901-38 New Image Processing, Segmentation, Extraction and Display Algorithms Allow Easier and More Versatile Dynamic Volume-rendered Three-dimensional Echocardiographic Examination: Application in Experimental and Clinical Studies

The segmentation and display algorithms available for our previous 3-D echo (3DE) studies were limited, and cutting planes had to be used for displaying specific regions (objects) within the heart. In this study, we explored the potential of newer 3D segmentation, extraction and display algorithms in performing volume-rendered 3DE in 10 dog studies and in 30 clinical studies. Sequential 2D image data for 3DE were collected using tomographic linear slicing, fan-like scanning or rotational imaging modalities with the use of computer-controlled motors, gated to ECG and respiration. The 3DE data matrix obtained from 2DE images of cardiac cycles acquired at 1mm or 1–2degrees (over 180 degree span) underwent further processing. Besides developing dynamic 3DE image projections as done in the past, the use of new filtering techniques (low pass, Sigma, anisotropic diffusion), and the new segmentation and extraction software (special contouring, morphing, and Boolean process) enabled us to perform the following: (1) We were able to extract any desired region or part of the heart (original data set), and display it in a dynamic 3D mode, (2) Color-encode the regions of pathology, (3) Change opacification grade of structures such as myocardium (from solid to transparent), (4) Visualize in 3D myocardial regions of selective contrast enhancement following contrast administration, (5) View chambers alone as cavity casts if desired, (6) Display one or more extracted labels (regions) together, side-by-side. or separately, (7) Quantify the volume of any extracted region, and (8) Perform measurements directly on 3D images (which was not possible before). These capabilities allowed us to better delineate and quantify pathology of various types in the clinical studies and aided in the study of myocardial perfusion in experimental studies. Thus the newly developed image processing algorithms increase the clinical potential of 3-D echocardiography