Abstract:Objective: In this paper, we propose a fully automatic framework for myocardium segmentation of delayed-enhancement (DE) MRI images without relying on prior patient-specific information. Methods: We employ a multicomponent Gaussian mixture model to deal with the intensity heterogeneity of myocardium caused by the infarcts. To differentiate the myocardium from other tissues with similar intensities, while at the same time maintain spatial continuity, we introduce a coupled level set (CLS) to regularize the posterior probability. The CLS, as a spatial regularization, can be adapted to the image characteristics dynamically. We also introduce an image intensity gradient based term into the CLS, adding an extra force to the posterior probability based framework, to improve the accuracy of myocardium boundary delineation. The prebuilt atlases are propagated to the target image to initialize the framework. Results: The proposed method was tested on datasets of 22 clinical cases, and achieved Dice similarity coefficients of 87.43 ± 5.62% (endocardium), 90.53 ± 3.20% (epicardium) and 73.58 ± 5.58% (myocardium), which have outperformed three variants of the classic segmentation methods. Conclusion: The results can provide a benchmark for the myocardial segmentation in the literature. Significance: DE MRI provides an important tool to assess the viability of myocardium. The accurate segmentation of myocardium, which is a prerequisite for further quantitative analysis of myocardial infarction (MI) region, can provide important support for the diagnosis and treatment management for MI patients.Objective: In this paper, we propose a fully automatic framework for myocardium segmentation of delayed-enhancement (DE) MRI images without relying on prior patient-specific information. Methods: We employ a multicomponent Gaussian mixture model to deal with the intensity heterogeneity of myocardium caused by the infarcts. To differentiate the myocardium from other tissues with similar intensities, while at the same time maintain spatial continuity, we introduce a coupled level set (CLS) to regularize the posterior probability. The CLS, as a spatial regularization, can be adapted to the image characteristics dynamically. We also introduce an image intensity gradient based term into the CLS, adding an extra force to the posterior probability based framework, to improve the accuracy of myocardium boundary delineation. The prebuilt atlases are propagated to the target image to initialize the framework. Results: The proposed method was tested on datasets of 22 clinical cases, and achieved Dice similarity coefficients of 87.43 ± 5.62% (endocardium), 90.53 ± 3.20% (epicardium) and 73.58 ± 5.58% (myocardium), which have outperformed three variants of the classic segmentation methods. Conclusion: The results can provide a benchmark for the myocardial segmentation in the literature. Significance: DE MRI provides an important tool to assess the viability of myocardium. The accurate segmentation of myocardium, which is a prerequisite for further quantitative analysis of myocardial infarction (MI) region, can provide important support for the diagnosis and treatment management for MI patients.

HOCM-Net: 3D Coarse-to-fine Structural Prior Fusion Based Segmentation Network for the Surgical Planning of Hypertrophic Obstructive Cardiomyopathy

A Fully Automated Two-Stage Segmentation Approach for Late Gadolinium-Enhanced Cardiac Magnetic Resonance Images in Personalized Cardiac Modeling

Automatic Segmentation of Aortic and Mitral Valves for Heart Surgical Planning of Hypertrophic Obstructive Cardiomyopathy.

Segmentation of Cardiac Tissues and Organs for CCTA Images Based on a Deep Learning Model

Three-Dimensional Heart Modeling of Hypertrophic Obstructive Cardiomyopathy for In Situ Patient-Specific Simulation to Optimize Septal Myectomy

Improving myocardial pathology segmentation with U-Net++ and EfficientSeg from multi-sequence cardiac magnetic resonance images

PC-U Net: Learning to Jointly Reconstruct and Segment the Cardiac Walls in 3D from CT Data

A preoperative planning procedure of septal myectomy for hypertrophic obstructive cardiomyopathy using image-based computational fluid dynamics simulations and shape optimization

Regional perception and multi-scale feature fusion network for cardiac segmentation

Patient-specific computational fluid dynamics for hypertrophic obstructive cardiomyopathy

Fused Segmentation of Geometric Models for Myocardium and Coronary Artery via Medial Axis

MyoPS-Net: Myocardial pathology segmentation with flexible combination of multi-sequence CMR images

Fully Automatic initialization and segmentation of left and right ventricles for large-scale cardiac MRI using a deeply supervised network and 3D-ASM

Deep U-Net architecture with curriculum learning for myocardial pathology segmentation in multi-sequence cardiac magnetic resonance images

Myocardium Segmentation from DE MRI Using Multicomponent Gaussian Mixture Model and Coupled Level Set

AMCNet: Attention-Based Multiscale Convolutional Network for DCM MRI Segmentation

Automatic right ventricular segmentation for cine cardiac magnetic resonance images based on a new deep atlas network

Toward Right Ventricle Segmentation in Cardiac MRIs via Feature Multiplexing and Multiscale Weighted Convolution

A Multi-Task Network For Cardiac Magnetic Resonance Image Segmentation And Classification

A Multi-Scale Network with the Encoder-Decoder Structure for CMR Segmentation

Automatic Segmentation of Different Pathologies from Cardiac Cine MRI Using Registration and Multiple Component EM Estimation