Signal-processing techniques for cochlear implants

Cochlear implants have been successful in restoring partial hearing to profoundly deaf people. The success of cochlear implants can be attributed to the combined efforts of scientists from various disciplines, including bioengineering, physiology, otolaryngology, speech science, and signal processing. Each of these disciplines contributed to various aspects of the cochlear implant design. Signal processing, in particular, played an important role in the development of different techniques for deriving electrical stimuli from the speech signal. The purpose of this article is to present a review of various signal-processing techniques that have been used for cochlear prosthesis over the past 25 years     

Electroacoustic evaluation of assistive hearing devices

The tests described here are the first steps in consumer-based hearing aid measures. Maximum length sequence broadband frequency response measurements provide a fast method to determine the accuracy of a hearing aid "fit". The aided output sound pressure level (SPL) graph adds valuable information to determining the fit: it clearly shows how the hearing aid compensates for a hearing loss; and relates the output SPL to the user's threshold, loudness discomfort level (LDL), and the desired sensation level long-term average spectrum of speech target for 70 dB input SPL. At a basic level, the question of user comfort is addressed by the SPL graph and biased frequency response testing. Exceeding a user's LDL may result in discomfort. Signal biased testing can be used to determine if environmental sounds are likely to cause unusual hearing aid operation, which may lead to discomfort. The type of processing used in a hearing aid can also be determined via pure-tone biased testing. The final questions of speech quality/intelligibility in various environments and the effect of hearing aid processing on listeners' performance and perception of sound quality are being addressed by the authors' current and future work. Objective measures of speech quality based on coherence have been used to predict subjective speech quality for normal listeners. There have also been a number of other objective measures of speech quality that have given good results over a wide range of distortions. The next step will be to apply these measures to hearing aid testing. This will allow one to make objective measurements of distortion and determine its influence on user comfort and judgment of speech quality     

Reducing correlated noise in digital hearing aids

The intelligibility of speech in communication systems is generally reduced by interfering noise. This interference, which can take the form of environmental noise, reverberation, competing speech, or electronic channel noise, reduces intelligibility by masking the signal of interest. The reduction in intelligibility is particularly troublesome for listeners with hearing impairments, who have greater difficulty understanding speech in the presence of noise than do normal-hearing listeners. Numerous digital signal processing (DSP)-based speech enhancement systems have been proposed to improve intelligibility in the presence of noise. Several of these systems have difficulty distinguishing between noise and consonants, and consequently attenuate both. Other methods, which use imprecise estimates of the noise, create audible artifacts that further mask consonants. The objective of the present study is to develop a new noise-reduction method that can reduce additive noise without impairing intelligibility. The new method could be used to improve intelligibility in a wide variety of applications, with special attention given to digital hearing aids and other portable communication systems (e.g., cellular telephones). Here, the authors present a new wavelet-based method for reducing correlated noise in noisy speech signals. The authors provide background information on the intelligibility problem and on previous attempts to address it. A theoretical framework is then proposed for reduction of correlated noise, along with some preliminary experimental results     

A phone-assistive device based on Bluetooth technology for cochlear implant users

Hearing-impaired people, and particularly hearing-aid and cochlear-implant users, often have difficulty communicating over the telephone. The intelligibility of telephone speech is considerably lower than the intelligibility of face-to-face speech. This is partly because of lack of visual cues, limited telephone bandwidth, and background noise. In addition, cellphones may cause interference with the hearing aid or cochlear implant. To address these problems that hearing-impaired people experience with telephones, this paper proposes a wireless phone adapter that can be used to route the audio signal directly to the hearing aid or cochlear implant processor. This adapter is based on Bluetooth technology. The favorable features of this new wireless technology make the adapter superior to traditional assistive listening devices. A hardware prototype was built and software programs were written to implement the headset profile in the Bluetooth specification. Three cochlear implant users were tested with the proposed phone-adapter and reported good speech quality.     

基于声门激励的语音语速、音量调整方法

在小词汇量孤立词语音识别的研究中,录入的语音指令语速过快(慢)或音量过大(小)都会影响语音识别系统对于这些指令的识别率.为此,需要对这些语音指令进行调整.本文提出一种基于线性预测分析技术的语音语速和音量调整方法,该方法由已知语音信号反解出声门激励信号,对反解出的声门激励信号进行语速和音量的调整,将调整后的声门激励信号再次激励声道系统输出语音信号,得到调整后的语音信号.仿真研究结果表明,该方法具有实际可行性并且可以显著改善语音识别系统对于语速过快(慢)、音量过大(小)指令的识别率.     

Neck and Face Surface Electromyography for Prosthetic Voice Control After Total Laryngectomy

The electrolarynx (EL) is a common rehabilitative speech aid for individuals who have undergone total laryngectomy, but they typically lack pitch control and require the exclusive use of one hand. The viability of using neck and face surface electromyography (sEMG) to control the onset, offset, and pitch of an EMG-controlled EL (EMG-EL) was studied. Eight individuals who had undergone total laryngectomy produced serial and running speech using a typical handheld EL and the EMG-EL while attending to real-time visual sEMG biofeedback. Running speech tokens produced with the EMG-EL were examined for naturalness by 10 listeners relative to those produced with a typical EL using a visual analog scale. Serial speech performance was assessed as the percentage of words that were fully voiced and pauses that were successfully produced. Results of the visual analog scale assessment indicated that individuals were able to use the EMG-EL without training to produce running speech perceived as natural as that produced with a typical handheld EL. All participants were able to produce running and serial speech with the EMG-EL controlled by sEMG from multiple recording locations, with the superior ventral neck or submental surface locations providing at least one of the two best control locations.     

Multiband sound source localization algorithm for directional enhancement in hearing aids

In order to improve speech comprehension for hearing‐impaired patients under complex scenarios, directional enhancement algorithm based on the sound location will be one of the future research directions. A multiband sound localization algorithm based on auditory bionics is proposed in this paper. First, according to the property of cochlear frequency‐division and hearing‐masking characteristic, the sound is divided into multiband signals. Then, the interaural time differences (ITDs) of the sensitive frequency bands are taken as the localization cues. To improve the location precision, the real ITD is extracted based on the Haas effect. Finally, the HRTF (head‐related transfer function) model is used to convert the ITD to the sound position. To improve the robustness of the algorithm, a multiband weighted strategy is adopted. In addition, a parameter adjustment method is proposed to effectively fit the hearing aid. Simulation and scenario experiments show that the algorithm has strong anti‐interference and high location accuracy. Furthermore, it is designed based on the existing algorithms for hearing aids and has low computation, so it is very suitable for hearing aids and other low‐power real‐time devices. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于动态信噪比估计的小波阈值去噪新算法

由于语音信号的非平稳性,传统的小波阈值去噪算法虽然能够衰减一部分语音信号中的噪声,但这些算法会不可避免地造成有用语音信号尤其是清音部分的损失,以至于去噪后的语音听觉质量下降,达不到很高的信噪比。针对这一问题,本文介绍了一种基于小波包变换和动态信噪比估计的阈值方法。这种方法可以有效地保护有用信号不被去除,但对于带噪信号在不同尺度下的信噪比,它的自适应性不强,因此本文对这种算法做了一些改进,使之可以达到更好的去噪效果。     

Low-Power Circuits for the Bidirectional Wireless Monitoring System of the Orthopedic Implants

This paper proposes an architecture of the wireless monitoring system for the real-time monitoring of the orthopedic implants, which monitors the implant duty cycle, detects abnormal asymmetry, high amounts of force, and other conditions of the orthopedic implants. Data for diagnosis are communicated wirelessly by the radio-frequency (RF) signal between the embedded chip and the remote circuit. In different working modes, the system can be powered by the RF signal or stiff lead zirconate–titanate (PZT) ceramics which are able to convert mechanical energy inside the orthopedic implant into electrical energy. The power circuits with a variable ratio switched-capacitor (SC) dc–dc converter have been taped out with 0.35-$mu$m complementary metal–oxide semiconductor (CMOS) technology. The test results show that the SC converter can transfer the input voltage that ranges from 5 V to 14 V from the PZT ceramics into the voltage ranging from 2 V to 2.5 V which will be dealt with by a low drop-out circuit in the future work. The total efficiency of the SC converter is from 28% to 42% at full-time working mode. The analog-to-digital converter (ADC) circuits have been fabricated in a 0.18-$mu$m 1P6M CMOS process. The test results show that the ADC chip consumes only 12.5 $mu$W in working mode and 150 nW in the sleep mode. The circuits, including RF circuits, ADC, and the microcontrol unit, have been implemented in a 0.18-$mu$ m CMOS process. Future work includes some clinical experiments test in the application where PZT elements are used for power generation in total knee-replacement implants.      

Speech coding with wavelets

Speech is a very basic way for humans to convey information to one another. From a communications standpoint, information can be sent efficiently by sending it as just text. However, with a bandwidth of only 4 kHz, speech can convey information with the emotion of a human voice. People want to be able to hear someone's voice from anywhere in the world-as if the person was in the same room. As a result, new, efficient speech coding techniques have impacted areas such as cellular telephony, mobile radios and voice mail systems     

A joint echo cancellation algorithm for quick suppression of howls in hearing aids

When the echo path of a hearing aid suddenly changes, howls easily occur. To quickly suppress the howls, a joint echo cancellation (JEC ) algorithm, which combines the variable step normalized least mean square (VNLMS ) algorithm with the notch filter algorithm, is proposed. According to whether the hearing aid howls or not, different strategies are used. First, when there are no howls, the echo signal is estimated using VNLMS and the step factor is computed according to three types of filter states, which are defined based on the normalized distance between the short‐term average and the long‐term average of the filter coefficients. Then, different step factors are used for different states. Second, when there are howls, the update of VNLMS is frozen to stabilize the howl frequency. To improve the detection accuracy, a howling detection algorithm based on the zoom‐fast Fourier transformation (ZoomFFT ) is proposed. The ZoomFFT algorithm can analyze the spectrum of a narrowband signal in a specified high sampling frequency. Then, the notch filters based on the estimated howl frequencies are dynamically generated to restrain the howls. Finally, when the howls are suppressed, VNLMS is reactivated. Compared to other echo cancellation algorithms, the proposed algorithm can quickly suppress the howls, and JEC has the best comprehensive performance. Furthermore, the quality of the processed speech is high, and the operation time is short. Thus, the proposed algorithm is suitable for low‐power‐consumption and small‐volume products such as hearing aids. © 2017 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Robotics courses for children as a motivation tool: the Chilean experience

Encouraging children's interest in science and technology, as well as increasing their technological literacy, may be regarded as one of the educational paradigms of this century. The authors of this paper, affiliated with the Department of Electrical Engineering at the University of Chile, Santiago, have designed schemes to contribute to both goals and set guidelines for curricular and extracurricular school activities related to technology. In this context, this paper reviews their experience concerning practical robotics courses for children developed since 2000. More than 700 children and 90 school teachers have already attended these robotics courses, and the model is now being implemented in several schools and institutions in Chile. The robotics courses evolved to their present form from ideas developed during the late 1990s, mostly in the United States. Some preliminary assessment data is presented to support this approach. Current projects are also outlined. It is believed that the authors' experience might be of interest to engineering schools elsewhere.     

A Microcomputer-Based Reading Aid for Blind Students

In schools and colleges for the blind, Braille is the popular medium of accessing textual educational information. However, in situations when the text does not contain diagrams, graphs, or maps, converting the text to speech would increase the reading speed of the blind student considerably. In this paper, the techniques used in designing and evaluating the performance of a microcomputer-based system which automatically reads out textual information in synthetic speech is described.     

Solutions to electromagnetic interference problems between cochlear implants and GSM phones.

For persons using cochlear implants, electromagnetic compatibility (EMC) problems may sometimes be an obstacle to using digital cellular telephones. This study aimed at exploring the benefit of three new assistive listening device prototypes that eliminate or diminish EMC problems. Ten experienced cochlear implant users listened in quiet to running speech samples and a sentence test on a landline phone, a digital cellular phone with and without three prototypes. The subjects' performance was assessed using a sentence test, a subjective visual analog scale, and by ranking the best and the poorest listening condition. Compared to the other test conditions, listening to a digital cellular phone alone revealed, on average, the poorest sentence recognition scores (29%) and the poorest results in four different subjective judgments (the amount of disturbances, the clarity of the message, the quality of the sound, overall judgment) with all three implant systems tested. The prototypes generally helped the implantees to recognize speech better on the cellular telephone (by 10-21 percent units, on average). Use of assistive listening devices and further development of EMC of both cochlear implant systems and digital cellular phones needs to take place to enable smooth use of digital cellular phones for all implantees.     

一种基于小波变换的语音信号去噪净化方法

语音信号在传输之前快速有效的去噪净化,对于3G通信具有重要的现实意义。由于语音信号的非平稳性,传统的小波阈值去噪算法虽然能够去除一部分语音信号中的噪声,但造成有用语音信号尤其是清音部分的损失,导致去噪后的语音听觉质量下降,达不到很高的信噪比。针对这一问题,本文在分析小波变换原理和传统软硬阈值法的基础上,建立了一种改进的阈值语音信号净化方法。仿真实验结果表明,本方法能有效去除信号中的噪声和较好保留语音细节,达到更佳的语音净化效果。     

基于DSP的语音录放和数字回声的实现

数字信号处理是现代数字语音通信的核心技术之一,DSP技术的应用为实现语音信号的采集、处理和播放奠定了基础.设计基于DSP的语音信号处理系统,运用TMS320VC5416处理器和AIC23语音芯片构建出硬件平台.分析DSP处理器和Codec芯片性能特点,配置软件控制接口,实现数字音频数据通信传输.采用C语言和汇编语言混合方式编写系统主程序和模块子程序,调试实验语音处理系统的信号采集、数字回声和播放,实现语音处理系统设计功能.     

Emerging Technologies

One of the most promising medical device therapies is neurostimulation. Advances in implantable device technologies make it now possible for patients suffering from a wide array of debilitating neurological conditions to receive adequate treatment. Implantable neurostimulation systems include cochlear implants, cortical stimulators, deep brain stimulators (DBSs), gastric nerve stimulators, spinal cord stimulators (SCSs), and vagus nerve stimulators (VNSs). Neurostimulation currently treats several debilitating conditions, including major treatmentresistant depression, epilepsy, gastroparesis, hearing loss, incontinence, chronic, untreatable pain, Parkinson disease, essential tremor, and dystonia [1]?[3].     

MRI and implanted medical devices: basic interactions with an emphasis on heating

There are three principal magnetic fields in magnetic resonance imaging (MRI) that may interact with medical implants. The static field will induce force and torque on ferromagnetic objects. The pulsed gradients are of audio frequency and the implant may concentrate the induced currents, with a potential for nerve stimulation or electrical inference. The currents induced in the body by the radio frequency (RF) field may also be concentrated by an implant, resulting in potentially dangerous heating of surrounding tissues. This paper presents basic information about MRI interactions with implants with an emphasis on RF-induced heating of leads used for deep brain stimulation (DBS). The temperature rise at the electrodes was measured in vitro as a function of the overall length of a DBS lead at an RF frequency of 64 MHz. The maximal temperature rise occurred for an overall length of 41 cm. The method of moments was used to calculate the current induced in the lead. From the induced currents, the RF power deposition near the electrodes was calculated and the heat equation was used to model the temperature rise. The calculated temperature rises as a function of lead length were in good agreement with the measured values.     

基于Bootstrap方法的独立分量分析可靠性研究

语音信号盲分离受到随机采样的数据所影响,产生统计可靠性问题,从而影响分离的效果.Bootstrap方法是以原始数据为基础的抽样统计推断法,可方便的应用于实际的数据处理之中,分析ICA的统计可靠性,但传统Bootstrap方法本身的计算特性限制了自助样本的生成范围,从而使得自助概率分布产生了一定的偏离,使之无法渐近于真实情形.本文通过对Bootstrap样本生成范围的拓展,改进了独立分量分析在语音信号应用的可靠性算法,获得了更加精确的可靠性参数.