Detection of gastric slow wave uncoupling from multi-channel electrogastrogram: validations and applications.

Current methodology of single channel electrogastrography is unable to detect coupling or uncoupling of gastric slow waves, which is crucial for gastric emptying. In this study, a new methodology, called cross-spectral analysis method, was established to compute the coupling percentage of multi-channel gastric slow waves recorded using serosal electrodes and electrogastrogram (EGG). Two experiments were performed to validate the method and demonstrate its applications in clinical research. In experiment 1, simultaneous recordings of gastric slow waves were made in five dogs from serosal electrodes and cutaneous electrodes. In experiment 2, four-channel fasting EGGs were made in 10 volunteers for 30 min during waking and 30 min during non-rapid eye movement (REM) sleep. The validation study (experiment 1) showed that the slow wave coupling calculated from the EGGs was correlated with that computed from the serosal recordings. The gastric slow wave coupling percentages detected from both serosal and cutaneous recordings were significantly impaired during vasopressin infusion (6.3 +/- 2.6 vs 62.4 +/- 6.3, P < 0.001 for serosal recordings; 6.7 +/- 3.0 vs 57.2 +/- 2.7, P < 0.001 for cutaneous recordings), and the coupling percentages respectively calculated from serosal and cutaneous recordings were significantly correlated during the baseline recording period (R = 0.922, P < 0.05) and vasopressin infusion period (R = 0.916, P < 0.05). In experiment 2, the gastric slow wave became less coupled when healthy volunteers fell asleep. The percentage of slow wave coupling calculated from the EGGs was 68.2 +/- 17.9% during waking but 41.9 +/- 20.8 during non-REM sleep (P < 0.05). The method developed in this study is reliable for the detection of slow wave uncoupling from multi-channel EGGs. Gastric slow wave coupling is impaired during vasopressin infusion and sleep. These data suggest that this method has potential applications in physiological and clinical studies.