Separation of gastric electrical control activity from simultaneous MGG/EGG recordings using independent component analysis

Spatiotemporal parameters of gastric electrical control activity such as its amplitude, direction and prop- agation velocity are physiological parameters of distinctive clinical interest due to their potential use for differentiating between the healthy and diseased states of the human stomach. Whereas their time evolution is relatively well behaved in the case of healthy subjects, significant deviations from normal have been observed in patients suffering from a number of gastric diseases such as gastroparesis and gastropathy. For this reason, monitoring ECA parameters noninvasively may offer a useful test for the presence of such diseases whose diagnosis remains problematic. Here, we describe a method for computing ECA direction and orientation from simultaneous, noninvasive magnetogastrographic (MGG) and electrogastrographic (EGG) recordings. We demonstrate how independent component anal- ysis and standard frequency analysis methods can be used to predict the locations and orientations of gastric current dipoles from MGG/EGG data. We compare our MGG-based dipole parameters to analogous ones obtained from simultaneous EGG recordings within the experimental framework of a human model. We find that magnetic recordings are superior in their ability to portray the underlying physiology of the stomach. I. INTRODUCTION The study of gastric motility is of clinical interest due to the relationship between gastrointestinal (GI) disorders and the characteristics of gastric electrical control activity (ECA). ECA is a slow, sinusoidal wave with a frequency of 3 cycles per minute (cpm) originating in the antral gastric region and propagated along smooth muscle cells. Abnormal ECA has been associated with many GI disorders, such as gastroparesis, diabetic gastropathy, and gastric myoelectrical dysrhythmia (12). Two methods for characterizing ECA are electrogastrography (EGG) and magnetogastrography (MGG). EGG involves the placement of cutaneous electrodes on the abdomen to record the bioelectric fields due to ECA at the body surface. The reliability of this procedure has been questioned due to the high dependence of electrical record- ings upon tissue conductivity, the thickness of the abdominal wall, and the variable propagation velocity of ECA (10). MGG involves the positioning of a magnetometer in close proximity to the abdomen to detect the biomagnetic field generated by the electric current of ECA. The measurement of biomagnetic fields is advantageous because they are more strongly dependent on tissue permeability, which is nearly equivalent to that of free space. In this study, we describe our application of principal and independent component analysis (PCA and ICA, respec- tively) for measuring ECA parameters from simultaneous, noninvasive MGG/EGG recordings. Our approach makes use of ICA to isolate the gastric signal, whereafter gastric dipole locations and orientations can be determined from ICA field maps. We compare our results based on magnetic data to analogous ones obtained using simultaneous electrode recordings.