Theoretical and computational methods for the noninvasive detection of gastric electrical source coupling

The ability to study the pathology of the stomach noninvasively from magnetic field measurements is important due to the significant practical advantages offered by noninvasive methods over other techniques of investigation. The inverse biomagnetic problem can play a central role in this process due to the information that inverse solutions can yield concerning the characteristics of the gastric electrical activity (GEA). To analyze gastrointestinal (GI) magnetic fields noninvasively, we have developed a computer implementation of a least-squares minimization algorithm that obtains numerical solutions to the biomagnetic inverse problem for the stomach. In this paper, we show how electric current propagation and the mechanical coupling of gastric smooth muscle cells during electrical control activity can be studied using such solutions. To validate our model, two types of numerical simulations of the GEA were developed and successfully used to demonstrate the ability of our computer algorithm to detect and accurately analyze these two phenomena. We also describe our analysis of experimental, noninvasively acquired gastric biomagnetic data as well as the information of interest that our numerical method can yield in clinical studies. Most importantly, we present experimental evidence that the coupling of gastric electrical sources can be observed using noninvasive techniques of measurement, in our case with the use of a superconducting quantum interference device magnetometer. We discuss the relevance and implications of our achievement to the future of GI research.

Figure 1

Spatial distribution of 190 simulated GI electric current dipoles used for inverse procedure testing (units of m). Gastric dipoles are visible at the top of the image, in an arrangement that matches the anatomy of the stomach. Duodenum dipoles are located immediately below, connecting the stomach to the small bowel that covers most of the volume used in the simulation.

Figure 2

Simulated (rhombi) and a few inverse (squares) dipoles identified by the least-squares inverse problem algorithm for the anatomical region of the stomach.

Figure 3

Same as Fig. 2, but for the duodenum.

Figure 4

Same as Fig. 2, but for the ileum.