Lag phase quantification for solid gastric emptying studies

This study compared the different calculation methods of the solid gastric emptying lag phase and evaluated the effect of the temporal sampling interval on the calculated value using the modified power exponential (MPE) method.

Twenty normal control subjects and 42 patients had anterior and posterior image acquisition on a multihead gamma camera, one frame per minute x 90. ROIs were selected for the stomach, gastric antrum and small bowel. Time-activity curves (TACs) were generated for anterior, posterior and geometric mean data. The lag phase was calculated using various methods such as transition point, starting index, first appearance of bowel activity (FABA), MPE and antral peak filling time. To determine the importance of the temporal sampling rate on the calculation of the lag phase by the MPE, intervals between 1 and 20 min were analyzed.

The transition point, starting index and FABA correlated extremely high (r = > or = 0.92) in normal control subjects and patients. Normal lag phase values were mean 22-24 +/- 10 min for transition point, starting index and FABA compared with 47 +/- 18 min for the MPE method (p < 0.0001). The MPE correlated poorly with the other method (r = 0.74). Antral peak filling time correlated poorly (r = 0.47) with transition point, starting index and FABA, but somewhat better with the MPE (r = 0.70). Comparing 15-min versus 1-min sampling intervals using the MPE, 35% of subjects had values that differed by > or = 7.5 min and 10% had values differing by > or = 15 min.

The lag phase calculated by the MPE correlated poorly with other methods, and its accuracy was limited by the rate of the temporal sampling. The transition point, starting index and FABA all highly correlated with each other; the latter is a particularly reliable physiological indicator and is easily quantified using a small-bowel TAC.