Classifier of intestinal contractile activity degree based on internal electroenterogram recording

The study of the intestinal interdigestive motor migratory complex (IMMC) is relevant in gastroenterology because most of the gastrointestinal pathologies are reflected in anomalies of the IMMC. The aim of this work is to develop an automatic classifier to discriminate among the different intestinal contractile activity degrees (quiescence, irregular, and maximum contractile activity) that compound the IMMC from the internal recordings of electroenterogram. Spectral and statistical parameters estimated from the internal electroenterogram have been used as features to the classifiers based on Linear Discriminant Analysis (LDA) and linear Support Vector Machines (SVM). The accuracy obtained by the SVM classifier is slightly higher than that of the LDA classifier. An accuracy of around 91% was obtained for the binary SVM classifier (quiescence vs maximum activity) and around 74% for the multiclass one. The use of additional features, and non-linear SVM classifiers could yield better classification accuracy values. Nevertheless, preliminary results suggest that SVM classifiers could be a very helpful tool for automatic classification of intestinal contractile activity degrees and for the identification of the IMMC which could be used for diagnosing anomalies in the intestinal motor function.