Cardiovascular magnetic resonance imaging in the UK Biobank: a major international health research resource

Device-measured physical activity and cardiac structure by magnetic resonance

BACKGROUND AND AIMS

Although extreme cardiac adaptions mirroring phenotypes of cardiomyopathy have been observed in endurance athletes, adaptions to high levels of physical activity within the wider population are under-explored. Therefore, in this study, associations between device-measured physical activity and clinically relevant cardiac magnetic resonance volumetric indices were investigated.

METHODS

Individuals without known cardiovascular disease or hypertension were included from the UK Biobank. Cardiac magnetic resonance data were collected between 2015 and 2019, and measures of end-diastolic chamber volume, left ventricular (LV) wall thickness, and LV ejection fraction were extracted. Moderate-to-vigorous-intensity physical activity (MVPA), vigorous-intensity physical activity (VPA), and total physical activity were assessed via wrist-worn accelerometers.

RESULTS

A total of 5977 women (median age and MVPA: 62 years and 46.8 min/day, respectively) and 4134 men (64 years and 49.8 min/day, respectively) were included. Each additional 10 min/day of MVPA was associated with a 0.70 [95% confidence interval (CI): 0.62, 0.79] mL/m2 higher indexed LV end-diastolic volume (LVEDVi) in women and a 1.08 (95% CI: 0.95, 1.20) mL/m2 higher LVEDVi in men. However, even within the top decile of MVPA, LVEDVi values remained within the normal ranges [79.1 (95% CI: 78.3, 80.0) mL/m2 in women and 91.4 (95% CI: 90.1, 92.7) mL/m2 in men]. Associations with MVPA were also observed for the right ventricle and the left/right atria, with an inverse association observed for LV ejection fraction. Associations of MVPA with maximum or average LV wall thickness were not clinically meaningful. Results for total physical activity and VPA mirrored those for MVPA.

CONCLUSIONS

High levels of device-measured physical activity were associated with cardiac remodelling within normal ranges.

Unlocking the diagnostic potential of electrocardiograms through information transfer from cardiac magnetic resonance imaging

Cardiovascular diseases (CVD) can be diagnosed using various diagnostic modalities. The electrocardiogram (ECG) is a cost-effective and widely available diagnostic aid that provides functional information of the heart. However, its ability to classify and spatially localise CVD is limited. In contrast, cardiac magnetic resonance (CMR) imaging provides detailed structural information of the heart and thus enables evidence-based diagnosis of CVD, but long scan times and high costs limit its use in clinical routine. In this work, we present a deep learning strategy for cost-effective and comprehensive cardiac screening solely from ECG. Our approach combines multimodal contrastive learning with masked data modelling to transfer domain-specific information from CMR imaging to ECG representations. In extensive experiments using data from 40,044 UK Biobank subjects, we demonstrate the utility and generalisability of our method for subject-specific risk prediction of CVD and the prediction of cardiac phenotypes using only ECG data. Specifically, our novel multimodal pre-training paradigm improves performance by up to 12.19% for risk prediction and 27.59% for phenotype prediction. In a qualitative analysis, we demonstrate that our learned ECG representations incorporate information from CMR image regions of interest. Our entire pipeline is publicly available at https://github.com/oetu/MMCL-ECG-CMR.

Use of AI in Cardiac CT and MRI: A Scientific Statement from the ESCR, EuSoMII, NASCI, SCCT, SCMR, SIIM, and RSNA

Artificial intelligence (AI) offers promising solutions for many steps of the cardiac imaging workflow, from patient and test selection through image acquisition, reconstruction, and interpretation, extending to prognostication and reporting. Despite the development of many cardiac imaging AI algorithms, AI tools are at various stages of development and face challenges for clinical implementation. This scientific statement, endorsed by several societies in the field, provides an overview of the current landscape and challenges of AI applications in cardiac CT and MRI. Each section is organized into questions and statements that address key steps of the cardiac imaging workflow, including ethical, legal, and environmental sustainability considerations. A technology readiness level range of 1 to 9 summarizes the maturity level of AI tools and reflects the progression from preliminary research to clinical implementation. This document aims to bridge the gap between burgeoning research developments and limited clinical applications of AI tools in cardiac CT and MRI.

Post-hoc out-of-distribution detection for cardiac MRI segmentation

In real-world scenarios, medical image segmentation models encounter input images that may deviate from the training images in various ways. These differences can arise from changes in image scanners and acquisition protocols, or even the images can come from a different modality or domain. When the model encounters these out-of-distribution (OOD) images, it can behave unpredictably. Therefore, it is important to develop a system that handles such out-of-distribution images to ensure the safe usage of the models in clinical practice. In this paper, we propose a post-hoc out-of-distribution (OOD) detection method that can be used with any pre-trained segmentation model. Our method utilizes multi-scale representations extracted from the encoder blocks of the segmentation model and employs Mahalanobis distance as a metric to measure the similarity between the input image and the in-distribution images. The segmentation model is pre-trained on a publicly available cardiac short-axis cine MRI dataset. The detection performance of the proposed method is evaluated on 13 different OOD datasets, which can be categorized as near, mild, and far OOD datasets based on their similarity to the in-distribution dataset. The results show that our method outperforms state-of-the-art feature space-based and uncertainty-based OOD detection methods across the various OOD datasets. Our method successfully detects near, mild, and far OOD images with high detection accuracy, showcasing the advantage of using the multi-scale and semantically rich representations of the encoder. In addition to the feature-based approach, we also propose a Dice coefficient-based OOD detection method, which demonstrates superior performance for adversarial OOD detection and shows a high correlation with segmentation quality. For the uncertainty-based method, despite having a strong correlation with the quality of the segmentation results in the near OOD datasets, they failed to detect mild and far OOD images, indicating the weakness of these methods when the images are more dissimilar. Future work will explore combining Mahalanobis distance and uncertainty scores for improved detection of challenging OOD images that are difficult to segment.

Virtual MOLLI Target: Generative Adversarial Networks Toward Improved Motion Correction in MRI Myocardial T1 Mapping

BACKGROUND

The modified Look-Locker inversion recovery (MOLLI) sequence is commonly used for myocardial T1 mapping. However, it acquires images with different inversion times, which causes difficulty in motion correction for respiratory-induced misregistration to a given target image.

HYPOTHESIS

Using a generative adversarial network (GAN) to produce virtual MOLLI images with consistent heart positions can reduce respiratory-induced misregistration of MOLLI datasets.

STUDY TYPE

Retrospective.

POPULATION

1071 MOLLI datasets from 392 human participants.

FIELD STRENGTH/SEQUENCE

Modified Look-Locker inversion recovery sequence at 3 T.

ASSESSMENT

A GAN model with a single inversion time image as input was trained to generate virtual MOLLI target (VMT) images at different inversion times which were subsequently used in an image registration algorithm. Four VMT models were investigated and the best performing model compared with the standard vendor-provided motion correction (MOCO) technique.

STATISTICAL TESTS

The effectiveness of the motion correction technique was assessed using the fitting quality index (FQI), mutual information (MI), and Dice coefficients of motion-corrected images, plus subjective quality evaluation of T1 maps by three independent readers using Likert score. Wilcoxon signed-rank test with Bonferroni correction for multiple comparison. Significance levels were defined as P < 0.01 for highly significant differences and P < 0.05 for significant differences.

RESULTS

The best performing VMT model with iterative registration demonstrated significantly better performance (FQI 0.88 ± 0.03, MI 1.78 ± 0.20, Dice 0.84 ± 0.23, quality score 2.26 ± 0.95) compared to other approaches, including the vendor-provided MOCO method (FQI 0.86 ± 0.04, MI 1.69 ± 0.25, Dice 0.80 ± 0.27, quality score 2.16 ± 1.01).

DATA CONCLUSION

Our GAN model generating VMT images improved motion correction, which may assist reliable T1 mapping in the presence of respiratory motion. Its robust performance, even with considerable respiratory-induced heart displacements, may be beneficial for patients with difficulties in breath-holding.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY: Stage 1.

Deep learning for cardiac imaging: focus on myocardial diseases, a narrative review

The integration of computational technologies into cardiology has significantly advanced the diagnosis and management of cardiovascular diseases. Computational cardiology, particularly, through cardiovascular imaging and informatics, enables a precise diagnosis of myocardial diseases utilizing techniques such as echocardiography, cardiac magnetic resonance imaging, and computed tomography. Early-stage disease classification, especially in asymptomatic patients, benefits from these advancements, potentially altering disease progression and improving patient outcomes. Automatic segmentation of myocardial tissue using deep learning (DL) algorithms improves efficiency and consistency in analyzing large patient populations. Radiomic analysis can reveal subtle disease characteristics from medical images and can enhance disease detection, enable patient stratification, and facilitate monitoring of disease progression and treatment response. Radiomic biomarkers have already demonstrated high diagnostic accuracy in distinguishing myocardial pathologies and promise treatment individualization in cardiology, earlier disease detection, and disease monitoring. In this context, this narrative review explores the current state of the art in DL applications in medical imaging (CT, CMR, echocardiography, and SPECT), focusing on automatic segmentation, radiomic feature phenotyping, and prediction of myocardial diseases, while also discussing challenges in integration of DL models in clinical practice.

The diabetic myocardial transcriptome reveals Erbb3 and Hspa2 as a novel biomarkers of incident heart failure

AIMS

Diabetes mellitus (DM) increases heart failure incidence and worsens prognosis, but its molecular basis is poorly defined in humans. We aimed to define the diabetic myocardial transcriptome and validate hits in their circulating protein form to define disease mechanisms and biomarkers.

METHODS AND RESULTS

RNA-sequencing data from the Genotype-Tissue Expression (GTEx) project was used to define differentially expressed genes (DEGs) in right atrial (RA) and left ventricular (LV) myocardium from people with vs. without DM (type 1 or 2). DEGs were validated as plasma proteins in the UK Biobank cohort, searching for directionally concordant differential expression. Validated plasma proteins were characterized in UK Biobank participants, irrespective of diabetes status, using cardiac magnetic resonance imaging, incident heart failure, and cardiovascular mortality. We found 32 and 32 DEGs associated with DM in the RA and LV, respectively, with no overlap between these. Plasma proteomic data were available for 12, with ERBB3, NRXN3, and HSPA2 (all LV hits) exhibiting directional concordance. Irrespective of DM status, lower circulating ERBB3 and higher HSPA2 were associated with impaired LV contractility and higher LV mass. Participants in the lowest quartile of circulating ERBB3 or highest quartile of circulating HSPA2 had increased incident heart failure and cardiovascular death vs. all other quartiles.

CONCLUSION

DM is characterized by lower Erbb3 and higher Hspa2 expression in the myocardium, with directionally concordant differences in their plasma protein concentration. These are associated with LV dysfunction, incident heart failure, and cardiovascular mortality.

Risk factors for raised left ventricular filling pressure by cardiovascular magnetic resonance: Prognostic insights

BACKGROUND

Cardiovascular magnetic resonance (CMR) imaging shows promise in estimating pulmonary capillary wedge pressure (PCWP) non-invasively. At the population level, the prognostic role of CMR-modelled PCWP remains unknown. Furthermore, the relationship between CMR-modelled PCWP and established risk factors for cardiovascular disease has not been well characterized.

OBJECTIVE

The main aim of this study was to investigate the prognostic value of CMR-modelled PCWP at the population level.

METHODS

Employing data from the imaging substudy of the UK Biobank, a very large prospective population-based cohort study, CMR-modelled PCWP was calculated using a model incorporating left atrial volume, left ventricular mass and sex. Logistic regression explored the relationships between typical cardiovascular risk factors and raised CMR-modelled PCWP (≥15 mmHg). Cox regression was used to examine the impact of typical risk factors and CMR-modelled PCWP on heart failure (HF) and major adverse cardiovascular events (MACE).

RESULTS

Data from 39 163 participants were included in the study. Median age of all participants was 64 years (inter-quartile range: 58 to 70), and 47% were males. Clinical characteristics independently associated with raised CMR-modelled PCWP included hypertension [odds ratio (OR) 1.57, 95% confidence interval (CI) 1.44-1.70, P < 0.001], body mass index (BMI) [OR 1.57, 95% CI 1.52-1.62, per standard deviation (SD) increment, P < 0.001], male sex (OR 1.37, 95% CI 1.26-1.47, P < 0.001), age (OR 1.33, 95% CI 1.27-1.41, per decade increment, P < 0.001) and regular alcohol consumption (OR 1.10, 95% CI 1.02-1.19, P = 0.012). After adjusting for potential confounders, CMR-modelled PCWP was independently associated with incident HF [hazard ratio (HR) 2.91, 95% CI 2.07-4.07, P < 0.001] and MACE (HR 1.48, 95% CI 1.16-1.89, P = 0.002).

CONCLUSIONS

Raised CMR-modelled PCWP is an independent risk factor for incident HF and MACE. CMR-modelled PCWP should be incorporated into routine CMR reports to guide HF diagnosis and further management.

Rare Genetic Variants in LDLR, APOB, and PCSK9 Are Associated With Aortic Stenosis

BACKGROUND

Despite a proposed causal role for LDL-C (low-density lipoprotein cholesterol) in aortic stenosis (AS), randomized controlled trials of lipid-lowering therapy failed to prevent severe AS. We aimed to assess the impact on AS and peak velocity across the aortic valve conferred by lifelong alterations in LDL-C levels mediated by protein-disrupting variants in 3 clinically significant genes for LDL (low-density lipoprotein) metabolism (LDLR, APOB, and PCSK9).

METHODS

We used sequencing data and electronic health records from UK Biobank (UKB) and All of Us and magnetic resonance imaging data from UKB. We identified predicted protein-disrupting variants with the Loss Of Function Transcript Effect Estimator (LOFTEE) and AlphaMissense algorithms and evaluated their associations with LDL-C and peak velocity across the aortic valve (UK Biobank), as well as diagnosed AS and aortic valve replacement (UK Biobank and All of Us).

RESULTS

We included 421 049 unrelated participants (5621 with AS) in UKB and 195 519 unrelated participants (1087 with AS) in All of Us. Carriers of protein-disrupting variants in LDLR had higher mean LDL-C (UKB: +42.6 mg/dL; P=4.4e-237) and greater risk of AS (meta-analysis: odds ratio, 3.52 [95% CI, 2.39-5.20]; P=2.3e-10) and aortic valve replacement (meta-analysis: odds ratio, 3.78 [95% CI, 2.26-6.32]; P=4.0e-7). Carriers of protein-disrupting variants in APOB or PCSK9 had lower mean LDL-C (UKB: -32.3 mg/dL; P<5e-324) and lower risk of AS (meta-analysis: odds ratio, 0.49 [95% CI, 0.31-0.75]; P=0.001) and aortic valve replacement (meta-analysis: odds ratio, 0.54 [95% CI, 0.30-0.97]; P=0.04). Among 57 371 UKB imaging substudy participants, peak velocities across the aortic valve were greater in carriers of protein-disrupting variants in LDLR (+12.2 cm/s; P=1.6e-5) and lower in carriers of protein-disrupting variants in PCSK9 (-6.9 cm/s; P=0.022).

CONCLUSIONS

Rare genetic variants that confer lifelong higher or lower LDL-C levels are associated with substantially increased and decreased risk of AS, respectively. Early and sustained lipid-lowering therapy may slow or prevent AS development.

Association of left ventricular ejection fraction with risk of cardiovascular diseases: a prospective cohort study

Left ventricular ejection fraction (LVEF) is the most ubiquitous parameter in cardiac imaging examinations, we aimed to investigate the associations between subtle changes of LVEF and risk of common cardiovascular diseases. This is a prospective cohort study based on UK Biobank. LVEF was obtained from cardiac magnetic resonance. Incident cardiovascular disease was the outcome, including heart failure, atrial fibrillation, ischemic heart disease, and myocardial infarction. Cox proportional hazard model was the main method. A U-shaped relationship was observed between quantified LVEF and cardiovascular events risk with the nadir at the LVEF of 55-64%. As compared to moderate LVEF (55-64%), both low (40-54%) and high LVEF ( ≥ 5%) were related to higher risk of cardiovascular diseases after adjusting for confounders (HRlow = 1.15, 95%CI = 1.02-1.30; HRhigh = 1.34, 95%CI = 1.05-1.72). Specifically, low LVEF was associated with increased risk of heart failure while high LVEF predominantly predicted elevated risk of ischemic heart diseases, both low and high LVEF were related to a borderline higher incidence of atrial fibrillation. Besides, associations with specific cardiovascular diseases varied by age, sex or comorbidities. There was a U-shaped relationship between LVEF and cardiovascular events risk with the nadir at the LVEF of 55-64%, while these associations were disease-specific and varied by age, sex or comorbidities.

Association of biological age acceleration with cardiac morphology, function, and incident heart failure: insights from UK Biobank participants

AIMS

Advanced age is associated with an increased risk of adverse cardiovascular events. The relationship between biological age acceleration (BAA), cardiac size, cardiac function, and heart failure (HF) is not well-defined.

METHODS AND RESULTS

Utilizing the UK Biobank cohort, we assessed biological age using the Klemera-Doubal and PhenoAge methods. BAA was quantified by residual analysis compared with chronological age. Cardiovascular magnetic resonance (CMR) imaging provided detailed insights into cardiac structure and function. We employed multivariate regression to examine links between BAA and CMR-derived cardiac phenotypes. Cox proportional hazard regression models analysis was applied to explore the causative relationship between BAA and HF. Additionally, Mendelian randomization was used to investigate the genetic underpinnings of these associations. A significant correlation was found between increased BAA and deleterious changes in cardiac structure, such as diminished left ventricular mass, lower overall ventricular volume, and reduced stroke volumes across ventricles and atria. Throughout a median follow-up of 13.8 years, participants with greater biological aging showed a heightened risk of HF [26% per standard deviation (SD) increase in KDM-BA acceleration, 95% confidence intervals (CI): 23-28%; 33% per SD increase in PhenoAge acceleration, 95% CI: 32-35%]. Mendelian randomization analysis suggests a likely causal link between BAA, vital cardiac metrics, and HF risk.

CONCLUSION

In this cohort, accelerated biological aging may serve as a risk indicator for altered cardiac dimensions, functionality, and the onset of heart failure among middle-aged and elderly adults. It holds promise as a focal point for evaluating risk and developing targeted interventions.

Myocardial Strain Measured by Cardiac Magnetic Resonance Predicts Cardiovascular Morbidity and Death

BACKGROUND

Myocardial strain using cardiac magnetic resonance (CMR) is a sensitive marker for predicting adverse outcomes in many cardiac disease states, but the prognostic value in the general population has not been studied conclusively.

OBJECTIVES

The goal of this study was to assess the independent prognostic value of CMR feature tracking (FT)-derived LV global longitudinal (GLS), circumferential (GCS), and radial strain (GRS) metrics in predicting adverse outcomes (heart failure, myocardial infarction, stroke, and death).

METHODS

Participants from the UK Biobank population imaging study were included. Univariable and multivariable Cox models were used for each outcome and each strain marker (GLS, GCS, GRS) separately. The multivariable models were tested with adjustment for prognostically important clinical features and conventional global LV imaging markers relevant for each outcome.

RESULTS

Overall, 45,700 participants were included in the study (average age 65 ± 8 years), with a median follow-up period of 3 years. All univariable and multivariable models demonstrated that lower absolute GLS, GCS, and GRS were associated with increased incidence of heart failure, myocardial infarction, stroke, and death. All strain markers were independent predictors (incrementally above some respective conventional LV imaging markers) for the morbidity outcomes, but only GLS predicted death independently: (HR: 1.18; 95% CI: 1.07-1.30).

CONCLUSIONS

In the general population, LV strain metrics derived using CMR-FT in radial, circumferential, and longitudinal directions are strongly and independently predictive of heart failure, myocardial infarction, and stroke, but only GLS is independently predictive of death in an adult population cohort.

Daytime Napping, Incident Atrial Fibrillation, and Dynamic Transitions With Dementia

Background

Associations between napping and incident atrial fibrillation (AF) remain unknown, and few studies have accounted for dynamic transitions between AF and dementia.

Objectives

The purpose of this study was to evaluate associations between napping with incident AF and the dynamic transitions of AF and dementia, as well as the mediation pathway of left ventricular (LV) size and function.

Methods

A total of 476,588 participants from UK Biobank were included. Napping frequency and other sleep behaviors were evaluated. Incident AF, dementia, and mortality were ascertained via linkage to external registry databases. LV size and function indices were obtained from cardiovascular magnetic resonance imaging phenotypes. A multistate survival analysis was conducted to examine daytime napping in relation to dynamic transitions. Weighed AF genetic risk score was calculated.

Results

Frequent daytime napping, compared to never/rarely napping, was associated with a 1.17-fold AF risk (HR: 1.17; 95% CI: 1.12-1.22), which persisted after controlling for other sleep behaviors. Genetic predisposition significantly modified associations between napping and AF (P for interaction <0.001), with stronger associations observed in those of low and moderate genetic risk. LV ejection fraction significantly mediated 26.2% (95% CI: 4.2%-74.1%) of associations between napping and AF. Frequent napping was also associated with a 1.27-fold risk of transition from AF to comorbidity of AF and dementia.

Conclusions

Our findings highlight the potential importance of screening for napping in view of the association with incident AF and dementia. Routine evaluations of the LV ejection fraction could be warranted to timely identify early indications of AF onset among habitual nappers.

Ventricular volume asymmetry as a novel imaging biomarker for disease discrimination and outcome prediction

Aims

Disruption of the predictable symmetry of the healthy heart may be an indicator of cardiovascular risk. This study defines the population distribution of ventricular asymmetry and its relationships across a range of prevalent and incident cardiorespiratory diseases.

Methods and results

The analysis includes 44 796 UK Biobank participants (average age 64.1 ± 7.7 years; 51.9% women). Cardiovascular magnetic resonance (CMR) metrics were derived using previously validated automated pipelines. Ventricular asymmetry was expressed as the ratio of left and right ventricular (LV and RV) end-diastolic volumes. Clinical outcomes were defined through linked health records. Incident events were those occurring for the first time after imaging, longitudinally tracked over an average follow-up time of 4.75 ± 1.52 years. The normal range for ventricular symmetry was defined in a healthy subset. Participants with values outside the 5th-95th percentiles of the healthy distribution were classed as either LV dominant (LV/RV > 112%) or RV dominant (LV/RV < 80%) asymmetry. Associations of LV and RV dominant asymmetry with vascular risk factors, CMR features, and prevalent and incident cardiovascular diseases (CVDs) were examined using regression models, adjusting for vascular risk factors, prevalent diseases, and conventional CMR measures. Left ventricular dominance was linked to an array of pre-existing vascular risk factors and CVDs, and a two-fold increased risk of incident heart failure, non-ischaemic cardiomyopathies, and left-sided valvular disorders. Right ventricular dominance was associated with an elevated risk of all-cause mortality.

Conclusion

Ventricular asymmetry has clinical utility for cardiovascular risk assessment, providing information that is incremental to traditional risk factors and conventional CMR metrics.

Quality control of cardiac magnetic resonance imaging segmentation, feature tracking, aortic flow, and native T1 analysis using automated batch processing in the UK Biobank study

Aims

Automated algorithms are regularly used to analyse cardiac magnetic resonance (CMR) images. Validating data output reliability from this method is crucial for enabling widespread adoption. We outline a visual quality control (VQC) process for image analysis using automated batch processing. We assess the performance of automated analysis and the reliability of replacing visual checks with statistical outlier (SO) removal approach in UK Biobank CMR scans.

Methods and results

We included 1987 CMR scans from the UK Biobank COVID-19 imaging study. We used batch processing software (Circle Cardiovascular Imaging Inc.-CVI42) to automatically extract chamber volumetric data, strain, native T1, and aortic flow data. The automated analysis outputs (∼62 000 videos and 2000 images) were visually checked by six experienced clinicians using a standardized approach and a custom-built R Shiny app. Inter-observer variability was assessed. Data from scans passing VQC were compared with a SO removal QC method in a subset of healthy individuals (n = 1069). Automated segmentation was highly rated, with over 95% of scans passing VQC. Overall inter-observer agreement was very good (Gwet's AC2 0.91; 95% confidence interval 0.84, 0.94). No difference in overall data derived from VQC or SO removal in healthy individuals was observed.

Conclusion

Automated image analysis using CVI42 prototypes for UK Biobank CMR scans demonstrated high quality. Larger UK Biobank data sets analysed using these automated algorithms do not require in-depth VQC. SO removal is sufficient as a QC measure, with operator discretion for visual checks based on population or research objectives.

The associations of leukocyte telomere length and intermediary cardiovascular phenotype with adverse cardiovascular outcomes in the white population

The evidence about the associations of leukocyte telomere length (LTL) and intermediary cardiovascular phenotypes with adverse cardiovascular outcomes is inconclusive. This study assessed these relationships with cardiovascular imaging, electrocardiography, and the risks of sudden cardiac death (SCD), coronary events, and heart failure (HF) admission. We conducted a cross-sectional analysis of UK Biobank participants enrolled between 2006 and 2010. LTL was measured using quantitative polymerase chain reactions. Electronic health records were used to determine the incidence of SCD, coronary events, and HF admission. Cardiovascular measurements were made using cardiovascular magnetic resonance imaging and machine learning. The associations of LTL with SCD, coronary events, and HF admission and cardiac magnetic resonance imaging, electrocardiogram parameters of 33,043 and 19,554 participants were evaluated by multivariate regression. The median (interquartile range) follow-up period was 11.9 (11.2-12.6) years. Data was analyzed from January to May 2023. Among the 403,382 white participants without coronary artery disease or HF, 181,637 (45.0%) were male with a mean age of 57.1 years old. LTL was independently negatively associated with a risk of SCD (LTL third quartile vs first quartile: hazard ratio [HR]: 0.81, 95% confidence interval [CI]: 0.72-0.92), coronary events (LTL third quartile vs first quartile: HR: 0.88, 95% CI: 0.84-0.92), and HF admission (LTL fourth quartile vs first quartile: HR: 0.84, 95% CI: 0.74-0.95). LTL was also independently positively associated with cardiac remodeling, specifically left ventricular mass index, left-ventricular-end systolic and diastolic volumes, mean left ventricular myocardial wall thickness, left ventricular stroke volume, and with electrocardiogram changes along the negative degree of T-axis. Cross-sectional study results showed that LTL was positively associated with heart size and cardiac function in middle age, but electrocardiography results did not show these associations, which could explain the negative association between LTL and risk of SCD, coronary events, and HF admission in UK Biobank participants.

Bone health, cardiovascular disease, and imaging outcomes in UK Biobank: a causal analysis

This study examined the association of estimated heel bone mineral density (eBMD, derived from quantitative ultrasound) with: (1) prevalent and incident cardiovascular diseases (CVDs: ischemic heart disease (IHD), myocardial infarction (MI), heart failure (HF), non-ischemic cardiomyopathy (NICM), arrhythmia), (2) mortality (all-cause, CVD, IHD), and (3) cardiovascular magnetic resonance (CMR) measures of left ventricular and atrial structure and function and aortic distensibility, in the UK Biobank. Clinical outcomes were ascertained using health record linkage over 12.3 yr of prospective follow-up. Two-sample Mendelian randomization (MR) was conducted to assess causal associations between BMD and CMR metrics using genetic instrumental variables identified from published genome-wide association studies. The analysis included 485 257 participants (55% women, mean age 56.5 ± 8.1 yr). Higher heel eBMD was associated with lower odds of all prevalent CVDs considered. The greatest magnitude of effect was seen in association with HF and NICM, where 1-SD increase in eBMD was associated with 15% lower odds of HF and 16% lower odds of NICM. Association between eBMD and incident IHD and MI was non-significant; the strongest relationship was with incident HF (SHR: 0.90 [95% CI, 0.89-0.92]). Higher eBMD was associated with a decreased risk in all-cause, CVD, and IHD mortality, in the fully adjusted model. Higher eBMD was associated with greater aortic distensibility; associations with other CMR metrics were null. Higher heel eBMD is linked to reduced risk of a range of prevalent and incident CVD and mortality outcomes. Although observational analyses suggest associations between higher eBMD and greater aortic compliance, MR analysis did not support a causal relationship between genetically predicted BMD and CMR phenotypes. These findings support the notion that bone-cardiovascular associations reflect shared risk factors/mechanisms rather than direct causal pathways.

Cardiovascular Significance and Genetics of Epicardial and Pericardial Adiposity

Importance

Epicardial and pericardial adipose tissue (EPAT) has been associated with cardiovascular diseases such as atrial fibrillation or flutter (AF) and coronary artery disease (CAD), but studies have been limited in sample size or drawn from selected populations. It has been suggested that the association between EPAT and cardiovascular disease could be mediated by local or paracrine effects.

Objective

To evaluate the association of EPAT with prevalent and incident cardiovascular disease and to elucidate the genetic basis of EPAT in a large population cohort.

Design, Setting, and Participants

A deep learning model was trained to quantify EPAT area from 4-chamber magnetic resonance images using semantic segmentation. Cross-sectional and prospective cardiovascular disease associations were evaluated, controlling for sex and age. Prospective associations were additionally controlled for abdominal visceral adipose tissue (VAT) volumes. A genome-wide association study was performed, and a polygenic score (PGS) for EPAT was examined in independent FinnGen cohort study participants. Data analyses were conducted from March 2022 to December 2023.

Exposures

The primary exposures were magnetic resonance imaging-derived continuous measurements of epicardial and pericardial adipose tissue area and visceral adipose tissue volume.

Main Outcomes and Measures

Prevalent and incident CAD, AF, heart failure (HF), stroke, and type 2 diabetes (T2D).

Results

After exclusions, this study included 44 475 participants (mean [SD] age, 64.1 [7.7] years; 22 972 female [51.7%]) from the UK Biobank. Cross-sectional and prospective cardiovascular disease associations were evaluated for a mean (SD) of 3.2 (1.5) years of follow-up. Prospective associations were additionally controlled for abdominal VAT volumes for 38 527 participants. A PGS for EPAT was examined in 453 733 independent FinnGen cohort study participants. EPAT was positively associated with male sex (β = +0.78 SD in EPAT; P < 3 × 10-324), age (Pearson r = 0.15; P = 9.3 × 10-229), body mass index (Pearson r = 0.47; P < 3 × 10-324), and VAT (Pearson r = 0.72; P < 3 × 10-324). EPAT was more elevated in prevalent HF (β = +0.46 SD units) and T2D (β = +0.56) than in CAD (β = +0.23) or AF (β = +0.18). EPAT was associated with incident HF (hazard ratio [HR], 1.29 per +1 SD in EPAT; 95% CI, 1.17-1.43), T2D (HR, 1.63; 95% CI, 1.51-1.76), and CAD (HR, 1.19; 95% CI, 1.11-1.28). However, the associations were no longer significant when controlling for VAT. Seven genetic loci were identified for EPAT, implicating transcriptional regulators of adipocyte morphology and brown adipogenesis (EBF1, EBF2, and CEBPA) and regulators of visceral adiposity (WARS2 and TRIB2). The EPAT PGS was associated with T2D (odds ratio [OR], 1.06; 95% CI, 1.05-1.07; P =3.6 × 10-44), HF (OR, 1.05; 95% CI, 1.04-1.06; P =4.8 × 10-15), CAD (OR, 1.04; 95% CI, 1.03-1.05; P =1.4 × 10-17), AF (OR, 1.04; 95% CI, 1.03-1.06; P =7.6 × 10-12), and stroke in FinnGen (OR, 1.02; 95% CI, 1.01-1.03; P =3.5 × 10-3) per 1 SD in PGS.

Conclusions and Relevance

Results of this cohort study suggest that epicardial and pericardial adiposity was associated with incident cardiovascular diseases, but this may largely reflect a metabolically unhealthy adiposity phenotype similar to abdominal visceral adiposity.

Assessment of new-onset heart failure prediction in a diabetic population using left ventricular global strain: a prospective cohort study based on UK Biobank

Background

Impaired glucose utilization influences myocardial contractile function. However, the prognostic importance of left ventricular global radial strain (LV-GRS), left ventricular global circumferential strain (LV-GCS), and left ventricular global longitudinal strain (LV-GLS) in predicting new-onset heart failure (HF) in a population with diabetes is unclear.

Methods

The study design is prospective cohort from the UK Biobank. Totally 37,899 participants had a complete data of cardiac magnetic resonance (CMR), of which 940 patients with diabetes were included, and all the participants completed follow-up. LV-GRS, LV-GCS, and LV-GLS were measured by completely automated CMR with tissue tagging. Cox proportional hazards regression analysis and C-index was performed to evaluate the association between the strain parameters and the new-onset HF in patients suffering from diabetes.

Results

The average age of the 940 participants was 57.67 ± 6.97 years, with males comprising 66.4% of the overall population. With an average follow-up period of 166.82 ± 15.26 months, 35 (3.72%) patients reached the endpoint (emergence of new-onset HF). Significant associations were found for the three strain parameters and the new-onset HF (LV-GRS-hazard ratio [HR]: 0.946, 95% CI: 0.916-0.976; LV-GCS-HR: 1.162, 95% CI: 1.086-1.244; LV-GCS-HR: 1.181, 95% CI: 1.082-1.289). LV-GRS, LV-GCS, and LV-GLS were closely related to the related indicators to HF, and showed a high relationship to new-onset HF in individuals with diabetes at 5 and 10 years: LV-GRS: 0.75 (95% CI, 0.41-0.94) and 0.76 (95% CI, 0.44-0.98), respectively; LV-GCS: 0.80 (95% CI, 0.50-0.96) and 0.75 (95% CI, 0.41-0.98), respectively; LV-GLS: 0.72 (95% CI, 0.40-0.93) and 0.76 (95% CI, 0.48-0.97), respectively. In addition, age, sex, body mass index (BMI), and presence of hypertension or coronary artery disease (CAD) made no impacts on the association between the global strain parameters and the incidence of HF.

Conclusion

LV-GRS, LV-GCS, and LV-GLS is significantly related to new-onset HF in patients with diabetes at 5 and 10 years.

Developing Cardiac Digital Twins at Scale: Insights from Personalised Myocardial Conduction Velocity

Large-cohort studies using cardiovascular imaging and diagnostic datasets have assessed cardiac anatomy, function, and outcomes, but typically do not reveal underlying biological mechanisms. Cardiac digital twins (CDTs) provide personalized physics- and physiology-constrained in-silico representations, enabling inference of multi-scale properties tied to these mechanisms. We constructed 3464 anatomically-accurate CDTs using cardiac magnetic resonance images from UK biobank and personalised their myocardial conduction velocities (CVs) from electrocardiograms (ECG), through an automated framework. We found well-known sex-specific differences in QRS duration were fully explained by myocardial anatomy, as CV remained consistent across sexes. Conversely, significant associations of CV with ageing and increased BMI suggest myocardial tissue remodelling. Novel associations were observed with left ventricular ejection fraction and mental-health phenotypes, through a phenome-wide association study, and CV was also linked with adverse clinical outcomes. Our study highlights the utility of population-based CDTs in assessing intersubject variability and uncovering strong links with mental health.

Deep Learning for Cardiovascular Imaging: A Review

Importance

Artificial intelligence (AI), driven by advances in deep learning (DL), has the potential to reshape the field of cardiovascular imaging (CVI). While DL for CVI is still in its infancy, research is accelerating to aid in the acquisition, processing, and/or interpretation of CVI across various modalities, with several commercial products already in clinical use. It is imperative that cardiovascular imagers are familiar with DL systems, including a basic understanding of how they work, their relative strengths compared with other automated systems, and possible pitfalls in their implementation. The goal of this article is to review the methodology and application of DL to CVI in a simple, digestible fashion toward demystifying this emerging technology.

Observations

At its core, DL is simply the application of a series of tunable mathematical operations that translate input data into a desired output. Based on artificial neural networks that are inspired by the human nervous system, there are several types of DL architectures suited to different tasks; convolutional neural networks are particularly adept at extracting valuable information from CVI data. We survey some of the notable applications of DL to tasks across the spectrum of CVI modalities. We also discuss challenges in the development and implementation of DL systems, including avoiding overfitting, preventing systematic bias, improving explainability, and fostering a human-machine partnership. Finally, we conclude with a vision of the future of DL for CVI.

Conclusions and Relevance

Deep learning has the potential to meaningfully affect the field of CVI. Rather than a threat, DL could be seen as a partner to cardiovascular imagers in reducing technical burden and improving efficiency and quality of care. High-quality prospective evidence is still needed to demonstrate how the benefits of DL CVI systems may outweigh the risks.

Multi-PGS enhances polygenic prediction by combining 937 polygenic scores

The predictive performance of polygenic scores (PGS) is largely dependent on the number of samples available to train the PGS. Increasing the sample size for a specific phenotype is expensive and takes time, but this sample size can be effectively increased by using genetically correlated phenotypes. We propose a framework to generate multi-PGS from thousands of publicly available genome-wide association studies (GWAS) with no need to individually select the most relevant ones. In this study, the multi-PGS framework increases prediction accuracy over single PGS for all included psychiatric disorders and other available outcomes, with prediction R2 increases of up to 9-fold for attention-deficit/hyperactivity disorder compared to a single PGS. We also generate multi-PGS for phenotypes without an existing GWAS and for case-case predictions. We benchmark the multi-PGS framework against other methods and highlight its potential application to new emerging biobanks.

Relationship Among Diabetes, Obesity, and Cardiovascular Disease Phenotypes: A UK Biobank Cohort Study

OBJECTIVE

Obesity and diabetes frequently coexist, yet their individual contributions to cardiovascular risk remain debated. We explored cardiovascular disease biomarkers, events, and mortality in the UK Biobank stratified by BMI and diabetes.

RESEARCH DESIGN AND METHODS

A total of 451,355 participants were stratified by ethnicity-specific BMI categories (normal, overweight, obese) and diabetes status. We examined cardiovascular biomarkers including carotid intima-media thickness (CIMT), arterial stiffness, left ventricular ejection fraction (LVEF), and cardiac contractility index (CCI). Poisson regression models estimated adjusted incidence rate ratios (IRRs) for myocardial infarction, ischemic stroke, and cardiovascular death, with normal-weight nondiabetes as comparator.

RESULTS

Five percent of participants had diabetes (10% normal weight, 34% overweight, and 55% obese vs. 34%, 43%, and 23%, respectively, without diabetes). In the nondiabetes group, overweight/obesity was associated with higher CIMT, arterial stiffness, and CCI and lower LVEF (P < 0.005); these relationships were diminished in the diabetes group. Within BMI classes, diabetes was associated with adverse cardiovascular biomarker phenotype (P < 0.005), particularly in the normal-weight group. After 5,323,190 person-years follow-up, incident myocardial infarction, ischemic stroke, and cardiovascular mortality rose across increasing BMI categories without diabetes (P < 0.005); this was comparable in the diabetes groups (P-interaction > 0.05). Normal-weight diabetes had comparable adjusted cardiovascular mortality to obese nondiabetes (IRR 1.22 [95% CI 0.96-1.56]; P = 0.1).

CONCLUSIONS

Obesity and diabetes are additively associated with adverse cardiovascular biomarkers and mortality risk. While adiposity metrics are more strongly correlated with cardiovascular biomarkers than diabetes-oriented metrics, both correlate weakly, suggesting that other factors underpin the high cardiovascular risk of normal-weight diabetes.

The Cardiovascular Impact and Genetics of Pericardial Adiposity

Background

While previous studies have reported associations of pericardial adipose tissue (PAT) with cardiovascular diseases such as atrial fibrillation and coronary artery disease, they have been limited in sample size or drawn from selected populations. Additionally, the genetic determinants of PAT remain largely unknown. We aimed to evaluate the association of PAT with prevalent and incident cardiovascular disease and to elucidate the genetic basis of PAT in a large population cohort.

Methods

A deep learning model was trained to quantify PAT area from four-chamber magnetic resonance images in the UK Biobank using semantic segmentation. Cross-sectional and prospective cardiovascular disease associations were evaluated, controlling for sex and age. A genome-wide association study was performed, and a polygenic score (PGS) for PAT was examined in 453,733 independent FinnGen study participants.

Results

A total of 44,725 UK Biobank participants (51.7% female, mean [SD] age 64.1 [7.7] years) were included. PAT was positively associated with male sex (β = +0.76 SD in PAT), age (r = 0.15), body mass index (BMI; r = 0.47) and waist-to-hip ratio (r = 0.55) (P < 1×10-230). PAT was more elevated in prevalent heart failure (β = +0.46 SD units) and type 2 diabetes (β = +0.56) than in coronary artery disease (β = +0.22) or AF (β = +0.18). PAT was associated with incident heart failure (HR = 1.29 per +1 SD in PAT [95% CI 1.17-1.43]) and type 2 diabetes (HR = 1.63 [1.51-1.76]) during a mean 3.2 (±1.5) years of follow-up; the associations remained significant when controlling for BMI. We identified 5 novel genetic loci for PAT and implicated transcriptional regulators of adipocyte morphology and brown adipogenesis (EBF1, EBF2 and CEBPA) and regulators of visceral adiposity (WARS2 and TRIB2). The PAT PGS was associated with T2D, heart failure, coronary artery disease and atrial fibrillation in FinnGen (ORs 1.03-1.06 per +1 SD in PGS, P < 2×10-10).

Conclusions

PAT shares genetic determinants with abdominal adiposity and is an independent predictor of incident type 2 diabetes and heart failure.

Heart-brain connections: Phenotypic and genetic insights from magnetic resonance images

Cardiovascular health interacts with cognitive and mental health in complex ways, yet little is known about the phenotypic and genetic links of heart-brain systems. We quantified heart-brain connections using multiorgan magnetic resonance imaging (MRI) data from more than 40,000 subjects. Heart MRI traits displayed numerous association patterns with brain gray matter morphometry, white matter microstructure, and functional networks. We identified 80 associated genomic loci (P < 6.09 × 10-10) for heart MRI traits, which shared genetic influences with cardiovascular and brain diseases. Genetic correlations were observed between heart MRI traits and brain-related traits and disorders. Mendelian randomization suggests that heart conditions may causally contribute to brain disorders. Our results advance a multiorgan perspective on human health by revealing heart-brain connections and shared genetic influences.

Assessment of valvular function in over 47,000 people using deep learning-based flow measurements

Valvular heart disease is associated with a high global burden of disease. Even mild aortic stenosis confers increased morbidity and mortality, prompting interest in understanding normal variation in valvular function at scale. We developed a deep learning model to study velocity-encoded magnetic resonance imaging in 47,223 UK Biobank participants. We calculated eight traits, including peak velocity, mean gradient, aortic valve area, forward stroke volume, mitral and aortic regurgitant volume, greatest average velocity, and ascending aortic diameter. We then computed sex-stratified reference ranges for these phenotypes in up to 31,909 healthy individuals. In healthy individuals, we found an annual decrement of 0.03cm ² in the aortic valve area. Participants with mitral valve prolapse had a 1 standard deviation [SD] higher mitral regurgitant volume (P=9.6 × 10 ^-12 ), and those with aortic stenosis had a 4.5 SD-higher mean gradient (P=1.5 × 10 ^-431 ), validating the derived phenotypes' associations with clinical disease. Greater levels of ApoB, triglycerides, and Lp(a) assayed nearly 10 years prior to imaging were associated with higher gradients across the aortic valve. Metabolomic profiles revealed that increased glycoprotein acetyls were also associated with an increased aortic valve mean gradient (0.92 SD, P=2.1 x 10 ^-22 ). Finally, velocity-derived phenotypes were risk markers for aortic and mitral valve surgery even at thresholds below what is considered relevant disease currently. Using machine learning to quantify the rich phenotypic data of the UK Biobank, we report the largest assessment of valvular function and cardiovascular disease in the general population.

Incident Clinical and Mortality Associations of Myocardial Native T1 in the UK Biobank

BACKGROUND

Cardiac magnetic resonance native T1-mapping provides noninvasive, quantitative, and contrast-free myocardial characterization. However, its predictive value in population cohorts has not been studied.

OBJECTIVES

The associations of native T1 with incident events were evaluated in 42,308 UK Biobank participants over 3.17 ± 1.53 years of prospective follow-up.

METHODS

Native T1-mapping was performed in 1 midventricular short-axis slice using the Shortened Modified Look-Locker Inversion recovery technique (WIP780B) in 1.5-T scanners (Siemens Healthcare). Global myocardial T1 was calculated using an automated tool. Associations of T1 with: 1) prevalent risk factors (eg, diabetes, hypertension, and high cholesterol); 2) prevalent and incident diseases (eg, any cardiovascular disease [CVD], any brain disease, valvular heart disease, heart failure, nonischemic cardiomyopathies, cardiac arrhythmias, atrial fibrillation [AF], myocardial infarction, ischemic heart disease [IHD], and stroke); and 3) mortality (eg, all-cause, CVD, and IHD) were examined. Results are reported as odds ratios (ORs) or HRs per SD increment of T1 value with 95% CIs and corrected P values, from logistic and Cox proportional hazards regression models.

RESULTS

Higher myocardial T1 was associated with greater odds of a range of prevalent conditions (eg, any CVD, brain disease, heart failure, nonischemic cardiomyopathies, AF, stroke, and diabetes). The strongest relationships were with heart failure (OR: 1.41 [95% CI: 1.26-1.57]; P = 1.60 × 10-9) and nonischemic cardiomyopathies (OR: 1.40 [95% CI: 1.16-1.66]; P = 2.42 × 10-4). Native T1 was positively associated with incident AF (HR: 1.25 [95% CI: 1.10-1.43]; P = 9.19 × 10-4), incident heart failure (HR: 1.47 [95% CI: 1.31-1.65]; P = 4.79 × 10-11), all-cause mortality (HR: 1.24 [95% CI: 1.12-1.36]; P = 1.51 × 10-5), CVD mortality (HR: 1.40 [95% CI: 1.14-1.73]; P = 0.0014), and IHD mortality (HR: 1.36 [95% CI: 1.03-1.80]; P = 0.0310).

CONCLUSIONS

This large population study demonstrates the utility of myocardial native T1-mapping for disease discrimination and outcome prediction.

The year 2021 in the European Heart Journal: Cardiovascular Imaging Part II

The European Heart Journal-Cardiovascular Imaging was launched in 2012 and has during these years become one of the leading multimodality cardiovascular imaging journals. The journal is currently ranked as Number 19 among all cardiovascular journals. It has an impressive impact factor of 9.130. The most important studies published in our Journal from 2021 will be highlighted in two reports. Part II will focus on valvular heart disease, heart failure, cardiomyopathies, and congenital heart disease, while Part I of the review has focused on studies about myocardial function and risk prediction, myocardial ischaemia, and emerging techniques in cardiovascular imaging.

Left atrial structure and function are associated with cardiovascular outcomes independent of left ventricular measures: a UK Biobank CMR study

AIMS

We evaluated the associations of left atrial (LA) structure and function with prevalent and incident cardiovascular disease (CVD), independent of left ventricular (LV) metrics, in 25 896 UK Biobank participants.

METHODS AND RESULTS

We estimated the association of cardiovascular magnetic resonance (CMR) metrics [LA maximum volume (LAV), LA ejection fraction (LAEF), LV mass : LV end-diastolic volume ratio (LVM : LVEDV), global longitudinal strain, and LV global function index (LVGFI)] with vascular risk factors (hypertension, diabetes, high cholesterol, and smoking), prevalent and incident CVDs [atrial fibrillation (AF), stroke, ischaemic heart disease (IHD), myocardial infarction], all-cause mortality, and CVD mortality. We created uncorrelated CMR variables using orthogonal principal component analysis rotation. All five CMR metrics were simultaneously entered into multivariable regression models adjusted for sex, age, ethnicity, deprivation, education, body size, and physical activity. Lower LAEF was associated with diabetes, smoking, and all the prevalent and incident CVDs. Diabetes, smoking, and high cholesterol were associated with smaller LAV. Hypertension, IHD, AF (incident and prevalent), incident stroke, and CVD mortality were associated with larger LAV. LV and LA metrics were both independently informative in associations with prevalent disease, however LAEF showed the most consistent associations with incident CVDs. Lower LVGFI was associated with greater all-cause and CVD mortality. In secondary analyses, compared with LVGFI, LV ejection fraction showed similar but less consistent disease associations.

CONCLUSION

LA structure and function measures (LAEF and LAV) demonstrate significant associations with key prevalent and incident cardiovascular outcomes, independent of LV metrics. These measures have potential clinical utility for disease discrimination and outcome prediction.

New Horizons: the value of UK Biobank to research on endocrine and metabolic disorders

UK Biobank is an intensively characterized prospective study of 500 000 men and women, aged 40 to 69 years when recruited, between 2006 and 2010, from the general population of the United Kingdom. Established as an open-access resource for researchers worldwide to perform health research that is in the public interest, UK Biobank has collected (and continues to collect) a vast amount of data on genetic, physiological, lifestyle, and environmental factors, with prolonged follow-up of heath conditions through linkage to administrative electronic health records. The study has already demonstrated its unique value in enabling research into the determinants of common endocrine and metabolic diseases. The importance of UK Biobank, heralded as a flagship project for UK health research, will only increase over time as the number of incident disease events accrue, and the study is enhanced with additional data from blood assays (such as whole-genome sequencing, metabolomics, and proteomics), wearable technologies (including physical activity and cardiac monitors), and body imaging (magnetic resonance imaging and dual-energy X-ray absorptiometry). This unique research resource is likely to transform our understanding of the causes, diagnosis, and treatment of many endocrine and metabolic disorders.

Estimation of biological heart age using cardiovascular magnetic resonance radiomics

We developed a novel interpretable biological heart age estimation model using cardiovascular magnetic resonance radiomics measures of ventricular shape and myocardial character. We included 29,996 UK Biobank participants without cardiovascular disease. Images were segmented using an automated analysis pipeline. We extracted 254 radiomics features from the left ventricle, right ventricle, and myocardium of each study. We then used Bayesian ridge regression with tenfold cross-validation to develop a heart age estimation model using the radiomics features as the model input and chronological age as the model output. We examined associations of radiomics features with heart age in men and women, observing sex-differential patterns. We subtracted actual age from model estimated heart age to calculate a "heart age delta", which we considered as a measure of heart aging. We performed a phenome-wide association study of 701 exposures with heart age delta. The strongest correlates of heart aging were measures of obesity, adverse serum lipid markers, hypertension, diabetes, heart rate, income, multimorbidity, musculoskeletal health, and respiratory health. This technique provides a new method for phenotypic assessment relating to cardiovascular aging; further studies are required to assess whether it provides incremental risk information over current approaches.

Automated Torso Contour Extraction from Clinical Cardiac MR Slices for 3D Torso Reconstruction

Whilst the electrocardiogram (ECG) is an essential tool for diagnosing cardiac electrical abnormalities, its characteristics are dependent on anatomical variability. Specifically variation in torso geometry affects relative positions of the leads with respect to the heart. We propose a novel pipeline that uses standard cardiac magnetic resonance images to reconstruct the torso and heart, and recreate the ECG considering torso and cardiac anatomy. This requires automated extraction of the torso contours. Our method combines an initial u-net segmenter with a second network that refines contours and removes spurious segments. The networks were evaluated on a cross validation study dataset and an independent test set. The use of two-channel input, including both original image and initial segmentation, in the refinement network significantly improved performance on the independent test set, reducing the Hausdorff distance from 9.1 pixels to 4.3 pixels and increasing Dice coefficient from 0.75 to 0.93. Clinical Relevance- This method can be utilized to allow ECG simulations with personalized torso geometry which has previously been demonstrated to significantly effect QRS parameters. A clinical tool can be developed using this method that accounts for torso geometry in ECG interpretation.

Associations of cognitive performance with cardiovascular magnetic resonance phenotypes in the UK Biobank

AIMS

Existing evidence suggests links between brain and cardiovascular health. We investigated associations between cognitive performance and cardiovascular magnetic resonance (CMR) phenotypes in the UK Biobank, considering a range of potential confounders.

METHODS AND RESULTS

We studied 29 763 participants with CMR and cognitive testing, specifically, fluid intelligence (FI, 13 verbal-numeric reasoning questions), and reaction time (RT, a timed pairs matching exercise); both were considered continuous variables for modelling. We included the following CMR metrics: left and right ventricular (LV and RV) volumes in end-diastole and end-systole, LV/RV ejection fractions, LV/RV stroke volumes, LV mass, and aortic distensibility. Multivariable linear regression models were used to estimate the association of each CMR measure with FI and RT, adjusting for age, sex, smoking, education, deprivation, diabetes, hypertension, high cholesterol, prior myocardial infarction, alcohol intake, and exercise level. We report standardized beta-coefficients, 95% confidence intervals, and P-values adjusted for multiple testing. In this predominantly healthy cohort (average age 63.0 ± 7.5 years), better cognitive performance (higher FI, lower RT) was associated with larger LV/RV volumes, higher LV/RV stroke volumes, greater LV mass, and greater aortic distensibility in fully adjusted models. There was some evidence of non-linearity in the relationship between FI and LV end-systolic volume, with reversal of the direction of association at very high volumes. Associations were consistent for men and women and in different ages.

CONCLUSION

Better cognitive performance is associated with CMR measures likely representing a healthier cardiovascular phenotype. These relationships remained significant after adjustment for a range of cardiometabolic, lifestyle, and demographic factors, suggesting possible involvement of alternative disease mechanisms.

Cardiac Magnetic Resonance Radiomics Reveal Differential Impact of Sex, Age, and Vascular Risk Factors on Cardiac Structure and Myocardial Tissue

Background: Cardiovascular magnetic resonance (CMR) radiomics analysis provides multiple quantifiers of ventricular shape and myocardial texture, which may be used for detailed cardiovascular phenotyping. Objectives: We studied variation in CMR radiomics phenotypes by age and sex in healthy UK Biobank participants. Then, we examined independent associations of classical vascular risk factors (VRFs: smoking, diabetes, hypertension, high cholesterol) with CMR radiomics features, considering potential sex and age differential relationships. Design: Image acquisition was with 1.5 Tesla scanners (MAGNETOM Aera, Siemens). Three regions of interest were segmented from short axis stack images using an automated pipeline: right ventricle, left ventricle, myocardium. We extracted 237 radiomics features from each study using Pyradiomics. In a healthy subset of participants (n = 14,902) without cardiovascular disease or VRFs, we estimated independent associations of age and sex with each radiomics feature using linear regression models adjusted for body size. We then created a sample comprising individuals with at least one VRF matched to an equal number of healthy participants (n = 27,400). We linearly modelled each radiomics feature against age, sex, body size, and all the VRFs. Bonferroni adjustment for multiple testing was applied to all p-values. To aid interpretation, we organised the results into six feature clusters. Results: Amongst the healthy subset, men had larger ventricles with dimmer and less texturally complex myocardium than women. Increasing age was associated with smaller ventricles and greater variation in myocardial intensities. Broadly, all the VRFs were associated with dimmer, less varied signal intensities, greater uniformity of local intensity levels, and greater relative presence of low signal intensity areas within the myocardium. Diabetes and high cholesterol were also associated with smaller ventricular size, this association was of greater magnitude in men than women. The pattern of alteration of radiomics features with the VRFs was broadly consistent in men and women. However, the associations between intensity based radiomics features with both diabetes and hypertension were more prominent in women than men. Conclusions: We demonstrate novel independent associations of sex, age, and major VRFs with CMR radiomics phenotypes. Further studies into the nature and clinical significance of these phenotypes are needed.

Brain age estimation at tract group level and its association with daily life measures, cardiac risk factors and genetic variants

Brain age can be estimated using different Magnetic Resonance Imaging (MRI) modalities including diffusion MRI. Recent studies demonstrated that white matter (WM) tracts that share the same function might experience similar alterations. Therefore, in this work, we sought to investigate such issue focusing on five WM bundles holding that feature that is Association, Brainstem, Commissural, Limbic and Projection fibers, respectively. For each tract group, we estimated brain age for 15,335 healthy participants from United Kingdom Biobank relying on diffusion MRI data derived endophenotypes, Bayesian ridge regression modeling and 10 fold-cross validation. Furthermore, we estimated brain age for an Ensemble model that gathers all the considered WM bundles. Association analysis was subsequently performed between the estimated brain age delta as resulting from the six models, that is for each tract group as well as for the Ensemble model, and 38 daily life style measures, 14 cardiac risk factors and cardiovascular magnetic resonance imaging features and genetic variants. The Ensemble model that used all tracts from all fiber groups (FG) performed better than other models to estimate brain age. Limbic tracts based model reached the highest accuracy with a Mean Absolute Error (MAE) of 5.08, followed by the Commissural ([Formula: see text]), Association ([Formula: see text]), and Projection ([Formula: see text]) ones. The Brainstem tracts based model was the less accurate achieving a MAE of 5.86. Accordingly, our study suggests that the Limbic tracts experience less brain aging or allows for more accurate estimates compared to other tract groups. Moreover, the results suggest that Limbic tract leads to the largest number of significant associations with daily lifestyle factors than the other tract groups. Lastly, two SNPs were significantly (p value [Formula: see text]) associated with brain age delta in the Projection fibers. Those SNPs are mapped to HIST1H1A and SLC17A3 genes.

Artificial Intelligence in Computer Vision: Cardiac MRI and Multimodality Imaging Segmentation

Purpose of Review

Anatomical segmentation has played a major role within clinical cardiology. Novel techniques through artificial intelligence-based computer vision have revolutionized this process through both automation and novel applications. This review discusses the history and clinical context of cardiac segmentation to provide a framework for a survey of recent manuscripts in artificial intelligence and cardiac segmentation. We aim to clarify for the reader the clinical question of "Why do we segment?" in order to understand the question of "Where is current research and where should be?".

Recent Findings

There has been increasing research in cardiac segmentation in recent years. Segmentation models are most frequently based on a U-Net structure. Multiple innovations have been added in terms of pre-processing or connection to analysis pipelines. Cardiac MRI is the most frequently segmented modality, which is due in part to the presence of publically-available, moderately sized, computer vision competition datasets. Further progress in data availability, model explanation, and clinical integration are being pursued.

Summary

The task of cardiac anatomical segmentation has experienced massive strides forward within the past five years due to convolutional neural networks. These advances provide a basis for streamlining image analysis, and a foundation for further analysis both by computer and human systems. While technical advances are clear, clinical benefit remains nascent. Novel approaches may improve measurement precision by decreasing inter-reader variability and appear to also have the potential for larger-reaching effects in the future within integrated analysis pipelines.

Automated Quality-Controlled Cardiovascular Magnetic Resonance Pericardial Fat Quantification Using a Convolutional Neural Network in the UK Biobank

Background: Pericardial adipose tissue (PAT) may represent a novel risk marker for cardiovascular disease. However, absence of rapid radiation-free PAT quantification methods has precluded its examination in large cohorts. Objectives: We developed a fully automated quality-controlled tool for cardiovascular magnetic resonance (CMR) PAT quantification in the UK Biobank (UKB). Methods: Image analysis comprised contouring an en-bloc PAT area on four-chamber cine images. We created a ground truth manual analysis dataset randomly split into training and test sets. We built a neural network for automated segmentation using a Multi-residual U-net architecture with incorporation of permanently active dropout layers to facilitate quality control of the model's output using Monte Carlo sampling. We developed an in-built quality control feature, which presents predicted Dice scores. We evaluated model performance against the test set (n = 87), the whole UKB Imaging cohort (n = 45,519), and an external dataset (n = 103). In an independent dataset, we compared automated CMR and cardiac computed tomography (CCT) PAT quantification. Finally, we tested association of CMR PAT with diabetes in the UKB (n = 42,928). Results: Agreement between automated and manual segmentations in the test set was almost identical to inter-observer variability (mean Dice score = 0.8). The quality control method predicted individual Dice scores with Pearson r = 0.75. Model performance remained high in the whole UKB Imaging cohort and in the external dataset, with medium-good quality segmentation in 94.3% (mean Dice score = 0.77) and 94.4% (mean Dice score = 0.78), respectively. There was high correlation between CMR and CCT PAT measures (Pearson r = 0.72, p-value 5.3 ×10-18). Larger CMR PAT area was associated with significantly greater odds of diabetes independent of age, sex, and body mass index. Conclusions: We present a novel fully automated method for CMR PAT quantification with good model performance on independent and external datasets, high correlation with reference standard CCT PAT measurement, and expected clinical associations with diabetes.

Associations of Meat and Fish Consumption With Conventional and Radiomics Cardiovascular Magnetic Resonance Phenotypes in the UK Biobank

Background: Greater red and processed meat consumption has been linked to adverse cardiovascular outcomes. However, the impact of these exposures on cardiovascular magnetic resonance (CMR) phenotypes has not been adequately studied.

Objective: We describe novel associations of meat intake with cardiovascular phenotypes and investigate underlying mechanisms through consideration of a range of covariates.

Design: We studied 19,408 UK Biobank participants with CMR data available. Average daily red and processed meat consumption was determined through food frequency questionnaires and expressed as a continuous variable. Oily fish was studied as a comparator, previously associated with favourable cardiac outcomes. We considered associations with conventional CMR indices (ventricular volumes, ejection fraction, stroke volume, left ventricular mass), novel CMR radiomics features (shape, first-order, texture), and arterial compliance measures (arterial stiffness index, aortic distensibility). We used multivariable linear regression to investigate relationships between meat intake and cardiovascular phenotypes, adjusting for confounders (age, sex, deprivation, educational level, smoking, alcohol intake, exercise) and potential covariates on the causal pathway (hypertension, hypercholesterolaemia, diabetes, body mass index).

Results: Greater red and processed meat consumption was associated with an unhealthy pattern of biventricular remodelling, worse cardiac function, and poorer arterial compliance. In contrast, greater oily fish consumption was associated with a healthier cardiovascular phenotype and better arterial compliance. There was partial attenuation of associations between red meat and conventional CMR indices with addition of covariates potentially on the causal pathway, indicating a possible mechanistic role for these cardiometabolic morbidities. However, other associations were not altered with inclusion of these covariates, suggesting importance of alternative biological mechanisms underlying these relationships. Radiomics analysis provided deeper phenotyping, demonstrating association of the different dietary habits with distinct ventricular geometry and left ventricular myocardial texture patterns.

Conclusions: Greater red and processed meat consumption is associated with impaired cardiovascular health, both in terms of markers of arterial disease and of cardiac structure and function. Cardiometabolic morbidities appeared to have a mechanistic role in the associations of red meat with ventricular phenotypes, but less so for other associations suggesting importance of alternative mechanism for these relationships.

Adverse cardiovascular magnetic resonance phenotypes are associated with greater likelihood of incident coronavirus disease 2019: findings from the UK Biobank

BACKGROUND

Coronavirus disease 2019 (COVID-19) disproportionately affects older people. Observational studies suggest indolent cardiovascular involvement after recovery from acute COVID-19. However, these findings may reflect pre-existing cardiac phenotypes.

AIMS

We tested the association of baseline cardiovascular magnetic resonance (CMR) phenotypes with incident COVID-19.

METHODS

We studied UK Biobank participants with CMR imaging and COVID-19 testing. We considered left and right ventricular (LV, RV) volumes, ejection fractions, and stroke volumes, LV mass, LV strain, native T1, aortic distensibility, and arterial stiffness index. COVID-19 test results were obtained from Public Health England. Co-morbidities were ascertained from self-report and hospital episode statistics (HES). Critical care admission and death were from HES and death register records. We investigated the association of each cardiovascular measure with COVID-19 test result in multivariable logistic regression models adjusting for age, sex, ethnicity, deprivation, body mass index, smoking, diabetes, hypertension, high cholesterol, and prior myocardial infarction.

RESULTS

We studied 310 participants (n = 70 positive). Median age was 63.8 [57.5, 72.1] years; 51.0% (n = 158) were male. 78.7% (n = 244) were tested in hospital, 3.5% (n = 11) required critical care admission, and 6.1% (n = 19) died. In fully adjusted models, smaller LV/RV end-diastolic volumes, smaller LV stroke volume, and poorer global longitudinal strain were associated with significantly higher odds of COVID-19 positivity.

DISCUSSION

We demonstrate association of pre-existing adverse CMR phenotypes with greater odds of COVID-19 positivity independent of classical cardiovascular risk factors.

CONCLUSIONS

Observational reports of cardiovascular involvement after COVID-19 may, at least partly, reflect pre-existing cardiac status rather than COVID-19 induced alterations.