基于深度学习的心脏图像分割研究现状

面对严重的医学影像分析缺口,深度学习的发展能够满足国内医疗行业的需求。心脏图像的处理方法可大致分为传统的图像处理技术、基于图谱的方法（atlas-based methods）、基于模型的方法（model-based methods）以及目前热门的采用机器学习和深度学习的方法。在深度学习兴起之前,传统的机器学习技术如模型法和图集法在心脏图像分割中有良好表现,但通常需要大量的特征工程知识或先验知识才能获得令人满意的精度。而基于深度学习的算法能从数据中自动发现复杂的特征以进行对象检测和分割。得益于先进的计算机硬件以及更多可用于训练的数据集,基于深度学习的分割算法已超越了以往的传统方法。本文回顾了2012—2022年有关心室、心外膜和心包脂肪的图像处理的各项方法、衡量指标及其目前的研究现状,并结合分割技术的发展,讨论了心脏分割的发展趋势。     

CCBE1 Is Essential for Epicardial Function during Myocardium Development

The epicardium is a single cell layer of mesothelial cells that plays a critical role during heart development contributing to different cardiac cell types of the developing heart through epithelial-to-mesenchymal transition (EMT). Moreover, the epicardium is a source of secreted growth factors that promote myocardial growth. CCBE1 is a secreted extracellular matrix protein expressed by epicardial cells that is required for the formation of the primitive coronary plexus. However, the role of CCBE1 during epicardial development was still unknown. Here, using a Ccbe1 knockout (KO) mouse model, we observed that loss of CCBE1 leads to congenital heart defects including thinner and hyper-trabeculated ventricular myocardium. In addition, Ccbe1 mutant hearts displayed reduced proliferation of cardiomyocyte and epicardial cells. Epicardial outgrowth culture assay to assess epicardial-derived cells (EPDC) migration showed reduced invasion of the collagen gel by EPDCs in Ccbe1 KO epicardial explants. Ccbe1 KO hearts also displayed fewer nonmyocyte/nonendothelial cells intramyocardially with a reduced proliferation rate. Additionally, RNA-seq data and experimental validation by qRT-PCR showed a marked deregulation of EMT-related genes in developing Ccbe1 mutant hearts. Together, these findings indicate that the myocardium defects in Ccbe1 KO mice arise from disruption of epicardial development and function.     

Regulation of Epicardial Cell Fate during Cardiac Development and Disease: An Overview

The epicardium is the outermost cell layer in the vertebrate heart that originates during development from mesothelial precursors located in the proepicardium and septum transversum. The epicardial layer plays a key role during cardiogenesis since a subset of epicardial-derived cells (EPDCs) undergo an epithelial–mesenchymal transition (EMT); migrate into the myocardium; and differentiate into distinct cell types, such as coronary vascular smooth muscle cells, cardiac fibroblasts, endothelial cells, and presumably a subpopulation of cardiomyocytes, thus contributing to complete heart formation. Furthermore, the epicardium is a source of paracrine factors that support cardiac growth at the last stages of cardiogenesis. Although several lineage trace studies have provided some evidence about epicardial cell fate determination, the molecular mechanisms underlying epicardial cell heterogeneity remain not fully understood. Interestingly, seminal works during the last decade have pointed out that the adult epicardium is reactivated after heart damage, re-expressing some embryonic genes and contributing to cardiac remodeling. Therefore, the epicardium has been proposed as a potential target in the treatment of cardiovascular disease. In this review, we summarize the previous knowledge regarding the regulation of epicardial cell contribution during development and the control of epicardial reactivation in cardiac repair after damage.     

In Vivo and In Vitro Cartilage Differentiation from Embryonic Epicardial Progenitor Cells

The presence of cartilage tissue in the embryonic and adult hearts of different vertebrate species is a well-recorded fact. However, while the embryonic neural crest has been historically considered as the main source of cardiac cartilage, recently reported results on the wide connective potential of epicardial lineage cells suggest they could also differentiate into chondrocytes. In this work, we describe the formation of cardiac cartilage clusters from proepicardial cells, both in vivo and in vitro. Our findings report, for the first time, cartilage formation from epicardial progenitor cells, and strongly support the concept of proepicardial cells as multipotent connective progenitors. These results are relevant to our understanding of cardiac cell complexity and the responses of cardiac connective tissues to pathologic stimuli.     

基于MRI图像的左心室分割方法研究现状与发展

心脏的磁共振图像（Magnetic Resonance Imaging，MRI）对比度强、分辨率高、能够准确描述心脏的解剖功能，因此被认为是准确评估心脏挤压能力的金标准。在心脏的核磁共振图像上准确分割出左心室是准确评估左心室功能的前提。阐述了磁共振图像左心室分割的基本特点和难点，针对现阶段具有代表性的4种MRI左心室分割算法的基本原理、分割效果及时间效率等特点进行了归纳与总结。总结分析了MRI左心室分割领域面临的一些问题和发展方向。     

Automatically detecting endocardium and epicardium of left ventricle by digital image processing for 2‐D echocardiography

Abstract

Detecting the endocardium and epicardium of the left ventricle is important for further quantitative analysis of cardiac functions and three‐dimensional reconstruction. However, to detect the boundaries automatically is difficult. In this paper, we proposed a novel, single frame method, based on binary morphology and region growing techniques, to detect the boundaries of the left ventricle automatically except an A.O.I. (Area Of Interest) adjustment operation must be performed during the process of getting image data from video tape and storing it onto the disk. The method is suited for boundary detection required in three‐dimensional reconstruction for the observation of congenital abnormalities of ventricular septum or for generating cross sectional images of arbitrary orientation. Moreover, the method may be improved by further processings for accurate endocardial and epicardial boundary estimation for quantitative cardiac function evaluations.