnancy Following Physician-Optimized Blood Pressure Self-Management: The POP-HT Randomized Clinical Trial Imaging Substudy.

BACKGROUND: Hypertensive pregnancy disorders are associated with adverse cardiac remodeling, which can fail to reverse in the postpartum period in some women. The Physician-Optimized Postpartum Hypertension Treatment trial demonstrated that improved blood pressure control while the cardiovascular system recovers postpartum associates with persistently reduced blood pressure. We now report the effect on cardiac remodeling. METHODS: In this prospective, randomized, open-label, blinded end point trial, in a single UK hospital, 220 women were randomly assigned 1:1 to self-monitoring with research physician-optimized antihypertensive titration or usual postnatal care from a primary care physician and midwife. Participants were 18 years of age or older, with preeclampsia or gestational hypertension, requiring antihypertensives on hospital discharge postnatally. Prespecified secondary cardiac imaging outcomes were recorded by echocardiography around delivery, and again at blood pressure primary outcome assessment, around 9 months postpartum, when cardiovascular magnetic resonance was also performed. RESULTS: A total of 187 women (101 intervention; 86 usual care) underwent echocardiography at baseline and follow-up, at a mean 258±14.6 days postpartum, of which 174 (93 intervention; 81 usual care) also had cardiovascular magnetic resonance at follow-up. Relative wall thickness by echocardiography was 0.06 (95% CI, 0.07-0.05; P<0.001) lower in the intervention group between baseline and follow-up, and cardiovascular magnetic resonance at follow-up demonstrated a lower left ventricular mass (-6.37 g/m2; 95% CI, -7.99 to -4.74; P<0.001), end-diastolic volume (-3.87 mL/m2; 95% CI, -6.77 to -0.98; P=0.009), and end-systolic volume (-3.25 mL/m2; 95% CI, 4.87 to -1.63; P<0.001) and higher left and right ventricular ejection fraction by 2.6% (95% CI, 1.3-3.9; P<0.001) and 2.8% (95% CI, 1.4-4.1; P<0.001), respectively. Echocardiography-assessed left ventricular diastolic function demonstrated a mean difference in average E/E' of 0.52 (95% CI, -0.97 to -0.07; P=0.024) and a reduction in left atrial volumes of -4.33 mL/m2 (95% CI, -5.52 to -3.21; P<0.001) between baseline and follow-up when adjusted for baseline differences in measures. CONCLUSIONS: Short-term postnatal optimization of blood pressure control after hypertensive pregnancy, through self-monitoring and physician-guided antihypertensive titration, associates with long-term changes in cardiovascular structure and function, in a pattern associated with more favorable cardiovascular outcomes. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04273854.

Misclassification of females and males in cardiovascular magnetic resonance parametric mapping: the importance of sex-specific normal ranges for diagnosis of health vs. disease.

AIMS: Cardiovascular magnetic resonance parametric mapping enables non-invasive quantitative myocardial tissue characterization. Human myocardium has normal ranges of T1 and T2 values, deviation from which may indicate disease or change in physiology. Normal myocardial T1 and T2 values are affected by biological sex. Consequently, normal ranges created with insufficient numbers of each sex may result in sampling biases, misclassification of healthy values vs. disease, and even misdiagnoses. In this study, we investigated the impact of using male normal ranges for classifying female cases as normal or abnormal (and vice versa). METHODS AND RESULTS: One hundred and forty-two healthy volunteers (male and female) were scanned on two Siemens 3T MR systems, providing averaged global myocardial T1 and T2 values on a per-subject basis. The Monte Carlo method was used to generate simulated normal ranges from these values to estimate the statistical accuracy of classifying healthy female or male cases correctly as 'normal' when using sex-specific vs. mixed-sex normal ranges. The normal male and female T1- and T2-mapping values were significantly different by sex, after adjusting for age and heart rate. CONCLUSION: Using 15 healthy volunteers who are not sex specific to establish a normal range resulted in a typical misclassification of up to 36% of healthy females and 37% of healthy males as having abnormal T1 values and up to 16% of healthy females and 12% of healthy males as having abnormal T2 values. This paper highlights the potential adverse impact on diagnostic accuracy that can occur when local normal ranges contain insufficient numbers of both sexes. Sex-specific reference ranges should thus be routinely adopted in clinical practice.

Myocardial Involvement After Hospitalization for COVID-19 Complicated by Troponin Elevation: A Prospective, Multicenter, Observational Study.

BACKGROUND: Acute myocardial injury in hospitalized patients with coronavirus disease 2019 (COVID-19) has a poor prognosis. Its associations and pathogenesis are unclear. Our aim was to assess the presence, nature, and extent of myocardial damage in hospitalized patients with troponin elevation. METHODS: Across 25 hospitals in the United Kingdom, 342 patients with COVID-19 and an elevated troponin level (COVID+/troponin+) were enrolled between June 2020 and March 2021 and had a magnetic resonance imaging scan within 28 days of discharge. Two prospective control groups were recruited, comprising 64 patients with COVID-19 and normal troponin levels (COVID+/troponin-) and 113 patients without COVID-19 or elevated troponin level matched by age and cardiovascular comorbidities (COVID-/comorbidity+). Regression modeling was performed to identify predictors of major adverse cardiovascular events at 12 months. RESULTS: Of the 519 included patients, 356 (69%) were men, with a median (interquartile range) age of 61.0 years (53.8, 68.8). The frequency of any heart abnormality, defined as left or right ventricular impairment, scar, or pericardial disease, was 2-fold greater in cases (61% [207/342]) compared with controls (36% [COVID+/troponin-] versus 31% [COVID-/comorbidity+]; P<0.001 for both). More cases than controls had ventricular impairment (17.2% versus 3.1% and 7.1%) or scar (42% versus 7% and 23%; P<0.001 for both). The myocardial injury pattern was different, with cases more likely than controls to have infarction (13% versus 2% and 7%; P<0.01) or microinfarction (9% versus 0% and 1%; P<0.001), but there was no difference in nonischemic scar (13% versus 5% and 14%; P=0.10). Using the Lake Louise magnetic resonance imaging criteria, the prevalence of probable recent myocarditis was 6.7% (23/342) in cases compared with 1.7% (2/113) in controls without COVID-19 (P=0.045). During follow-up, 4 patients died and 34 experienced a subsequent major adverse cardiovascular event (10.2%), which was similar to controls (6.1%; P=0.70). Myocardial scar, but not previous COVID-19 infection or troponin, was an independent predictor of major adverse cardiovascular events (odds ratio, 2.25 [95% CI, 1.12-4.57]; P=0.02). CONCLUSIONS: Compared with contemporary controls, patients with COVID-19 and elevated cardiac troponin level have more ventricular impairment and myocardial scar in early convalescence. However, the proportion with myocarditis was low and scar pathogenesis was diverse, including a newly described pattern of microinfarction. REGISTRATION: URL: https://www.isrctn.com; Unique identifier: 58667920.

Left ventricular anatomy in obstructive hypertrophic cardiomyopathy: beyond basal septal hypertrophy.

AIMS: Obstructive hypertrophic cardiomyopathy (oHCM) is characterized by dynamic obstruction of the left ventricular (LV) outflow tract (LVOT). Although this may be mediated by interplay between the hypertrophied septal wall, systolic anterior motion of the mitral valve, and papillary muscle abnormalities, the mechanistic role of LV shape is still not fully understood. This study sought to identify the LV end-diastolic morphology underpinning oHCM. METHODS AND RESULTS: Cardiovascular magnetic resonance images from 2398 HCM individuals were obtained as part of the NHLBI HCM Registry. Three-dimensional LV models were constructed and used, together with a principal component analysis, to build a statistical shape model capturing shape variations. A set of linear discriminant axes were built to define and quantify (Z-scores) the characteristic LV morphology associated with LVOT obstruction (LVOTO) under different physiological conditions and the relationship between LV phenotype and genotype. The LV remodelling pattern in oHCM consisted not only of basal septal hypertrophy but a combination with LV lengthening, apical dilatation, and LVOT inward remodelling. Salient differences were observed between obstructive cases at rest and stress. Genotype negative cases showed a tendency towards more obstructive phenotypes both at rest and stress. CONCLUSIONS: LV anatomy underpinning oHCM consists of basal septal hypertrophy, apical dilatation, LV lengthening, and LVOT inward remodelling. Differences between oHCM cases at rest and stress, as well as the relationship between LV phenotype and genotype, suggest different mechanisms for LVOTO. Proposed Z-scores render an opportunity of redefining management strategies based on the relationship between LV anatomy and LVOTO.

Artificial Intelligence for Contrast-Free MRI: Scar Assessment in Myocardial Infarction Using Deep Learning-Based Virtual Native Enhancement.

BACKGROUND: Myocardial scars are assessed noninvasively using cardiovascular magnetic resonance late gadolinium enhancement (LGE) as an imaging gold standard. A contrast-free approach would provide many advantages, including a faster and cheaper scan without contrast-associated problems. METHODS: Virtual native enhancement (VNE) is a novel technology that can produce virtual LGE-like images without the need for contrast. VNE combines cine imaging and native T1 maps to produce LGE-like images using artificial intelligence. VNE was developed for patients with previous myocardial infarction from 4271 data sets (912 patients); each data set comprises slice position-matched cine, T1 maps, and LGE images. After quality control, 3002 data sets (775 patients) were used for development and 291 data sets (68 patients) for testing. The VNE generator was trained using generative adversarial networks, using 2 adversarial discriminators to improve the image quality. The left ventricle was contoured semiautomatically. Myocardial scar volume was quantified using the full width at half maximum method. Scar transmurality was measured using the centerline chord method and visualized on bull's-eye plots. Lesion quantification by VNE and LGE was compared using linear regression, Pearson correlation (R), and intraclass correlation coefficients. Proof-of-principle histopathologic comparison of VNE in a porcine model of myocardial infarction also was performed. RESULTS: VNE provided significantly better image quality than LGE on blinded analysis by 5 independent operators on 291 data sets (all P<0.001). VNE correlated strongly with LGE in quantifying scar size (R, 0.89; intraclass correlation coefficient, 0.94) and transmurality (R, 0.84; intraclass correlation coefficient, 0.90) in 66 patients (277 test data sets). Two cardiovascular magnetic resonance experts reviewed all test image slices and reported an overall accuracy of 84% for VNE in detecting scars when compared with LGE, with specificity of 100% and sensitivity of 77%. VNE also showed excellent visuospatial agreement with histopathology in 2 cases of a porcine model of myocardial infarction. CONCLUSIONS: VNE demonstrated high agreement with LGE cardiovascular magnetic resonance for myocardial scar assessment in patients with previous myocardial infarction in visuospatial distribution and lesion quantification with superior image quality. VNE is a potentially transformative artificial intelligence-based technology with promise in reducing scan times and costs, increasing clinical throughput, and improving the accessibility of cardiovascular magnetic resonance in the near future.

MOCOnet: Robust Motion Correction of Cardiovascular Magnetic Resonance T1 Mapping Using Convolutional Neural Networks.

Background: Quantitative cardiovascular magnetic resonance (CMR) T1 mapping has shown promise for advanced tissue characterisation in routine clinical practise. However, T1 mapping is prone to motion artefacts, which affects its robustness and clinical interpretation. Current methods for motion correction on T1 mapping are model-driven with no guarantee on generalisability, limiting its widespread use. In contrast, emerging data-driven deep learning approaches have shown good performance in general image registration tasks. We propose MOCOnet, a convolutional neural network solution, for generalisable motion artefact correction in T1 maps. Methods: The network architecture employs U-Net for producing distance vector fields and utilises warping layers to apply deformation to the feature maps in a coarse-to-fine manner. Using the UK Biobank imaging dataset scanned at 1.5T, MOCOnet was trained on 1,536 mid-ventricular T1 maps (acquired using the ShMOLLI method) with motion artefacts, generated by a customised deformation procedure, and tested on a different set of 200 samples with a diverse range of motion. MOCOnet was compared to a well-validated baseline multi-modal image registration method. Motion reduction was visually assessed by 3 human experts, with motion scores ranging from 0% (strictly no motion) to 100% (very severe motion). Results: MOCOnet achieved fast image registration (<1 second per T1 map) and successfully suppressed a wide range of motion artefacts. MOCOnet significantly reduced motion scores from 37.1±21.5 to 13.3±10.5 (p < 0.001), whereas the baseline method reduced it to 15.8±15.6 (p < 0.001). MOCOnet was significantly better than the baseline method in suppressing motion artefacts and more consistently (p = 0.007). Conclusion: MOCOnet demonstrated significantly better motion correction performance compared to a traditional image registration approach. Salvaging data affected by motion with robustness and in a time-efficient manner may enable better image quality and reliable images for immediate clinical interpretation.

Toward Replacing Late Gadolinium Enhancement With Artificial Intelligence Virtual Native Enhancement for Gadolinium-Free Cardiovascular Magnetic Resonance Tissue Characterization in Hypertrophic Cardiomyopathy.

BACKGROUND: Late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) imaging is the gold standard for noninvasive myocardial tissue characterization but requires intravenous contrast agent administration. It is highly desired to develop a contrast agent-free technology to replace LGE for faster and cheaper CMR scans. METHODS: A CMR virtual native enhancement (VNE) imaging technology was developed using artificial intelligence. The deep learning model for generating VNE uses multiple streams of convolutional neural networks to exploit and enhance the existing signals in native T1 maps (pixel-wise maps of tissue T1 relaxation times) and cine imaging of cardiac structure and function, presenting them as LGE-equivalent images. The VNE generator was trained using generative adversarial networks. This technology was first developed on CMR datasets from the multicenter Hypertrophic Cardiomyopathy Registry, using hypertrophic cardiomyopathy as an exemplar. The datasets were randomized into 2 independent groups for deep learning training and testing. The test data of VNE and LGE were scored and contoured by experienced human operators to assess image quality, visuospatial agreement, and myocardial lesion burden quantification. Image quality was compared using a nonparametric Wilcoxon test. Intra- and interobserver agreement was analyzed using intraclass correlation coefficients (ICC). Lesion quantification by VNE and LGE were compared using linear regression and ICC. RESULTS: A total of 1348 hypertrophic cardiomyopathy patients provided 4093 triplets of matched T1 maps, cines, and LGE datasets. After randomization and data quality control, 2695 datasets were used for VNE method development and 345 were used for independent testing. VNE had significantly better image quality than LGE, as assessed by 4 operators (n=345 datasets; P<0.001 [Wilcoxon test]). VNE revealed lesions characteristic of hypertrophic cardiomyopathy in high visuospatial agreement with LGE. In 121 patients (n=326 datasets), VNE correlated with LGE in detecting and quantifying both hyperintensity myocardial lesions (r=0.77-0.79; ICC=0.77-0.87; P<0.001) and intermediate-intensity lesions (r=0.70-0.76; ICC=0.82-0.85; P<0.001). The native CMR images (cine plus T1 map) required for VNE can be acquired within 15 minutes and producing a VNE image takes less than 1 second. CONCLUSIONS: VNE is a new CMR technology that resembles conventional LGE but without the need for contrast administration. VNE achieved high agreement with LGE in the distribution and quantification of lesions, with significantly better image quality.

Subclinical Changes in Cardiac Functional Parameters as Determined by Cardiovascular Magnetic Resonance (CMR) Imaging in Sleep Apnea and Snoring: Findings from UK Biobank.

Background and Objectives: Obstructive sleep apnea (OSA) is a common disorder with an increased risk for left ventricular and right ventricular dysfunction. Most studies to date have examined populations with manifest cardiovascular disease using echocardiography to analyze ventricular dysfunction with little or no reference to ventricular volumes or myocardial mass. Our aim was to explore these parameters with cardiac MRI. We hypothesized that there would be stepwise increase in left ventricular mass and right ventricular volumes from the unaffected, to the snoring and the OSA group. Materials and Methods: We analyzed cardiac MRI data from 4978 UK Biobank participants free from cardiovascular disease. Participants were allocated into three cohorts: with OSA, with self-reported snoring and without OSA or snoring (n = 118, 1886 and 2477). We analyzed cardiac parameters from balanced cine-SSFP sequences and indexed them to body surface area. Results: Patients with OSA were mostly males (47.3% vs. 79.7%; p < 0.001) with higher body mass index (25.7 ± 4.0 vs. 31.3 ± 5.3 kg/m²; p < 0.001) and higher blood pressure (135 ± 18 vs. 140 ± 17 mmHg; p = 0.012) compared to individuals without OSA or snoring. Regression analysis showed a significant effect for OSA in left ventricular end-diastolic index (LVEDVI) (ß = -4.9 ± 2.4 mL/m²; p = 0.040) and right ventricular end-diastolic index (RVEDVI) (ß = -6.2 ± 2.6 mL/m²; p = 0.016) in females and for right ventricular ejection fraction (RVEF) (ß = 1.7 ± 0.8%; p = 0.031) in males. A significant effect was discovered in snoring females for left ventricular mass index (LVMI) (ß = 3.5 ± 0.9 g/m²; p < 0.001) and in males for left ventricular ejection fraction (LVEF) (ß = 1.0 ± 0.3%; p = 0.001) and RVEF (ß = 1.2 ± 0.3%; p < 0.001). Conclusion: Our study suggests that OSA is highly underdiagnosed and that it is an evolving process with gender specific progression. Females with OSA show significantly lower ventricular volumes while males with snoring show increased ejection fractions which may be an early sign of hypertrophy. Separate prospective studies are needed to further explore the direction of causality.

Sex-specific associations between alcohol consumption, cardiac morphology, and function as assessed by magnetic resonance imaging: insights form the UK Biobank Population Study.

AIMS: Data regarding the effects of regular alcohol consumption on cardiac anatomy and function are scarce. Therefore, we sought to determine the relationship between regular alcohol intake and cardiac structure and function as evaluated with cardiac magnetic resonance imaging. METHODS AND RESULTS: Participants of the UK Biobank who underwent cardiac magnetic resonance were enrolled in our analysis. Data regarding regular alcohol consumption were obtained from questionnaires filled in by the study participants. Exclusion criteria were poor image quality, missing, or incongruent data regarding alcohol drinking habits, prior drinking, presence of heart failure or angina, and prior myocardial infarction or stroke. Overall, 4335 participants (61.5 ± 7.5 years, 47.6% male) were analysed. We used multivariate linear regression models adjusted for age, ethnicity, body mass index, smoking, hypertension, diabetes mellitus, physical activity, cholesterol level, and Townsend deprivation index to examine the relationship between regular alcohol intake and cardiac structure and function. In men, alcohol intake was independently associated with marginally increased left ventricular end-diastolic volume [ß = 0.14; 95% confidence interval (CI) = 0.05-0.24; P = 0.004], left ventricular stroke volume (ß = 0.08; 95% CI = 0.03-0.14; P = 0.005), and right ventricular stroke volume (ß = 0.08; 95% CI = 0.02-0.13; P = 0.006). In women, alcohol consumption was associated with increased left atrium volume (ß = 0.14; 95% CI = 0.04-0.23; P = 0.006). CONCLUSION: Alcohol consumption is independently associated with a marginal increase in left and right ventricular volumes in men, but not in women, whereas alcohol intake showed an association with increased left atrium volume in women. Our results suggest that there is only minimal relationship between regular alcohol consumption and cardiac morphology and function in an asymptomatic middle-aged population.

Standardization of T1-mapping in cardiovascular magnetic resonance using clustered structuring for benchmarking normal ranges.

BACKGROUND: Cardiovascular magnetic resonance T1-mapping is increasingly used for tissue characterization, commonly based on Modified Look-Locker Inversion recovery (MOLLI). However, there are numerous MOLLI variants with differing normal ranges. This lack of standardization presents confusion and difficulty in inter-center comparisons, hindering widespread adoption of T1-mapping. METHODS: To address this, we performed a structured literature search for native left ventricular myocardial T1-mapping in healthy humans measured using MOLLI variants at 1.5 and 3 Tesla, across scanner vendors. We then used k-means clustering to structure normal MOLLI-T1 values according to magnetic field strength, and investigated correlations between common imaging parameters: repetition time (TR), echo time (TE), flip angle (FA). RESULTS: We analyzed data from 2207 healthy controls in 76 independent reports. Normal MOLLI-T1 standard deviations varied by 11-fold, and dependencies on TE, TR, and FA differed between 1.5 T and 3 T, thwarting meaningful T1 standardization even within a single field strength, including the use of Z-score. However, divergent MOLLI-T1 norms may be structured using data clustering. For 1.5 T, two clusters emerged: Cluster11.5T: T1 = 958 ± 16 ms (n = 1280); Cluster21.5T: T1 = 1027 ± 19 ms (n = 386). For 3 T, three clusters emerged: Cluster13T: T1 = 1160 ± 21 ms (n = 330); Cluster23T: T1 = 1067 ± 18 ms (n = 178); Cluster33T: T1 = 1227 ± 19 ms (n = 41). We then propose the concept of an online calculator for assigning local norms to a known MOLLI-T1 cluster, allowing benchmarking against published norms. CONCLUSION: Clustered structuring allows T1 standardization of widely-divergent MOLLI variants, benchmarking local norms (usually based on smaller samples) against published norms (larger samples). This may increase confidence and quality control in method implementation, facilitating wider clinical adoption of T1-mapping.

Deep learning with attention supervision for automated motion artefact detection in quality control of cardiac T1-mapping.

Cardiac magnetic resonance quantitative T1-mapping is increasingly used for advanced myocardial tissue characterisation. However, cardiac or respiratory motion can significantly affect the diagnostic utility of T1-maps, and thus motion artefact detection is critical for quality control and clinically-robust T1 measurements. Manual quality control of T1-maps may provide reassurance, but is laborious and prone to error. We present a deep learning approach with attention supervision for automated motion artefact detection in quality control of cardiac T1-mapping. Firstly, we customised a multi-stream Convolutional Neural Network (CNN) image classifier to streamline the process of automatic motion artefact detection. Secondly, we imposed attention supervision to guide the CNN to focus on targeted myocardial segments. Thirdly, when there was disagreement between the human operator and machine, a second human validator reviewed and rescored the cases for adjudication and to identify the source of disagreement. The multi-stream neural networks demonstrated 89.8% agreement, 87.4% ROC-AUC on motion artefact detection with the human operator in the 2568 T1 maps. Trained with additional supervision on attention, agreements and AUC significantly improved to 91.5% and 89.1%, respectively (p < 0.001). Rescoring of disagreed cases by the second human validator revealed that human operator error was the primary cause of disagreement. Deep learning with attention supervision provides a quick and high-quality assurance of clinical images, and outperforms human operators.

Improving the Generalizability of Convolutional Neural Network-Based Segmentation on CMR Images.

Background: Convolutional neural network (CNN) based segmentation methods provide an efficient and automated way for clinicians to assess the structure and function of the heart in cardiac MR images. While CNNs can generally perform the segmentation tasks with high accuracy when training and test images come from the same domain (e.g., same scanner or site), their performance often degrades dramatically on images from different scanners or clinical sites. Methods: We propose a simple yet effective way for improving the network generalization ability by carefully designing data normalization and augmentation strategies to accommodate common scenarios in multi-site, multi-scanner clinical imaging data sets. We demonstrate that a neural network trained on a single-site single-scanner dataset from the UK Biobank can be successfully applied to segmenting cardiac MR images across different sites and different scanners without substantial loss of accuracy. Specifically, the method was trained on a large set of 3,975 subjects from the UK Biobank. It was then directly tested on 600 different subjects from the UK Biobank for intra-domain testing and two other sets for cross-domain testing: the ACDC dataset (100 subjects, 1 site, 2 scanners) and the BSCMR-AS dataset (599 subjects, 6 sites, 9 scanners). Results: The proposed method produces promising segmentation results on the UK Biobank test set which are comparable to previously reported values in the literature, while also performing well on cross-domain test sets, achieving a mean Dice metric of 0.90 for the left ventricle, 0.81 for the myocardium, and 0.82 for the right ventricle on the ACDC dataset; and 0.89 for the left ventricle, 0.83 for the myocardium on the BSCMR-AS dataset. Conclusions: The proposed method offers a potential solution to improve CNN-based model generalizability for the cross-scanner and cross-site cardiac MR image segmentation task.

Fully Automated Myocardial Strain Estimation from Cardiovascular MRI-tagged Images Using a Deep Learning Framework in the UK Biobank.

PURPOSE: To demonstrate the feasibility and performance of a fully automated deep learning framework to estimate myocardial strain from short-axis cardiac MRI-tagged images. MATERIALS AND METHODS: In this retrospective cross-sectional study, 4508 cases from the U.K. Biobank were split randomly into 3244 training cases, 812 validation cases, and 452 test cases. Ground truth myocardial landmarks were defined and tracked by manual initialization and correction of deformable image registration using previously validated software with five readers. The fully automatic framework consisted of (a) a convolutional neural network (CNN) for localization and (b) a combination of a recurrent neural network (RNN) and a CNN to detect and track the myocardial landmarks through the image sequence for each slice. Radial and circumferential strain were then calculated from the motion of the landmarks and averaged on a slice basis. RESULTS: Within the test set, myocardial end-systolic circumferential Green strain errors were -0.001 ± 0.025, -0.001 ± 0.021, and 0.004 ± 0.035 in the basal, mid-, and apical slices, respectively (mean ± standard deviation of differences between predicted and manual strain). The framework reproduced significant reductions in circumferential strain in participants with diabetes, hypertensive participants, and participants with a previous heart attack. Typical processing time was approximately 260 frames (approximately 13 slices) per second on a GPU with 12 GB RAM compared with 6-8 minutes per slice for the manual analysis. CONCLUSION: The fully automated combined RNN and CNN framework for analysis of myocardial strain enabled unbiased strain evaluation in a high-throughput workflow, with similar ability to distinguish impairment due to diabetes, hypertension, and previous heart attack.Published under a CC BY 4.0 license. Supplemental material is available for this article.

Standardized image post-processing of cardiovascular magnetic resonance T1-mapping reduces variability and improves accuracy and consistency in myocardial tissue characterization.

BACKGROUND: Myocardial T1-mapping is increasingly used in multicentre studies and trials. Inconsistent image analysis introduces variability, hinders differentiation of diseases, and results in larger sample sizes. We present a systematic approach to standardize T1-map analysis by human operators to improve accuracy and consistency. METHODS: We developed a multi-step training program for T1-map post-processing. The training dataset contained 42 left ventricular (LV) short-axis T1-maps (normal and diseases; 1.5 and 3 Tesla). Contours drawn by two experienced human operators served as reference for myocardial T1 and wall thickness (WT). Trainees (n = 26) underwent training and were evaluated by: (a) qualitative review of contours; (b) quantitative comparison with reference T1 and WT. RESULTS: The mean absolute difference between reference operators was 8.4 ±â€¯6.3 ms (T1) and 1.2 ±â€¯0.7 pixels (WT). Trainees' mean discrepancy from reference in T1 improved significantly post-training (from 8.1 ±â€¯2.4 to 6.7 ±â€¯1.4 ms; p < 0.001), with a 43% reduction in standard deviation (SD) (p = 0.035). WT also improved significantly post-training (from 0.9 ±â€¯0.4 to 0.7 ±â€¯0.2 pixels, p = 0.036), with 47% reduction in SD (p = 0.04). These experimentally-derived thresholds served to guide the training process: T1 (±8 ms) and WT (±1 pixel) from reference. CONCLUSION: A standardized approach to CMR T1-map image post-processing leads to significant improvements in the accuracy and consistency of LV myocardial T1 values and wall thickness. Improving consistency between operators can translate into 33-72% reduction in clinical trial sample-sizes. This work may: (a) serve as a basis for re-certification for core-lab operators; (b) translate to sample-size reductions for clinical studies; (c) produce better-quality training datasets for machine learning.

Does self-reported pregnancy loss identify women at risk of an adverse cardiovascular phenotype in later life? Insights from UK Biobank.

INTRODUCTION: Cardiovascular disease (CVD) is more common in women who have had pregnancy complications such as spontaneous pregnancy loss. We used cross-sectional data from the UK Biobank Imaging Enhancement Study to determine whether pregnancy loss is associated with cardiac or vascular remodelling in later life, which might contribute to this increased risk. METHODS: Pregnancy history was reported by women participating in UK Biobank between 2006 and 2010 at age 40-69 years using a self-completed touch-screen questionnaire. Associations between self-reported spontaneous pregnancy loss and cardiovascular measures, collected in women who participated in the Imaging Enhancement Study up to the end of 2015, were examined. Cardiac structure and function were assessed by magnetic resonance (CMR) steady-state free precession imaging at 1.5 Tesla. Carotid intima-media thickness (CIMT) measurements were taken for both common carotid arteries using a CardioHealth Station. Statistical associations with CMR and carotid measures were adjusted for age, BMI and other cardiovascular risk factors. RESULTS: Data were available on 2660 women of whom 111 were excluded because of pre-existing cardiovascular disease and 30 had no pregnancy information available. Of the remaining 2519, 446 were nulligravid and 2073 had a history of pregnancies, of whom 622 reported at least one pregnancy loss (92% miscarriages and 8% stillbirths) and 1451 reported no pregnancy loss. No significant differences in any cardiac or carotid parameters were evident in women who reported pregnancy loss compared to other groups (Table 1). CONCLUSION: Women who self-report pregnancy loss do not have significant differences in cardiac structure, cardiac function, or carotid structure in later life to explain their increased cardiovascular risk. This suggests any cardiovascular risks associated with pregnancy loss operate through other disease mechanisms. Alternatively, other characteristics of pregnancy loss, which we were not able to take account of, such as timing and number of pregnancy losses may be required to identify those at greatest cardiovascular risk.

Changes in Cardiac Morphology and Function in Individuals With Diabetes Mellitus: The UK Biobank Cardiovascular Magnetic Resonance Substudy.

BACKGROUND: Diabetes mellitus (DM) is associated with increased risk of cardiovascular disease. Detection of early cardiac changes before manifest disease develops is important. We investigated early alterations in cardiac structure and function associated with DM using cardiovascular magnetic resonance imaging. METHODS: Participants from the UK Biobank Cardiovascular Magnetic Resonance Substudy, a community cohort study, without known cardiovascular disease and left ventricular ejection fraction ≥50% were included. Multivariable linear regression models were performed. The investigators were blinded to DM status. RESULTS: A total of 3984 individuals, 45% men, (mean [SD]) age 61.3 (7.5) years, hereof 143 individuals (3.6%) with DM. There was no difference in left ventricular (LV) ejection fraction (DM versus no DM; coefficient [95% CI]: -0.86% [-1.8 to 0.5]; P=0.065), LV mass (-0.13 g/m2 [-1.6 to 1.3], P=0.86), or right ventricular ejection fraction (-0.23% [-1.2 to 0.8], P=0.65). However, both LV and right ventricular volumes were significantly smaller in DM, (LV end-diastolic volume/m2: -3.46 mL/m2 [-5.8 to -1.2], P=0.003, right ventricular end-diastolic volume/m2: -4.2 mL/m2 [-6.8 to -1.7], P=0.001, LV stroke volume/m2: -3.0 mL/m2 [-4.5 to -1.5], P<0.001; right ventricular stroke volume/m2: -3.8 mL/m2 [-6.5 to -1.1], P=0.005), LV mass/volume: 0.026 (0.01 to 0.04) g/mL, P=0.006. Both left atrial and right atrial emptying fraction were lower in DM (right atrial emptying fraction: -6.2% [-10.2 to -2.1], P=0.003; left atrial emptying fraction:-3.5% [-6.9 to -0.1], P=0.043). LV global circumferential strain was impaired in DM (coefficient [95% CI]: 0.38% [0.01 to 0.7], P=0.045). CONCLUSIONS: In a low-risk general population without known cardiovascular disease and with preserved LV ejection fraction, DM is associated with early changes in all 4 cardiac chambers. These findings suggest that diabetic cardiomyopathy is not a regional condition of the LV but affects the heart globally.

Quantitative CMR population imaging on 20,000 subjects of the UK Biobank imaging study: LV/RV quantification pipeline and its evaluation.

Population imaging studies generate data for developing and implementing personalised health strategies to prevent, or more effectively treat disease. Large prospective epidemiological studies acquire imaging for pre-symptomatic populations. These studies enable the early discovery of alterations due to impending disease, and enable early identification of individuals at risk. Such studies pose new challenges requiring automatic image analysis. To date, few large-scale population-level cardiac imaging studies have been conducted. One such study stands out for its sheer size, careful implementation, and availability of top quality expert annotation; the UK Biobank (UKB). The resulting massive imaging datasets (targeting ca. 100,000 subjects) has put published approaches for cardiac image quantification to the test. In this paper, we present and evaluate a cardiac magnetic resonance (CMR) image analysis pipeline that properly scales up and can provide a fully automatic analysis of the UKB CMR study. Without manual user interactions, our pipeline performs end-to-end image analytics from multi-view cine CMR images all the way to anatomical and functional bi-ventricular quantification. All this, while maintaining relevant quality controls of the CMR input images, and resulting image segmentations. To the best of our knowledge, this is the first published attempt to fully automate the extraction of global and regional reference ranges of all key functional cardiovascular indexes, from both left and right cardiac ventricles, for a population of 20,000 subjects imaged at 50 time frames per subject, for a total of one million CMR volumes. In addition, our pipeline provides 3D anatomical bi-ventricular models of the heart. These models enable the extraction of detailed information of the morphodynamics of the two ventricles for subsequent association to genetic, omics, lifestyle habits, exposure information, and other information provided in population imaging studies. We validated our proposed CMR analytics pipeline against manual expert readings on a reference cohort of 4620 subjects with contour delineations and corresponding clinical indexes. Our results show broad significant agreement between the manually obtained reference indexes, and those automatically computed via our framework. 80.67% of subjects were processed with mean contour distance of less than 1 pixel, and 17.50% with mean contour distance between 1 and 2 pixels. Finally, we compare our pipeline with a recently published approach reporting on UKB data, and based on deep learning. Our comparison shows similar performance in terms of segmentation accuracy with respect to human experts.

Automated quality control in image segmentation: application to the UK Biobank cardiovascular magnetic resonance imaging study.

BACKGROUND: The trend towards large-scale studies including population imaging poses new challenges in terms of quality control (QC). This is a particular issue when automatic processing tools such as image segmentation methods are employed to derive quantitative measures or biomarkers for further analyses. Manual inspection and visual QC of each segmentation result is not feasible at large scale. However, it is important to be able to automatically detect when a segmentation method fails in order to avoid inclusion of wrong measurements into subsequent analyses which could otherwise lead to incorrect conclusions. METHODS: To overcome this challenge, we explore an approach for predicting segmentation quality based on Reverse Classification Accuracy, which enables us to discriminate between successful and failed segmentations on a per-cases basis. We validate this approach on a new, large-scale manually-annotated set of 4800 cardiovascular magnetic resonance (CMR) scans. We then apply our method to a large cohort of 7250 CMR on which we have performed manual QC. RESULTS: We report results used for predicting segmentation quality metrics including Dice Similarity Coefficient (DSC) and surface-distance measures. As initial validation, we present data for 400 scans demonstrating 99% accuracy for classifying low and high quality segmentations using the predicted DSC scores. As further validation we show high correlation between real and predicted scores and 95% classification accuracy on 4800 scans for which manual segmentations were available. We mimic real-world application of the method on 7250 CMR where we show good agreement between predicted quality metrics and manual visual QC scores. CONCLUSIONS: We show that Reverse classification accuracy has the potential for accurate and fully automatic segmentation QC on a per-case basis in the context of large-scale population imaging as in the UK Biobank Imaging Study.