Search

We present a framework for combining a cardiac motion atlas with non-motion data. The atlas represents cardiac cycle motion across a number of subjects in a common space based on rich motion descriptors capturing 3D displacement, velocity, strain and strain rate. The non-motion data are derived from a variety of sources such as imaging, electrocardiogram (ECG) and clinical reports. Once in the atlas space, we apply a novel supervised learning approach based on random projections and ensemble learning to learn the relationship between the atlas data and some desired clinical output. We apply our framework to the problem of predicting response to Cardiac Resynchronisation Therapy (CRT). Using a cohort of 34 patients selected for CRT using conventional criteria, results show that the combination of motion and non-motion data enables CRT response to be predicted with 91.2% accuracy (100% sensitivity and 62.5% specificity), which compares favourably with the current state-of-the-art in CRT response prediction.

Segmentation of medical images, particularly late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) used for visualizing diseased atrial structures, is a crucial first step for ablation treatment of atrial fibrillation. However, direct segmentation of LGE-MRIs is challenging due to the varying intensities caused by contrast agents. Since most clinical studies have relied on manual, labor-intensive approaches, automatic methods are of high interest, particularly optimized machine learning approaches. To address this, we organized the 2018 Left Atrium Segmentation Challenge using 154 3D LGE-MRIs, currently the world's largest atrial LGE-MRI dataset, and associated labels of the left atrium segmented by three medical experts, ultimately attracting the participation of 27 international teams. In this paper, extensive analysis of the submitted algorithms using technical and biological metrics was performed by undergoing subgroup analysis and conducting hyper-parameter analysis, offering an overall picture of the major design choices of convolutional neural networks (CNNs) and practical considerations for achieving state-of-the-art left atrium segmentation. Results show that the top method achieved a Dice score of 93.2% and a mean surface to surface distance of 0.7 mm, significantly outperforming prior state-of-the-art. Particularly, our analysis demonstrated that double sequentially used CNNs, in which a first CNN is used for automatic region-of-interest localization and a subsequent CNN is used for refined regional segmentation, achieved superior results than traditional methods and machine learning approaches containing single CNNs. This large-scale benchmarking study makes a significant step towards much-improved segmentation methods for atrial LGE-MRIs, and will serve as an important benchmark for evaluating and comparing the future works in the field. Furthermore, the findings from this study can potentially be extended to other imaging datasets and modalities, having an impact on the wider medical imaging community. (C) 2020 Elsevier B.V. All rights reserved.

Background The genetic and environmental factors that define cardiac structure and function, particularly at the intersection between health and the earliest stages of disease, remain poorly characterised. Conventional cardiovascular imaging provides limited, semi-quantitative, and global metrics of the heart, and these methods are insensitive to regional variation. We aimed to assess whether three-dimensional (3D) cardiac magnetic resonance (CMR) phenotyping could provide methodological, statistical, and scaling advantages for population studies. Methods Volunteers without self-reported cardiovascular disease were recruited prospectively. A cardiac atlas-based software was developed to quantitatively analyse two-dimensional (2D) and 3D CMRs. Phenotypes (eg, wall thickness and function) were extracted at 46 000 points across the heart and genotyping and whole-exome sequencing performed. Associations between 3D phenotypes, clinical variables, and genotypes were studied with 3D regression models and Bayesian latent factor analysis. Findings 1850 volunteers (mean age 41 years, SD 13) took part in the study. Automatically analysed 3D images were more accurate than 2D images at defining cardiac surfaces, enabling a reduction in the sample size required for epidemiological and genetic studies of the heart. Computational, high-dimensional 3D phenotyping revealed that systolic blood pressure, body composition, and genetic variation (both rare and common) were associated with regional rather than global changes in cardiac phenotypes, which were not detected by 2D CMR. 3D CMR revealed latent precursors of the hypertensive heart phenotype in healthy individuals that were previously unappreciated. In a 3D genome-wide association study we detected a larger number of loci significantly (p −10 ) associated with regional cardiac variation, showing genetic control of localised cardiac physiological features. Interpretation We have shown that quantitative 3D CMR combined with computational modelling and advanced statistical analysis provide new insights into the genetic and anthropometric determinants of cardiac morphological and physiological features. Our approach provides new opportunities for understanding the structure and function of the human heart, and reveals hidden features only apparent through big data analytical methods. This approach can be applied at scale to very large datasets for mechanistic or interventional studies. Funding Medical Research Council, British Heart Foundation, Fondation Leducq grants.