Study on reconstruction of rectal sensation based on wavelet packet analysis and SVM strategy

To control anal incontinence, we have developed an artificial anal sphincter system with sensor feedback. The artificial anal sphincter system is a novel hydraulic-electric muscle which mainly comprises an artificial anal sphincter, a wireless power supply subsystem, and a rectal sensation reconstruction subsystem. To investigate the features of the patients' rectal sensation, we have developed an in vitro experimental platform of artificial anal sphincter. In vitro experiments have been performed, and demonstrate that the traditional threshold method is not suitable for predicting the time for defecation. The traditional threshold method only uses single-dimensional pressure time series which may contain a few interdependent components simultaneously. A wavelet packet analysis algorithm is employed to extract the feature vector of the rectal pressure signal, then the rectal sensation prediction model is constructed based on a support vector machine for defecation pattern recognition. The results show that the proposed method is an effective approach for the reconstruction of patients' rectal sensation.