(688c) Compartmental Modeling of the Gastrointestinal (GI) Tract: Model Development and Validation in Predicting Gastric Emptying of Liquids

Introduction

The stomach in the Gastrointestinal (GI) tract is responsible for the storage, breakdown, and transport of food to the duodenum for digestion. Previously, finite element methods and computational fluid dynamics (CFD) techniques have been reported that model this complex function involving interacting neural cell (Interstitial Cells of Cajal (ICC)) and smooth muscle cell (SMC) activity, deformable wall boundaries and the resulting pulsating flow. Invariably, these models are computationally expensive. In this work, we propose a compartmental modeling approach that is computationally cheaper and suitable for designing model based closed-loop controllers as neurostimulation-based therapy for treating GI disorders.

Methods

The stomach is divided into seven compartments: Fundus, Upper Corpus (pacemaker), Lower Corpus, Proximal/Middle/Terminal Antrum and Pyloric sphincter. Each compartment is modeled with a set of ordinary differential equations with respect to time. The stomach has three stimulation points 1) Fundus 2) Pacemaker Region and 3) Pyloric Sphincter. A simplified set of equations captures the electrochemo-mechanical coupling which converts membrane voltage changes caused by the stimulation in the ICC to smooth muscle contractions. During these contractions, the non-linear viscoelastic material of the stomach’s muscle wall is taken into consideration. The stimulus at the Fundus leads to long or tonic contractions. These contractions provide a pressure gradient which helps in the gastric emptying of the stomach. The stimulus at the Pacemaker Region results in slow waves in the SMC which results in frequent and periodic muscle contractions that give rise to peristaltic motion which propagates from the Pacemaker Region to the Terminal Antrum of the stomach. During the peristaltic motion, the contraction is stronger when the peristaltic wave propagates from the Corpus to the Proximal Antrum, so this leads to a peristaltic pump which helps in the gastric emptying of the fluid. The velocity profile for peristalsis is modeled using a transient Shapiro’s model. The stomach’s muscle walls also contain folds or rugae, so a transient Bernoulli’s equation is used to model the fluid flow by the pressure gradient from the Fundus and also to capture the effect of friction by rugae on the fluid velocity. Finally, the stimulus at the Pyloric Sphincter initiates a retrograde flow and controls the gastric fluid emptying since the time taken for gastric emptying depends on the volume of gastric contents in the stomach, the nutrient content, and the viscosity. When chyme from the pyloric sphincter enters the duodenum, a feedback signal may be sent to the pyloric sphincter to inhibit the chyme flow to the duodenum to prevent overloading if the liquid is high in calorie content. This feedback signal is also incorporated in the model. The resulting model is solved using differential equation solvers in MATLAB.

Results/Conclusions

The model computes membrane voltage from ICC, SMC, intracellular calcium concentration, stress and stretch values during the SMC contraction, fluid velocity and volume of the stomach during gastric emptying. These results are then compared and validated against other simulations: (1) the CFD study of liquid gastric emptying (Li, Gin (2020)); (2) FEM viscoelastic SMC model (Panda, Buist (2021)); (3) experimental data from Vella, Camilleri (2017) on mechanical and hormonal response to nutrient ingestion.