Principal components: a new approach to the analysis of gastric emptying

Digitized activity-time curves from gastric emptying studies with a gamma camera can best be represented by the principal components of the curves. These principal components can be used to construct a model for each curve which is a much better fit to the original data, as measured by the statistical criterion of goodness of fit, than a single exponential model. The emptying curves of patients from the same clinical population exhibit similar principal components which distinguish them from other clinical groups. So that, for example, the control patient group can be distinguished from the group of patients with duodenal ulcer. The implication of this separation between groups is that differences in the pattern of gastric emptying, which are detected by the principal component analysis, have physiological causes. Further, it may be possible to isolate the nature of these causes by inducing changes in pattern with known stimuli. (author)