Rectal sensation function rebuilding based on optimal wavelet packet and support vector machine. Issue 3 (1st May 2013)

• Record Type: Journal Article
• Authors: Jiang, Enyu; Zan, Peng; Zhu, Xiaojin; Liu, Jinding; Shao, Yong
• Is Part Of: IET science, measurement & technology. Volume 7:Issue 3(2013)
• Page Start: 139
• Page End: 144
• Publication Date: 2013-05-01
• Subjects: feedforward neural nets -- medical signal processing -- particle swarm optimisation -- support vector machines -- wavelet transforms; optimal wavelet packet -- support vector machine -- rectal sensation function rebuilding method -- OWP -- SVM -- rectal sensation loss -- anal incontinence -- human rectum characteristics -- high‐amplitude propagated contractions -- rectal contractions -- rectal pressure feature -- Davies‐Bouldin criterion -- normalised mean -- optimal bases coefficient energy -- feature vector -- rectal sensation prediction model -- parameter optimisation -- particle swarm optimisation -- prediction accuracy -- defecation intension -- feedforward neural networks -- kernel functions; Measurement -- Periodicals; Electrical engineering -- Periodicals; Electronics -- Periodicals; Nanotechnology -- Periodicals; Electromagnetism -- Periodicals; Medical instruments and apparatus -- Periodicals; 621.3
• DOI: 10.1049/iet-smt.2012.0015
• Languages: English
• ISSNs: 1751-8822

Abstract

Abstract : Rectal sensation function rebuilding method based on optimal wavelet packet (OWP) and support vector machine (SVM) is proposed for rectal sensation loss caused by anal incontinence. By analysing human rectum characteristics, high‐amplitude propagated contractions in rectal contractions are used to indicate an urge to defecate. Rectal pressure feature is extracted using OWP based on Davies‐Bouldin criterion. Normalised mean and energy of optimal bases coefficients are taken as feature vector. Rectal sensation prediction model is trained based on SVM whose parameters are optimised by particle swarm optimisation. Then the trained model is used to predict the urge to defecate. Meanwhile, contrast analysis of prediction accuracy of defecation intension using methods based on feedforward neural networks and SVM based on different kernel functions are given in this study. The prediction accuracy of the optimised SVM is compared with SVM using different kernel functions. Experiment results show that the proposed method is advantageous to rebuild patients' rectal sensation.