The use of gastric electrical signals for algorithm for automatic eating detection in dogs

Abstract  Food ingestion increases fundic impedance (FI) and reduces antral slow wave rate (SWR). Our aim was to determine if such changes can be integrated into an algorithm for automatic eating detection (AED). When incorporated in implantable medical devices, AED can time treatment to food intake without need for patient input. Four dogs were implanted with fundic and antral electrodes, connected to an implantable recording device. Changes in FI and SWR induced by fixed meals of different weights were determined, and were used to build an AED algorithm. Its performance was then tested on the same animals given an ad libitum access to food. The effects of gastric balloon distension and nitroglycerin on SWR and FI were also tested. Fixed meals reduced SWR in a weight-dependant manner, R2 = 0.936, P < 0.05 baseline compared to 50, 100, 200 and 400 g. Meals increased FI above baseline in a weight-dependant manner; R2 = 0.994, P < 0.05 baseline compared to 200 and 400 g. During ad libitum intake, the AED algorithm detected 86% of all meals ≥15 g. Gastric distension reduced SWR and increased FI. Nitroglycerin reduced SWR. AED, using changes in FI and gastric SWR is feasible. Changes in FI and SWR are induced primarily by the presence of food in the stomach.