Comparative analysis of quantitative gastric emptying indices and power-exponential modelling of gastric emptying curves.

A comparative study was undertaken on the problems of choosing a reliable measure of gastric emptying (GE). The study was based on 218 GE measurements of a radiolabelled solid meal. Power-exponential fitting of GE curves was shown to provide an excellent mathematical approximation to experimental data - a coefficient of determination r2 greater than equal to 0.90 was obtained in the case of 96.3% GE curves. A range of indices was derived from the data, namely, the time to reach the maximum of a GE curve, the area under a GE curve, the mean transit time (and mean transit time during the ninety minutes of the examination), the emptying index, the GE half-time t 1/2 and the shape factor S. On the basis of between-subject variability (3.9% coefficient of variation) and reproducibility (coefficient of variation for paired examinations, CVp = 3.2%), the mean transit time calculated over the period of the GE study (MTT90) was found to be the best GE measure. The versatility of the power-exponential method was confirmed by the analysis of mathematical curves designed to mimic a range of real life emptying patterns, and the relationships between the derived parameters were also investigated.