Endocardial border identification in two-dimensional echocardiographic images: review of methods

A basic goal in the analysis of echocardiographic images is to identify the locations of the endocardial borders. This is necessary in order to create three-dimensional reconstructions, and to derive precise quantitative parameters from the images. Endocardial borders can be identified manually; however, this is time-consuming, inconvenient, and liable to human subjectivity. For this reason there has been a great deal of interest in automatic border identification methods. In recent years numerous studies have been published, describing a wide variety of proposed border identification algorithms. In spite of the large number of published studies, however, the relative merits of the various methods remain unclear. In this paper the various analysis techniques are reviewed and discussed. The reasons for the lack of definitive conclusions regarding the utility of these methods are described. Finally, the basic methodological requirements necessary for empirical evaluation of border identification techniques are described.