The proposed image processing algorithm effectively automates the extraction of endocardial borders from 2D echocardiograms, overcoming common image quality issues.
Answers are grounded in this paper and cite their sources.
May aid automated echocardiogram analysis; leaves open prospective validation before clinical adoption.
An effective image processing algorithm for extracting the endocardial border from a series of 2-dimensional echocardiograms (2D echo) is proposed. The main steps involved in this algorithm are coordinate transformation, regional co-occurrence matrix thresholding, search region limiting, border detecting, and spatial and temporal smoothing. Using these steps, some major difficulties related to 2D echo image processing including low signal-to-noise ratio, echo drop out, and gray level intensity variability have been dramatically reduced. All endocardial borders in an entire cardiac cycle are computed. This is done automatically on the basis of observer-drawn outlines at end systole and end diastole. These outlines facilitate the quantitative analysis by defining a range in which both boundary dimension and shape changes should occur. Image processing is restricted to this range. The validation studies have been perfonned on test balloons where both image quality has been changed and complex motion abnormalities have been experimentally produced. Good results have also been obtained in patient studies.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (1984) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: