摘要
电子耳蜗通过将声能转换成电能,直接刺激耳蜗内残余的听神经纤维来恢复重度聋患者的听觉。本文基于电子耳蜗听觉仿真模型理论,建立了电子耳蜗NofM算法的听觉仿真模型。并通过这一模型在不同条件下与传统的CIS算法进行了听觉仿真试验的对比。从试验的结果得出:具有频带选择特性的NofM算法较之传统的CIS算法具有更好的抗噪特性。最后对此类算法进一步的改进方法提出了建议。
Cochlear implant is a neural prosthetic device used to provide the sensation of sound to those who are profoundly deaf by delivering electrical stimulus to auditory nerve directly. The goal of this study is to build the acoustic model of N of M strategy used in speech processing of cochlear implant. And the acoustic model provided a mechanism by which to investigate the effects of various situation. The improvement of this strategy was observed for speech recognition in noise compared with the CIS strategy. We also proposed the implementation methods to improve N of M strategy and stimulate further investigation.
出处
《北京生物医学工程》
2006年第6期576-578,610,共4页
Beijing Biomedical Engineering
基金
清华大学基础研究基金(JC2002019)
粤港科技合作基金(GHS/053/04)资助
