Melton H E, Collins S M, Skorton D J
Ultrason Imaging. 1983 Oct;5(4):300-7. doi: 10.1177/016173468300500402.
Previous approaches to automatic border recognition in two-dimensional echocardiography have utilized off-line computer image analysis. In this paper, we describe a hardware technique for real-time endocardial edge detection during the echocardiographic examination. This technique is based on a new method of line-by-line compensation for ultrasound attenuation. The detected edges and composite edge images can be displayed alone or superimposed upon the usual echocardiographic display. To test the accuracy of our real-time technique, we obtained echocardiograms of excised hearts and compared real-time edges with edges manually traced by an experienced observer. Left ventricular cavity areas and endocardial perimeters derived from real-time edges correlated well with those derived from observer-defined edges.
以往二维超声心动图自动边界识别方法采用的是离线计算机图像分析。在本文中，我们描述了一种在超声心动图检查期间进行实时心内膜边缘检测的硬件技术。该技术基于一种新的逐行超声衰减补偿方法。检测到的边缘和合成边缘图像可以单独显示，也可以叠加在常规超声心动图显示上。为了测试我们实时技术的准确性，我们获取了离体心脏的超声心动图，并将实时边缘与经验丰富的观察者手动描绘的边缘进行比较。从实时边缘得出的左心室腔面积和心内膜周长与从观察者定义的边缘得出的结果相关性良好。
