Chapter 6 Development of DPOAE Acquisition System for Hearing Screening 6.1 Introduction Evoked otoacoustic emission testing is one of the most commonly used method for hearing screening. Distortion Product Otoacoustic Emissions are generally accepted as a good tool for investigation of micro mechanism of the cochlea and hearing screening. DPOAEs are acoustic signals that can be detected in the ear canal of a person with normal OHC function, subsequent to stimulations of auditory system with a pair of pure tones at frequencies.f 1 and.f2. The resulting emission of interest is the distortion product tone at the frequency 2fi -.f2. DPOAEs are measured in the presence of the stimulus tones, but can be easily detected using narrow-band filtering (typically spectral analysis) because they are separated in frequency from the much larger stimulus tones. The current research indicates that the majority of hearing impaired individuals will be identified by a simple OAE test. Patients who fail to generate OAEs should be re-screened and/or.referred for additional audiological testing. This chapter deals with the development of DPOAE response acquisition system for the management of hearing screening [59]. The Figure 6.1 shows the block schematic of DPOAE measurement device. The first unit of the system is a high sensitive probe. Inorder to measure the OAEs, the probe incorporate a miniature microphone and miniature speakers. Two speakers are required to elicit DPOAEs, so that two stimulus tones are mixed acoustically 72
PROBE < - Data Acqusitio (mic & speaker) Card (DAQ) L_J Figure 6.1: Block diagram of DPOAE detection system rather than electrically to prevent artifactual intermodulation distortion products [91]. These stimulus are generated and presented to the ear canal via speakers located in the probe. A microphone in the probe measures the sound in the ear canal and transmits the signal to the ADC in the DAQ card after proper amplification. The function of amplifier unit is to raise the level of low voltage microphone output signal to a required value to improve the ADC performance. The DAQ output is then fed to the PC for further signal processing and detects any emission present [60]. 6.2 Hardware Systems 6.2.1 Probe Unit and Stimulus Generation The speaker/microphone pair chosen of the DPOAE measurement is OtoRead probe assembly which include two numbers of small speaker systems and one high sensitive low noise microphone. The stimulus tones generating circuit should be good in stability and linearity, proper in frequency band while not polluting the output signal. In this work, digital audiometry system is used for stimulus tone generation. These tones are generated in the frequency range 2 to 5 KHz with stimulus intensity level L 1 at 65 dbspl and L 2 at 55 dbspl respectively. These tones are then fed to probe system to elicit DPOAE. 73
6.2.2 Signal Detection and Amplification The high sensitive microphone in the probe collects re::;ponse signals from the ear canal. The signal level of microphone output is very small, order of 5 to 10 m V. This low level signal is converted by ADC, will resulting in poor resolution of the digitized incoming signal. Therefore, in order to increase the accuracy of DPOAE estimation and resolution of ADC, analog amplification is added between the DAQ device and the probe microphone as shown in Figure 6.1. In this work, two stage analog amplifier circuit is used. The first stage is an instrumentation amplifier based circuit, it is a low noise amplifier with RFI filter in the input side to eliminate RF interference. Main attraction of this stage is small size, battery operated amplifier and placed near to the probe system. The pre-amplified/filtered signal fed to next stage to provide required gain to increase the resolution of ADC. The amplifier circuit diagram for signal amplification is shown in Figure 6.2. R2 (var) Rg ADC USB6009 47k0 47kn Figure 6.2: Signal amplification circuit The main components used in the amplifier circuit are INA333 Instrumentation amplifier and OPA333 Operational amplifier. The first stage of the amplifier circuit is developed with INA333 amplifier. The INA333 is a low power precision instru- 74
mentation amplifier offering excellent accuracy. It provides very low offset voltage, excellent offset voltage drift and high common mode rejection. It operates with power supplies as low as + 1.8 V(±O. 9V) and quiescent current is only 50µA, ideal for battery operated system. A single external resistor sets any gain from 1 to 1000. Figure 6.3 shows the basic connections of INA333. The INA333 is designed to use an industry standard gain equation. G = 1 + looko/ R 9, where R 9 is external resistor for gain control, G is gain of the amplifier circuit. In this work, amplifier is designed to provide gain of lo. Therefore, R 9 llko. The two 47 Kn resistors in the input side provides a balanced input with possible advantages of lower input offset voltage as a result of bias current and better high frequency common-mode rejection. The second stage amplifier, the operational amplifier used in this circuit is OPA333 (µa741). Using the basic op-amp equation Vaut = ( 1 + ) n and letting R2 = 22KO, Rl = 1.2KO and n =50 to IOOmV (output of the first stage). Therefore, Vout approximately equal to 1 to 1.9 Volts. This value allows the ADC to use almost the full dynamic range available. Hence the value of R2 used in this circuit is 47 Kn, a variable resistor to adjust the voltage gain and Rl as 1.2Kn. The OPA333 series of CMOS operational amplifiers provide very low offset voltage (max 10 micro volts ) and near-zero drift over time and temperature. Single or dual supplies as low as+ 1.8V (±0.9v) and up to +5\, ( ±2. 75v) may be used. The OPA333 family offers excellent CMRR without the crossover associated with traditional complementary input stages. This design results in superior performance for driving Analog-to-Digital Converters (ADCs) without degradation of differential 75
V 0.1µF VIN- 150k0 150k0 RFI Filter 5/JkO 6 RFI Filter Load V 0 V --irfi Filter IN+ 4 0.1 INA333 Ref V- Figure 6.3: Basic connections of INA333 linearity. 6.2.3 USB 6009 DAQ Device The main part of DPOAE signal analysis is conversion of amplified microphone output analog signal into digital signal. This task is performed by ADC available in the Datta Acql!D.isition (DAQ) card. In this work USB6009 Data acquisition (DAQ) devire i'.s wed. This USB based device provides connection to eight analog inputs, ttwo ;aj!jl.atdog output channels, and a 32 bit counter when using full speed USB interface. The internal diagram of USB6009 device is shown in Figure 6.4. The amplified signal is fed to the computer through this USB device. This device convert the detected analog signal into digital format for further signal processing and analysis. Figure 6.5 illustrates the analog input circuitry of the USB6009. The USB 76
Power Su I 5 V 200mA USB Microcontroller 8 Channel 12/14b ADC 12b DAC 12b DAC Figure 6.4: The internal diagram of USB6009 device +2.5VREF ADC AI FIFO Input Si gn al Input Range Selection Figure 6.5: Analog input circuitry of USB6009 DAQ 6009 has one Analog-to-Digital converter (ADC). Multiplexer (MUX) routes one AI channel at a time to PGA. The programmable gain amplifier provides input gain of 1, 2, 3, 5, 10, 16, or 20 when configured in differential measurement and gain 77
of 1 when configured for single-ended measurements. ADC digitizes the AI signal by converting analog voltage into digital code. The main features of ADC in USB 6009 are 14 bits differential(13 bits single ended) AI resolution, maximum sampling rate 48 KS/s. 6.3 DPOAE Detection and Analysis The detection and analysis of DPOAE is developed in Lab VIEW 7.1 Software /NI -DAQmx module. The Figure 6.6 shows the block schematic of the system. Amp lifted Micro phone outp ut DAQ ASSIST XY graph Amplitude level Measurement._ DPOAE.. Preprocessing unit.. Band Pass Filter.. fd.. Spectral,. Smoothing Measurement Filter Figure 6.6: Block diagram of representation of DPOAE detection using Lab VIEW The amplified microphone output is fed to the computer system through DAQ device for further signal analysis. The DAQAssrsr module is used for settings of channels, acquisition modes, terminal configuration and sampling rate etc. The preprocessing module includes a low pass filter (LPF) and a high pass filter (HPF). The filter module processes the signals through filter and windows. The function of LPF is to exclude instrumentation noise and to prevent aliasing in FFT. It is programmed 78
with a cutoff frequency of 6 KHz. The HPF excludes the largest components of ambient and subject noise, which are predominantly at low frequency range. Cutoff frequency of HPF is selected as 600 Hz. The narrowband band pass filter with centre frequency fd (2fi -h) is used to attenuate all components except the DPOAE signal as much as possible to enhance the quality of input signal. A smoothing filter unit is also used to reduce the effect of any high frequency signal. The spectral analysis module performs spectral measurements such as peak spectrum and power spectrum of the signal. The XY graph gives the spectrum of detected DPOAE signal and amplitude level measurement module gives the voltage level of DPOAE signal and its noise level. 6.4 Results In this work, DPOAE data are collected from hearing subjects using the system developed for hearing screening. A total of 24 hearing subjects (48 Yrs) are participated in this test. Figure 6.7 (:a) shows the amplified microphone signal output which is fed to the data acquisition device, Figure 6.7 (b) shows the spectra of this signal which contains two strong stimulus and a weak DPOAE signal atf i, h, 2f i - h respectively and Figure 6.7 (c) shows the frequency spectrum of detected DPOAE signal. Screening test is also conducted using pure tone audiometric data whicht is measured using ELKON GIGA3 audiometer. The nom1al hearing subject is coded as PASS, if subject having mean hearing level (MHL) less than 30 db and hearing impaired subject is coded as REFER if MHL greater than or equal to 30 db. Table 6.1 shows the result of DPOAE screening and pure tone audiometric screening. The analysis of total PASSes and REFERs resulting from the DPOAE method are compared with that of pure tone screening and examined for sensitivity and specificity 79
1 0.75-05- r~- -0.25-05- -0.75-0 I I I I I 0.02 0.04 0.Q6 0.08 OJ. Time (a) -100.., -200- -a t -300 ~-400 I I I I, SOD 1000 1500 2000 2500 (b) -500- -60- -80-.., -100-." ~ "5.-120- ~ -140- -160- -180-1 I I I I 0 500 1000 1500 2000 Fr (c) Figure 6.7: (a) Amplified microphone output signal(b) Frequency spectrum ofdetected signal. (c)frequency spectrum ofdpoae signal 80
to establish the validity. Screening Screening Results Total No. Method PASS REFER of Ears Pure.Tone 37 11 48 DPOAE 35 13 48 Table 6.1: PASS/REFER outcome of Pure tone Audiometry and DPOAE REFER PASS REFER 10 3 PASS 1 34 Table 6.2: Screening results of pure tone and DPOAEs Comparison of pure tone with DPOAE procedures are given in Table 6.2. The analysis determined that 34(91.89%) of all ears PASSed both pure tone screening and DPOAE screening, where as 1(9.1 %)of the ear that REFER in pure tone PASSed in DPOAE. 10(90.9%) of the ears that REFER in pure tone screening are also REFERed in DPOAE screening. Of the pure tone screening PASSed, there are 3(8.11 %) that REFER in DPOAE screening. From the above results specificity, sensitivity and accuracy of DPOAE with respect to pure tone screening are 91.89%, 90.9% and 91.6% respectively. This DPOAE acquisition system doesn't require any behavioural response from the hearing subjects and provides good results in a faster way. 6.5 Conclusions A prototype of DPOAE measurement system has been designed and developed for hearing screening. Its results are verified with 48 ears and compared with pure tone audiometry. The main features of the system are its simplicity, low cost, fast 81
operation and its immunity to external noise. It is definitely an effective tool to medical practitioners for hearing screening. 82