中文摘要：

为重建肛门失禁患者的直肠感知功能,设计了一种新型的人工肛门括约肌;针对采集的直肠压力活动信号,利用小波分析在时域-频域提取其信息熵特征,并进行了信息熵特征提取实验;实验表明,信息熵特征对不同的信号具有较好的分类能力。利用提取的直肠压力信息熵特征,基于决策级信息融合方法构建直肠感知功能预测模型;最后通过仿真实验,验证了此预测模型在不同信噪比下都具有较高的预测准确率,能够重建患者直肠感知功能,预测患者排便时间。

英文摘要：

For reconstructing rectal sensation function of the patients with fecal incontinence,we have developed a novel artificial sphincter system.The information entropy feature of rectal pressure activity is extracted in time-frequency domain with wavelet analysis,and the extraction experiment is taken,which shows the information entropy feature can effectively classify different signals.Using the information entropy feature of rectal pressure activity,the predictive model of rectal sensation function is constructed based on decision-layer information fusion method.Finally,the simulation experiment is taken,and results show that the model has high predictive accuracy in different SNR,it can reconstruct patients＇ rectal sensation function and predict when need to defecate.