(7-1) Evoked potential and Acoustic emission

The following report is about experiments on evoked potential and acoustic emission how they are stimulated by pulsive sounds It was aimed to see how evoked potentials are different with a rectangular pulse or a click sound. The latter experiments were for the acoustic emission stimulated by a rectangular pulse of 0.05ms.


(1) Evoked potential stimulated by a rectangular pulse of 0.05ms
1-1) Experimental chamber
An experimental chamber was built with plywood having the ceiling sloped. Plan has 1.8m by 2.25m and the height at the higher side was 2.4m and the lower 1.8m.
The inside was shield with brass mesh first and a 10cm thick glass fiber layer was laid. Its surface was covered by a sheet of Venetian cloth to be sound absorbent and thermal insulated.

1-2) Evoked potential by a rectangular pulse and a click sound
1-2-1) Experiments
A full range unit loudspeaker was set at the end of a CV cylinder and fixed under the ceiling. A bed was cut into the shape of a man and covered with a urethane foam layer. A tested person can lay down on the bed relaxed. The experimental chamber is shown in Fig.1 with the measuring gears.


Fig.1 Chamber for evoked potential measurement

Evoked potential was measured at the vertex (Cz) referred to the potential at the left earlobe (A1). The trigger signal for them was fed when the loudspeaker got an input signal. It was added 500 times in phase. The repetition duration of stimuli was chosen 3 seconds after the experience of hearing response experiments to avoid any interaction. The negative reaction of the loudspeaker was canceled by the convolution of the inverse filter. Thus, a rectangular pulse of 0.05ms was made for the experiments. It is the same signal as used to find the linear part of the response of our hearing system. The response is assumed as the peripheral response and the interest is how it happens at the higher level in our hearing system.
There have been discussed on the response by a pulse sound input1), but it was a click sound with negative reaction and it is not clearly defined. Their experimental results must have been so much affected by input signals.
The output sound of the loudspeaker with a rectangular signal input had the following negative reaction. It is called a click sound for comparison. The rectangular pulse and the click sound are compared in Fig.2.

Fig.2 Rectangular pulse and click sound

1-2-2) Measured results
To see the short time latency I`VII the time axis was enlarged, and to see the middle and long latency P1AP2AN1AN2c it was shortened. They are shown in Fig.3.


Fig.3 Example of evoked potential by a rectangular pulse and click sound

1-2-3) Discussions
These two pulses did not show much difference. Even for the rectangular pulse, it was convolved with the head related transfer function and looked similar as the click sound.
If an impact sound has a quick rise and finishes in a short while, the evoked potential was strongly affected by the rise and the response to the negative reaction could have been masked.
Their difference could be found in their evoked potentials if we could estimate the details but it was beyond our ability.

(2) Evoked potential stimulated by successive pulses

2-1) At the time interval of 2.5ms
Evoked potentials were measured giving two successive pulses with 2.5ms time interval which are shown in Fig.4.

Fig.4 Two successive sounds with the interval of 2.5ms

It was expected two successive rectangular pulses would be heard separated but they were heard as one, as it happened on the click sounds. However, there occurred a slight difference recognized as shown in Fig.5.


Fig.5 Evoked potentials by two successive rectangular pulses and click sounds

2-2) At the time interval of 60ms
When two pulses were heard with the 60ms time interval, a group of seven waves is observed again 60ms later on the evoked potential as shown in Fig.6.
It is interesting too that it happened before the evoked potential completely finished.


Fig.6 Evoked potentials by two successive pulses with the time interval of 60ms

2-3) Discussions
The short time latency V wave was recognized only at the time of the second pulse for two successive rectangular pulses with the 2.5ms interval on two persons out of three.
As for other waves, the first pulse got seven waves corresponding on the time axis.
When the time interval of two pulses was 60ms, seven waves of short time latency emerged twice having the time interval of stimulation.
This time region corresponds to the one which happened to hear two rectangular pulses as three being referred to the Gestalt psychology2). The Gestalt psychological response for two successive pulses must be referred to the SVR or the alpha brain wave. Also the introduction of technique to measure till P300 should be discussed.
The rectangular pulse of 0.05ms is a signal which does not have any auto-correlation. When it is repeated with a certain time interval, where its response emerges is interesting, namely which part in the hearing system it can be observed. If we could know the propagation speed in the system, we can localize where and how it is accepted.
If we follow with a pure tone, we tend to go into a swamp. However, if it is discussed with the rectangular pulse of 0.05ms, it gives us clear and inclusive information.
If the amplitude of the rectangular pulse is given with a wide rage, how the system establishes the logarithm scale on the point of magnitude estimation.

(3) Acoustic emission by a rectangular pulse of 0.05ms
3-1) Experiment
The acoustic emission is the slight reaction from the inner ear at hearing sound. To detect it, the rectangular pulse of 0.05ms was used for stimulation. This can be detected by the superposition of it, but the acoustical response which is given by the physical response before it reaches the eardrum must be eliminated.
The earlobe and the external ear canal of a tested person were casted with plaster and they were molded with silicon rubber to simulate the physical parts before the eardrum.
They were attached to a dummy head which is filled with fiber glass. The eardrum was made by a thin plastic sheet.
A microphone was placed for the dummy head at the same place as the acoustic emission was measured for a tested person. The physical response thus measured was subtracted from the measured acoustic emission A 1/2 inch B&K probe microphone was placed a little bent towards the external ear canal. Stainless steel fiber was filled the probe for acoustical absorption.
A rectangular pulse of 0.05ms was repeatedly given 2000 times every 3 seconds. The 2000 times added result which is subtracted the physical reaction mentioned is shown in Fig.7.The early part has high frequency component and low frequency later stage as can be seen.
To see the frequency dependency Zwickerfs filters for 50Hz, 350Hz, 1000Hz, 2150Hz and 4800Hz were convolved to the above response. Each filter is shown in Fig.8 whose inverse transformation is a triangle as shown for an example of 1000Hz. For 50HZ, the Hanning low pass filter of 100Hz was used.

3-2) Measured results
First the acoustic emission in Fig.7 was cut at zero around 7.3ms after it reached the eardrum. It is shown with . It was used to be convolved with the filters in Fig.8.
As it is shown in Fig.9, the acoustic emission filtered with a critical band shows that high frequency comes earlier and low frequency comes later.
They are the responses from the basilar membrane.


Fig.7 Acoustic emission to a rectangular pulse of 0.05ms after 2000 times simultaneous addition


Fig.8 Zwickerfs critical band filters in the time domain and frequency domain


Fig.9 Acoustic emission in the time domain for each critical band filter

Reference
1jFor instance, I.J.Hirsh: hTemporal aspects of hearingh in the eNervous system, vol.III Human communication and its disordersf, E.L.Eagles Ed. (Raven Press, New York, 1975), p157-162.
2) Y.Sakurai and H.Morimoto,h The transient response of human hearing systemh, J.Acoust.Soc.Jpn.(E), 10, 4(1989).