Quantitative Evaluation of Regional Function

Generating parametric images for evaluation of regional left ventricular function implies performing a data reduction and consequently isolating specific functional features from a cardiac cycle. Whether the latter has been obtained by a gated equilibrium blood pool image or by contrast cineangio-

graphy is inconsequential. However, the following discussion will be based on gated equilibrium radionuclide angiography. The data reduction generated by parametric imaging ultimately compresses all the information found in the cardiac cycle sequence — pertaining to a certain functional component — into one or two images. This makes for a much easier interpretation and singles out the respective functional component. Such images remove a great part of subjectivity of interpreting a cinematic display and enables quick comparisons at subsequent patient follow-up. It is also important to remember that for parametric images, a color display is an absolute necessity and not a luxury [32,33]. The most commonly used parametric images are the simple conventional function images such as stroke volume image, the paradoxical image, the regional ejection fraction image, and more complex parametric images such as the phase analysis and the time rate factor analysis images.

Stroke Volume Image (SV)

This image is simply obtained by subtracting pixel-by-pixel end-systolic from end-diastolic counts (ED — ES). The resulting image is a regional map of the net volume changes between diastole and systole in each pixel [33,34]. The display usually includes only positive values (negative values are set to 0); only those structures that contain more blood in end-diastole than in end-systole will be presented. Therefore, under normal circumstances only the ventricles will be seen. The result will not change whether the operation is performed before or after background subtraction.

Paradox Image

Originally developed for dynamic radionuclide studies, this image is obtained by doing inverse pixel-by-pixel subtraction, which means subtracting the end-diastolic from the end-systolic counts (ES — ED). Similarly to the SV image, the pixels for which subtraction generates negative values are set to 0 and the results are not dependent on background subtraction. In normals, only the atria and great vessels area will be seen because ventricles will generate negative values. The purpose of this image is to detect ventricular paradoxical motion. This is possible because dyskinetic regions will have increased local blood volume during systole.

Regional Ejection Fraction Image (REF)

This image is obtained by dividing pixel-by-pixel the SV image by the background corrected end-diastolic frame:

In this case, the use of background subtraction is mandatory. The regional ejection fraction image is proportional to local ejection fraction values. In normals, its pattern is represented by a number of crescent-shaped areas of progressively lower value going from the free borders and apex to the septum and ventricular base. At the periphery, there are values of 100% or close to it because all of the blood volume present in end-diastole has been completely expelled by the time end-systole is reached. In regional wall motion abnormalities, the normal REF pattern is broken. If dyskinetic segments are present, the resulting negative values are set to 0. There is a significant potential for errors in the computation of REF image because of its dependence on adequate background subtraction, leading to significant overestimation or underestimation. In addition, for practical reasons ( presence of surrounding noise) the REF image requires masking, which confines the results to the ventricular region of interest.

0 0

Post a comment