Fully automatic segmentation of left ventricular anatomy in 3-D LGE-MRI

Kurzendorfer, Tanja; Forman, Christoph; Schmidt, Michaela; Tillmanns, Christoph; Maier, Andreas; Brost, Alexander

doi:10.1016/j.compmedimag.2017.05.001

Cited by 29 publications

(25 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The presented workflow could be speed up by reducing the number of needed interaction. Particularly, the delineation work on LGE-MRI SAX view could be automated such as seen in [15]. During the intervention, thickness information confirmed to be a driver of clinician's confidence when ablating, as the safety margin before perforating cardiac's wall was known.…”

Section: Resultsmentioning

confidence: 94%

Multimodal Image Integration for Catheter Ablation of Ventricular Tachycardia

Courtial¹,

Simon²,

Léderlin³

et al. 2017

Computing in Cardiology Conference (CinC)

View full text Add to dashboard Cite

Radiofrequency catheter ablation is an important therapeutic option for patients with recurrent ventricular tachycardia (VT). The procedure aims at cauterizing regions of the ventricle. A good knowledge of anatomical structures and tissues properties can improve procedure safety and efficacy. This work aimed at analysing the feasibility of a workflow based on multimodal image integration to assist the intervention. We focused here on a model including myocardial thickness and fibrosis extracted from late gadolinium enhanced MRI (LGE-MRI) sequence, and left ventricular anatomy segmented from multiphase computed tomography (CT). The produced model can be imported within Carto-3 c system (Biosense Webster Inc, Diamond Bar, CA), the software used during the ablation procedure. Four patients cases have been included in this preliminary study. The four produced patient specific models have been validated by a clinician, and two of them have been integrated within Carto-3 c system and used during the intervention. IntroductionCatheter ablation, used for 30 years, aims at cauterizing ventricle's regions responsible for VT initiation and maintenance. To do so, a localized delivery of radiofrequency energy is applied. It's an important therapeutic option for patients suffering of recurrent VT resistant to antiarrythmic drugs. Depending on the needs, an endocardial or an epicardial approach can be considered. Ablation of ventricle regions is not an innocuous procedure, and it raises important questions to proceed safely and efficiently. During the procedure, the physician has to know (i) the catheter's position against the VT sources, (ii) which spots to avoid (arteries, phrenic nerve), (iii) which regions could be potential targets, and (iv) what constraint can the catheter apply on the cardiac wall without perforating it (myocardium thickness).Answers have been found to these different requirements. Electroanotomical mapping systems have been early recognized as a powerful assistance tool for VT ablation [1]. Studies such as [2] have shown the benefit of visualizing structures at risk during VT ablation, segmented from CT series. A multimodal model integration has been proposed in [3], including both CT and MRI enhanced sequences. It contains wall thining (WT) zones, defined as zones where myocardium wall's thickness is less than 5 mm [4] segmented from CT, and scare core and grey zone, defined as more than 50% and 35-50% of maximal myocardial signal [5] respectively. The interest of such informations has been later confirmed in [6], as a good predictor of local abnormal ventricular activity (LAVA).Multimodality has been recognized in different heart surgery as a powerful tool to plan and guide interventions [7][8] [9]. We propose here a patient specific multimodal model built from CT and LGE-MRI, images including the left endocardium segmented on CT, and fibrosis and myocardium's thickness segmented on LGE-MRI. LGE-MRI is registered on CT in a process including Cine-MRI as an intermediate step. The objective is ...

show abstract

Section: Resultsmentioning

confidence: 94%

Multimodal Image Integration for Catheter Ablation of Ventricular Tachycardia

Courtial¹,

Simon²,

Léderlin³

et al. 2017

Computing in Cardiology Conference (CinC)

View full text Add to dashboard Cite

show abstract

“…Kurzendorfer et al utilized a two-step registration-based method to delineate the LV myocardium and evaluated their proposed method on 30 clinical 3D LGE-MRI datasets from individual subjects obtained at two different clinical sites, which reported DSC of 83% and 80% for the endocardium and epicardium, respectively. 48 We used a majority voting system for final label prediction from three orthogonal views that result in more confident prediction. This is particularly useful in ambiguous cases, where suppression of false-positive errors can be achieved.…”

Section: Discussionmentioning

confidence: 99%

Fully automated segmentation of left ventricular scar from 3D late gadolinium enhancement magnetic resonance imaging using a cascaded multi‐planar U‐Net (CMPU‐Net)

et al. 2020

View full text Add to dashboard Cite

Purpose: Three-dimensional (3D) late gadolinium enhancement magnetic resonance (LGE-MR) imaging enables the quantification of myocardial scar at high resolution with unprecedented volumetric visualization. Automated segmentation of myocardial scar is critical for the potential clinical translation of this technique given the number of tomographic images acquired. Methods: In this paper, we describe the development of cascaded multi-planar U-Net (CMPU-Net) to efficiently segment the boundary of the left ventricle (LV) myocardium and scar from 3D LGE-MR images. In this approach, two subnets, each containing three U-Nets, were cascaded to first segment the LV myocardium and then segment the scar within the presegmented LV myocardium. The U-Nets were trained separately using two-dimensional (2D) slices extracted from axial, sagittal, and coronal slices of 3D LGE-MR images. We used 3D LGE-MR images from 34 subjects with chronic ischemic cardiomyopathy. The U-Nets were trained using 8430 slices, extracted in three orthogonal directions from 18 images. In the testing phase, the outputs of U-Nets of each subnet were combined using the majority voting system for final label prediction of each voxel in the image. The developed method was tested for accuracy by comparing its results to manual segmentations of LV myocardium and LV scar from 7250 slices extracted from 16 3D LGE-MR images. Our method was also compared to numerous alternative methods based on machine learning, energy minimization, and intensity-thresholds. Results: Our algorithm reported a mean dice similarity coefficient (DSC), absolute volume difference (AVD), and Hausdorff distance (HD) of 85.14% AE 3.36%, 43.72 AE 27.18 cm 3 , and 19.21 AE 4.74 mm for determining the boundaries of LV myocardium from LGE-MR images. Our method also yielded a mean DSC, AVD, and HD of 88.61% AE 2.54%, 9.33 AE 7.24 cm 3 , and 17.04 AE 9.93 mm for LV scar segmentation on the unobserved test dataset. Our method significantly outperformed the alternative techniques in segmentation accuracy (P < 0.05). Conclusions: The CMPU-Net method provided fully automated segmentation of LV scar from 3DLGE-MR images and outperformed the alternative techniques.

show abstract

“…In the existing literature, two main families of techniques have been proposed to automatically segment LGE-MRI data. The first one segments directly the LGE-MRI images by using different techniques such as graph-cuts [1], atlasbased registration [2], or more recently Convolutional Neural Networks (CNNs) [3]. However, these techniques generally lack robustness due to the limited availability of LGE-MRI datasets for training.…”

Section: Introductionmentioning

confidence: 99%

Combining Multi-Sequence and Synthetic Images for Improved Segmentation of Late Gadolinium Enhancement Cardiac MRI

Campello

Martín-Isla

Izquierdo

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Accurate segmentation of the cardiac boundaries in late gadolinium enhancement magnetic resonance images (LGE-MRI) is a fundamental step for accurate quantification of scar tissue. However, while there are many solutions for automatic cardiac segmentation of cine images, the presence of scar tissue can make the correct delineation of the myocardium in LGE-MRI challenging even for human experts. As part of the Multi-Sequence Cardiac MR Segmentation Challenge, we propose a solution for LGE-MRI segmentation based on two components. First, a generative adversarial network is trained for the task of modality-tomodality translation between cine and LGE-MRI sequences to obtain extra synthetic images for both modalities. Second, a deep learning model is trained for segmentation with different combinations of original, augmented and synthetic sequences. Our results based on three magnetic resonance sequences (LGE, bSSFP and T2) from 45 different patients show that the multi-sequence model training integrating synthetic images and data augmentation improves in the segmentation over conventional training with real datasets. In conclusion, the accuracy of the segmentation of LGE-MRI images can be improved by using complementary information provided by non-contrast MRI sequences.

show abstract

Fully automatic segmentation of left ventricular anatomy in 3-D LGE-MRI

Cited by 29 publications

References 30 publications

Multimodal Image Integration for Catheter Ablation of Ventricular Tachycardia

Multimodal Image Integration for Catheter Ablation of Ventricular Tachycardia

Fully automated segmentation of left ventricular scar from 3D late gadolinium enhancement magnetic resonance imaging using a cascaded multi‐planar U‐Net (CMPU‐Net)

Combining Multi-Sequence and Synthetic Images for Improved Segmentation of Late Gadolinium Enhancement Cardiac MRI

Contact Info

Product

Resources

About