Brecht Heyde

Bio: Brecht Heyde is an academic researcher from Katholieke Universiteit Leuven. The author has contributed to research in topics: Segmentation & Image registration. The author has an hindex of 19, co-authored 70 publications receiving 1155 citations. Previous affiliations of Brecht Heyde include Catholic University of Leuven.

Current state of three-dimensional myocardial strain estimation using echocardiography.

Abstract: With the developments in ultrasound transducer technology and both hardware and software computing, real-time volumetric imaging has become widely available, accompanied by various methods of assessing three-dimensional (3D) myocardial strain, often referred to as 3D speckle-tracking echocardiographic methods. Indeed, these methods should provide cardiologists with a better view of regional myocardial mechanics, which might be important for diagnosis, prognosis, and therapy. However, currently available 3D speckle-tracking echocardiographic methods are based on different algorithms, which introduce substantial differences between them and make them not interchangeable with each other. Therefore, it is critical that each 3D speckle-tracking echocardiographic method is validated individually before being introduced into clinical practice. In this review, the authors discuss differences and similarities of the currently available 3D strain estimation approaches and provide an overview of the current status of their validation.

Standardized Evaluation System for Left Ventricular Segmentation Algorithms in 3D Echocardiography

Abstract: Real-time 3D Echocardiography (RT3DE) has been proven to be an accurate tool for left ventricular (LV) volume assessment. However, identification of the LV endocardium remains a challenging task, mainly because of the low tissue/blood contrast of the images combined with typical artifacts. Several semi and fully automatic algorithms have been proposed for segmenting the endocardium in RT3DE data in order to extract relevant clinical indices, but a systematic and fair comparison between such methods has so far been impossible due to the lack of a publicly available common database. Here, we introduce a standardized evaluation framework to reliably evaluate and compare the performance of the algorithms developed to segment the LV border in RT3DE. A database consisting of 45 multivendor cardiac ultrasound recordings acquired at different centers with corresponding reference measurements from three experts are made available. The algorithms from nine research groups were quantitatively evaluated and compared using the proposed online platform. The results showed that the best methods produce promising results with respect to the experts' measurements for the extraction of clinical indices, and that they offer good segmentation precision in terms of mean distance error in the context of the experts' variability range. The platform remains open for new submissions.

Cardiovascular magnetic resonance myocardial feature tracking using a non-rigid, elastic image registration algorithm: assessment of variability in a real-life clinical setting

Abstract: Cardiovascular magnetic resonance myocardial feature tracking (CMR-FT) is a promising technique for quantification of myocardial strain from steady-state free precession (SSFP) cine images. We sought to determine the variability of CMR-FT using a non-rigid elastic registration algorithm recently available in a commercial software package (Segment, Medviso) in a real-life clinical setting. Firstly, we studied the variability in a healthy volunteer who underwent 10 CMR studies over five consecutive days. Secondly, 10 patients were selected from our CMR database yielding normal findings (normal group). Finally, we prospectively studied 10 patients with known or suspected myocardial pathology referred for further investigation to CMR (patient group). In the patient group a second study was performed respecting an interval of 30 min between studies. All studies were manually segmented at the end-diastolic phase by three observers. In all subjects left ventricular (LV) circumferential and radial strain were calculated in the short-axis direction (EccSAX and ErrSAX, respectively) and longitudinal strain in the long-axis direction (EllLAX). The level of CMR experience of the observers was 2 weeks, 6 months and >20 years. Mean contouring time was 7 ± 1 min, mean FT calculation time 13 ± 2 min. Intra- and inter-observer variability was good to excellent with an coefficient of reproducibility (CR) ranging 1.6% to 11.5%, and 1.7% to 16.0%, respectively and an intraclass correlation coefficient (ICC) ranging 0.89 to 1.00 and 0.74 to 0.99, respectively. Variability considerably increased in the test-retest setting with a CR ranging 4.2% to 29.1% and an ICC ranging 0.66 to 0.95 in the patient group. Variability was not influenced by level of expertise of the observers. Neither did the presence of myocardial pathology at CMR negatively impact variability. However, compared to global myocardial strain, segmental myocardial strain variability increased with a factor 2–3, in particular for the basal and apical short-axis slices. CMR-FT using non-rigid, elastic registration is a reproducible approach for strain analysis in patients routinely scheduled for CMR, and is not influenced by the level of training. However, further improvement is needed to reliably depict small variations in segmental myocardial strain.

Elastic Image Registration Versus Speckle Tracking for 2-D Myocardial Motion Estimation: A Direct Comparison In Vivo

Abstract: Despite the availability of multiple solutions for assessing myocardial strain by ultrasound, little is currently known about the relative performance of the different methods. In this study, we sought to contrast two strain estimation techniques directly (speckle tracking and elastic registration) in an in vivo setting by comparing both to a gold standard reference measurement. In five open-chest sheep instrumented with ultrasonic microcrystals, 2-D images were acquired with a GE Vivid7 ultrasound system. Radial (eRR), longitudinal (eLL), and circumferential strain (eCC) were estimated during four inotropic stages: at rest, during esmolol and dobutamine infusion, and during acute ischemia. The correlation of the end-systolic strain values of a well-validated speckle tracking approach and an elastic registration method against sonomicrometry were comparable for eLL (r=0.70 versus r=0.61 , respectively; p=0.32) and eCC (r=0.73 versus r=0.80 respectively; p=0.31). However, the elastic registration method performed considerably better for eRR (r=0.64 versus r=0.85 respectively; p=0.09). Moreover, the bias and limits of agreement with respect to the reference strain estimates were statistically significantly smaller in this direction (p <; 0.001). This could be related to regularization which is imposed during the motion estimation process as opposed to an a posteriori regularization step in the speckle tracking method. Whether one method outperforms the other in detecting dysfunctional regions remains the topic of future research.

V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation

Abstract: Convolutional Neural Networks (CNNs) have been recently employed to solve problems from both the computer vision and medical image analysis fields. Despite their popularity, most approaches are only able to process 2D images while most medical data used in clinical practice consists of 3D volumes. In this work we propose an approach to 3D image segmentation based on a volumetric, fully convolutional, neural network. Our CNN is trained end-to-end on MRI volumes depicting prostate, and learns to predict segmentation for the whole volume at once. We introduce a novel objective function, that we optimise during training, based on Dice coefficient. In this way we can deal with situations where there is a strong imbalance between the number of foreground and background voxels. To cope with the limited number of annotated volumes available for training, we augment the data applying random non-linear transformations and histogram matching. We show in our experimental evaluation that our approach achieves good performances on challenging test data while requiring only a fraction of the processing time needed by other previous methods.

V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation

Abstract: Convolutional Neural Networks (CNNs) have been recently employed to solve problems from both the computer vision and medical image analysis fields. Despite their popularity, most approaches are only able to process 2D images while most medical data used in clinical practice consists of 3D volumes. In this work we propose an approach to 3D image segmentation based on a volumetric, fully convolutional, neural network. Our CNN is trained end-to-end on MRI volumes depicting prostate, and learns to predict segmentation for the whole volume at once. We introduce a novel objective function, that we optimise during training, based on Dice coefficient. In this way we can deal with situations where there is a strong imbalance between the number of foreground and background voxels. To cope with the limited number of annotated volumes available for training, we augment the data applying random non-linear transformations and histogram matching. We show in our experimental evaluation that our approach achieves good performances on challenging test data while requiring only a fraction of the processing time needed by other previous methods.

Deep Learning Techniques for Automatic MRI Cardiac Multi-Structures Segmentation and Diagnosis: Is the Problem Solved?

Abstract: Delineation of the left ventricular cavity, myocardium, and right ventricle from cardiac magnetic resonance images (multi-slice 2-D cine MRI) is a common clinical task to establish diagnosis. The automation of the corresponding tasks has thus been the subject of intense research over the past decades. In this paper, we introduce the “Automatic Cardiac Diagnosis Challenge” dataset (ACDC), the largest publicly available and fully annotated dataset for the purpose of cardiac MRI (CMR) assessment. The dataset contains data from 150 multi-equipments CMRI recordings with reference measurements and classification from two medical experts. The overarching objective of this paper is to measure how far state-of-the-art deep learning methods can go at assessing CMRI, i.e., segmenting the myocardium and the two ventricles as well as classifying pathologies. In the wake of the 2017 MICCAI-ACDC challenge, we report results from deep learning methods provided by nine research groups for the segmentation task and four groups for the classification task. Results show that the best methods faithfully reproduce the expert analysis, leading to a mean value of 0.97 correlation score for the automatic extraction of clinical indices and an accuracy of 0.96 for automatic diagnosis. These results clearly open the door to highly accurate and fully automatic analysis of cardiac CMRI. We also identify scenarios for which deep learning methods are still failing. Both the dataset and detailed results are publicly available online, while the platform will remain open for new submissions.

Myocardial strain imaging: how useful is it in clinical decision making?

Abstract: Myocardial strain is a principle for quantification of left ventricular (LV) function which is now feasible with speckle-tracking echocardiography. The best evaluated strain parameter is global longitudinal strain (GLS) which is more sensitive than left ventricular ejection fraction (LVEF) as a measure of systolic function, and may be used to identify sub-clinical LV dysfunction in cardiomyopathies. Furthermore, GLS is recommended as routine measurement in patients undergoing chemotherapy to detect reduction in LV function prior to fall in LVEF. Intersegmental variability in timing of peak myocardial strain has been proposed as predictor of risk of ventricular arrhythmias. Strain imaging may be applied to guide placement of the LV pacing lead in patients receiving cardiac resynchronization therapy. Strain may also be used to diagnose myocardial ischaemia, but the technology is not sufficiently standardized to be recommended as a general tool for this purpose. Peak systolic left atrial strain is a promising supplementary index of LV filling pressure. The strain imaging methodology is still undergoing development, and further clinical trials are needed to determine if clinical decisions based on strain imaging result in better outcome. With this important limitation in mind, strain may be applied clinically as a supplementary diagnostic method.

Anatomically Constrained Neural Networks (ACNNs): Application to Cardiac Image Enhancement and Segmentation

Abstract: Incorporation of prior knowledge about organ shape and location is key to improve performance of image analysis approaches In particular, priors can be useful in cases where images are corrupted and contain artefacts due to limitations in image acquisition The highly constrained nature of anatomical objects can be well captured with learning-based techniques However, in most recent and promising techniques such as CNN-based segmentation it is not obvious how to incorporate such prior knowledge State-of-the-art methods operate as pixel-wise classifiers where the training objectives do not incorporate the structure and inter-dependencies of the output To overcome this limitation, we propose a generic training strategy that incorporates anatomical prior knowledge into CNNs through a new regularisation model, which is trained end-to-end The new framework encourages models to follow the global anatomical properties of the underlying anatomy ( eg shape, label structure) via learnt non-linear representations of the shape We show that the proposed approach can be easily adapted to different analysis tasks ( eg image enhancement, segmentation) and improve the prediction accuracy of the state-of-the-art models The applicability of our approach is shown on multi-modal cardiac data sets and public benchmarks In addition, we demonstrate how the learnt deep models of 3-D shapes can be interpreted and used as biomarkers for classification of cardiac pathologies