Systems and methods for analyzing electrical activity in smooth muscle of the gastrointestinal tract of a patient are disclosed. The systems includes an electromyographic-sensing patch adapted for placement on the skin of the abdomen of the patient. The patch has at least one bipolar electrode pair, or a multitude arranged in an array, and is enabled for communication of a signal indicative of a sensed electromyographic signal. The methods include artifact removal, data normalization, and peak detection methods on data derived from the sensed electromyographic signal and employ, at least for example, machine learning methods.
