Parametric images as a tool for quantitative normative evaluation

Features extracted from static or dynamic scintigraphic images can be used to produce parametric or functional images. Those images typically map a dynamic parameter (temporal encoding) or a quantitative measure of the tracer distribution. In this report, we consider the cases in which the nature of the extracted feature is well suited for a quantitative normative evaluation, ie, where the values obtained can be directly or indirectly compared with expected normal values. The major difference between this approach and the more common heuristic or synoptic interpretation of "images" lies in the underlying modeling: the model "predicts" a minimal washout rate, a match between ventilation and perfusion rates in the lungs, homogeneous contraction in the left ventricle, an expected angular distribution of thallium in the myocardium, or the absence of an additional kinetic feature. The quantitative aspect of the analysis is based in all cases on an approach that overcomes or is less sensitive to morphological or structural biological variability: in some cases the patient provides the normalizing data, as in ventilation-perfusion ratios. In other cases, the model predicts homogeneous results (as in phase analysis) or a range of normal physiological values (for xenon washout). Less commonly, the analysis requires a transformation of the data, as is the case in the analysis of myocardial perfusion, which follows a polar transformation of the image, and an analysis based on angular coordinates. In a normative approach, pathology is defined as a deviation (in this case a quantitative deviation) from the norm. The next step is to model specific abnormalities: early right-sided reappearance of the bolus in a right-to-left shunt, the appearance of "abnormal" kinetic factors, or a comparison with the expected distribution of bloodflow (or location of defects) in certain coronary lesions. However, in all cases, one should note that the analysis does not compromise the major feature of "imaging," that is to recognize regional, rather than exclusively global malfunction.