Method and device for examining the faculty of hearing

Abstract

In a method for examining the faculty of hearing for at least one ear of a mammal, in which growth curves are determined on the basis of the measurement of DPOAE's evoked by pairs of excitation signals (f1, f2) for different excitation frequencies f2, the ear is presented with first excitation signals with a first excitation frequency f1 and a first noise level L1 and secondary excitation signals with a second excitation frequency f2 and a second noise level L2. Pulse pairs with a first pulse of the first excitation signal and a second pulse of the second excitation signal are presented in the ear, and the DPOAE's evoked thereby are captured and evaluated. A set of at least two different pulse pairs with different second excitation frequencies f2 is presented in one block that is repeated several times during a measuring period.

Claims

1. A method for the examination of the hearing ability for at least one ear of a mammal, the method comprising the steps; providing at least one ear probe on or in the ear, the at least one ear probe including one or two miniature loudspeakers designed for et emitting excitations signals and one receiver configured to capture and forward distortion product optoacoustic emissions (DPOAE's) evoked by the excitation signals; providing a computer unit that is operably connected to the at least one ear probe; generating, by the computer unit, a set of at least two different pulse pairs in a measurement block, each pulse pair in the set being presented in a slot having a length T, the slots occurring sequentially without overlapping, the block being a sum of all the slots, the block having a block time being defined by a product of the length of the slot T and a total number of the pulse pairs in the set, a pulse being an on-and-off process within the slot, each pulse pair having a first pulse of a first excitation signal and a second pulse of a second excitation signal, the first excitation signal having a first excitation frequency f1 and a first signal level L1, the second excitation signal having a second excitation frequency f2 and a second signal level L2, the at least two different pulse pairs in the set each having different second excitation frequencies f2, the set being repeated several times in a measuring period containing more than one block; presenting in the ear the pulse pairs of the first and second excitation signals one pulse pair at a time, and capturing and evaluating the DPOAE's evoked by pairs of excitation signals for different second excitation frequencies; outputting growth curves on the basis of the measurement DPOAE's evoked; whereby the outputted growth curves is used as basis for evaluating the hearing ability and diagnosing hearing conditions.

2. A method according to claim 1, wherein a duration of the first and the second pulses in a pulse pair is 2 to 20 ms.

3. A method according to claim 1 wherein the length of the slot T is >10 ms.

4. A method according to claim 1, wherein, in one block, the second excitation frequencies f2 of two immediately consecutive pulse pairs are at least one octave apart.

5. A method according to claim 1, wherein measured signal levels of the DPOAE's for pulse pairs of the same second excitation frequencies f2 are averaged during a measuring period.

6. A method according to claim 1, wherein a start of a first and a consecutive pulse pair with the same excitation frequency f2 have a time lag of 30 to 100 ms.

7. A t method according to claim 1, wherein, at a beginning of the measurements, a check occurs to determine whether the frequency fdp of one of the DPOAE's interferes with a spontaneous emission (SOAE).

8. A method according to claim 1, wherein, within one block, the second signal levels L2 of the pulse pairs have a level difference from one another that is smaller than 15 dB.

9. A method according to claim 1, wherein a sequence of the pulse pairs and a time lag between two immediately consecutive pulse pairs in a block remains constant.

10. A method according to claim 5, wherein, once a desired signal-to-noise ratio for an excitation frequency f2 has been reached, further averages planned for this excitation frequency f2 are skipped.

11. A method according to claim 1, wherein at least two sets of pulse pairs are generated that are at least partially different in terms of the second excitation frequency f2, each set presented as one block, wherein the blocks are presented in chronological sequence.

12. A method according to claim 11, wherein it is continually checked for each pulse pair whether a desired signal-to-noise ratio has been reached and that the pulse pairs for that excitation frequency f2 are eliminated with a further measurement, and the remaining pulse pairs are, if required, redistributed to the blocks.

13. A method according to claim 11, wherein pending pulse pairs are continually checked as to their compatibility with regard to noise level and time lag, and it is further checked whether pending pulse pairs in a block may be re-allocated to freed slots.

14. A method according to claim 1, wherein the noise levels of the DPOAE's for all the second excitation frequencies f2 contained in the or every set are measured and averaged for a second noise level L2 respectively allocated to the excitation frequency f2, and the measurements are conducted at least once for new noise levels L2.

15. A method according to claim 14. wherein for each excitation frequency f2, a growth curve of measured values of the signal levels of the DPOAE's is determined for various signal levels L2, and respective threshold values are then determined from said growth curves.

16. A method according to claim 15, wherein the growth curve is determined from at least three values for the signal level Ldp of the DPOAE determined at three different signal levels L2, but the same excitation frequency f2, wherein the at least three signal levels L2 include an upper signal level L2 that is used to determine a lower signal level L2, wherein the third, middle signal level is determined by means of the upper and the lower signal levels.

17. A method according to claim 16, wherein a preliminary lower noise level L2 is determined from the upper signal level L2 and population data, and then the middle signal level L2 is determined from the upper signal level L2 and the preliminary lower signal level, said middle signal level being in the middle between the upper and the preliminary lower signal levels L2.

18. A method according to claim 16, wherein a final lower signal level L2 is determined from the upper signal level L2 and the middle signal level L2.

19. A device for the examination of the hearing ability for at least one ear of a mammal, comprising: at least one ear probe configured to be placed on/in the ear, the at least one ear probe including one or two miniature loudspeakers and one receiver wherein the one or every miniature loudspeaker is designed for the presentation of a first excitation signal with a first excitation frequency f1 and a first signal level L1 and/or a second excitation signal with a second excitation frequency f2 and a second signal level L2, and wherein the receiver is designed for the capture and forwarding of a DPOAE's evoked by the first and second excitation signal; and a computer unit operable connected to the at least one ear probe, the computer unit configured to present a set of at least two pulse pairs in a block, each pulse pair having a first pulse of the first excitation signal and a second pulse of the second excitation signal, the at least two pulse pairs in the set having different second excitation frequencies f2, each pulse pair in the set occurring in a slot having a length T, the slot following a preceding slot sequentially without overlapping, the block being a sum of all the slots, the block having a block time defined by a product of the length of the slot T and a total number of the pulse pairs in the set, a pulse being an on-and-off process within the slot, the computer unit further configured such that said set is run repeatedly several times for one measuring period containing more than one block, the computer unit further configured to determine and output growth curves on the basis of the measurement of DPOAE's evoked by pairs of excitation signals for different second excitation frequencies used for evaluation of the hearing ability and diagnosis of hearing conditions.

20. A hearing aid comprising a device according to claim 19.

21. The method according to claim 1, wherein a ratio of f2/f1 is constant.

22. The method according to claim 1, wherein the pulse pair signals do not use the entire time of the slot.

23. The method according to claim 1, wherein the first pulse and the second pulse in a pulse pair are presented at a temporal offset.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Exemplary embodiments of the invention are explained in more detail in the following description with reference to the appended drawings. They show:

(2) FIG. 1 a basic sketch of a device, with which the new method is performed;

(3) FIG. 2 a basic growth curve as it can be recorded and extrapolated using the new method;

(4) FIG. 3 the envelopes for four different f2 excitation pulses with the frequencies 1.5; 4; 2; 3 kHz for the fixed-block time-frequency pulse combination method with fixed pulse pair arrangement;

(5) FIG. 3b the envelopes for four different pulse pairs with respectively four different excitation frequencies f2 and f1 with the frequencies f1 1.25; 1.67; 2.5; 3.33 kHz and the frequencies f2 1.5; 2; 3; 4 kHz for the fixed-block time-frequency pulse combination method with fixed pulse arrangement;

(6) FIG. 4 the distribution of a panel of four f2 pulse pairs to four slots in a block for the fixed-block time-frequency pulse combination method with fixed pulse pair arrangement;

(7) FIG. 5 the distribution of a total of seven f2 pulse pairs to two panels and two blocks with four slots each for the fixed-block time-frequency pulse combination method with fixed pulse pair arrangement;

(8) FIG. 6 the distribution of a total of seven f2 pulse pairs to three panels and three blocks with four slots each for the flexible-block time-frequency pulse combination method with fixed pulse pair arrangement;

(9) FIG. 7 the distribution of a total of seven f2 pulse pairs to four panels and four blocks with a variable number of slots for the flexible-block time-frequency pulse combination method with free pulse pair arrangement, under the simplified assumption that the measurement periods for the selected f2 pulse pairs are uniform; and

(10) FIG. 8 the distribution of a total of seven f2 pulse pairs to six panels and six blocks with a variable number of slots for the flexible-block time-frequency pulse combination method with free pulse pair arrangement, without the simplified assumption that the measurement periods for the selected f2 pulse pairs are uniform.

DETAILED DESCRIPTION OF THE DRAWINGS

(11) A device 10 to be used according to the invention is shown schematically in FIG. 1; it typically comprises at least one ear probe 11, which is connected to a sound card or another AD card in a computer unit 15in this case, a computervia cable 12, 14.

(12) According to DE 199 05 743 A1 mentioned above, a transmission path can be provided between a headphone carrying two ear probes and the computer, in order to avoid disturbing noises that could be caused by the cable. Because of the two ear probes, the measurements can be binaural in this case.

(13) Each ear probe 11 comprises at least one highly linear loudspeaker 16, which emits the two excitation signals f1, L1 and f2, L2, and at least one microphone 17, which measures the DPOAE's, i.e., their sound pressure levels Ldp at a frequency fdp=2 f1f2, adjusted using f2. Often times, each ear probe 11, however, comprises two loudspeakers 16, so that no technical distortions occur in the stimulation with the two tones f1, f2, said distortions being hard to separate from the physiological DPOAE's.

(14) The computer can be provided externally or integrated into the ear probe 11. The computer is designed to generate the pulse pairs (f1, f2) used according to the invention and to adjust them adaptively during the measurement. It can store the pulse pairs (f1, f2) and the measured DPOAE's (Ldp, fdp) for later analysis, and can make them available to be read.

(15) Alternatively, the computer can also perform the analyses in real time.

(16) The results of the measurement are so-called growth curves for the selected excitation frequencies f2, from which the computer 15 then determines the respective threshold values, which constitute an objective evaluation of the faculty of hearing and can be used for different purposes. It is important in this respect that the new method allow for a very quick determination of the growth curves, which promotes acceptance of the method.

(17) A basic growth curve 21 is shown in FIG. 2. The measured level Ldp of the DPOAE's is plotted for an excitation frequency f2 against the sound level L2. From the seven measured values 22 for the DPOAE's, the threshold value 23 is extrapolated.

(18) For the adjustment of the signals f1, f2 output by the computer to the loudspeaker 16 integrated into the ear probe 11, a final amplifier and/or an impedance adjustment and frequency response correction can be provided, which is realized in the form of a passive or active electronic circuit.

(19) For the adjustment of the electrical signals generated by the microphone 17 to the computer interface, a preamplifier and a frequency response correction can be provided, which is also realized in the form of a passive or active electronic circuit.

(20) If not all processing steps described are performed in the ear probe 11, either a cable connection or a wireless connection is provided. If all processing steps are performed in the ear probe 11, a wireless connection is provided for the transmission of the measurement data to a playback device, either after the measurement or in real time.

(21) The data can also be stored as clinical data in clinical information systems.

(22) The devices 10 described in this respect are used for examining the faculty of hearing by pediatricians, in ENT clinics, by ENT physicians, by hearing aid acousticians, and by patients at home. They can, however, also be integrated directly into hearing aids, in order to adjust them in a way during operation to a change in the hearing of the wearer.

(23) FIG. 1 schematically shows a hearing aid 20, in which the device 10 is arranged.

(24) The method to be performed with these devices was described above in its individual required and preferred steps. Below is a complete overview of preferred exemplary embodiments with, in part, slightly modified nomenclature, which exemplary embodiments are, however, not to be considered as limitations on the extent of protection or the scope of the present invention.

(25) 1. Introduction

(26) A time-optimal procedure for measuring DPOAE growth functions must take into account different levels of measured value acquisition and analysis, which are classified as follows:

(27) 1.1) Measurement of an Individual DPOAE

(28) This requires determining a minimally-achievable signal-to-noise ratio, determining how the signal and the noise should be calculated, a suitable artifact reduction procedure, and, in general, a maximum measurement time, after which the measurement is to be aborted as unsuccessful.

(29) 1.2) Threshold Value Approximation

(30) When measuring growth functions, there is the question of at which excitation levels and in what sequence the single measurement values must be taken, and of how many single measurement values there should be. For the extrapolation procedure, it has been shown that, often, only the three points closest to the threshold of a growth function contribute to the estimation of the extrapolated threshold, and that the accuracy of the estimation increases when other points are omitted. This results in the idea of targeting these three points when conducting measurements, in order to save time on the remaining points.

(31) As will be shown, the measuring point closest to the threshold is the most critical one. If its excitation level is chosen to be too low, either an average over too long a time must be calculated or it can even occur that, after the maximum single measurement time runs out, it cannot be determined with sufficient reliability. If it is too high, the extrapolation error grows, since the last measuring point is located farther from the estimated threshold.

(32) Thus, according to the invention, a procedure is provided, which, based upon a first valid single measuring point that preferably should be the measuring point that is the third and furthest from the threshold, determines both the following measuring points according to optimal criteria, wherein, with each following measuring point, the calculation of the optimal next excitation level shall be adaptively refined.

(33) The question of how the first single measurement value furthest from the threshold should be excited shall be considered separately. Essentially, what is considered here is a global strategy, which can emerge in different ways, depending upon the application.

(34) In general, it is useful to excite the first measurement value at around L2=45 dB SPL, since, above this level, according to the invention, growth functions for normal hearers are already saturated, i.e., they produce values that are not helpful for an optimal threshold estimate. If, for example, newborns were omitted from a sieve test, this would be a good option, because most newborns have normal hearing, and, therefore, the average measurement time for the sieve test would probably be minimized, since, for most of the population, only three single measurement values (per frequency) need to be obtained, and it is rare for the first threshold-furthest measuring point to be attempted without success.

(35) With patients who come to the clinic with hearing loss, growth functions that are appropriate for a normal-hearing population are much less common, while the number of cases whereby the maximum measurement time for a single measurement value is used up at L2=45 dB SPL without obtaining a valid value can already be considerable.

(36) In this area, procedures that newly position a growth function if there is no optimal excitation should also be considered, and so should those that adjust growth functions depending upon the results from other frequencies for multi-frequency measurements (i.e., pulse interleaving procedures with multiple frequencies).

(37) 1.3) Temporal Pulse Interleaving Procedure

(38) As a rule, the state of the inner ear should be determined, not at a single frequency, but at a suitable number of frequenciestypically at octave or half-octave intervals in the 1 kHzf28 kHz range. This means that multiple growth functions must be measured.

(39) When compared to continuous DPOAE, pulsed DPOAE have the severe drawback that the measurement at a frequency-noise level combination in general has a low duty factor and, therefore, produces a correspondingly low signal-to-noise ratio for the same measurement time.

(40) This drawback, however, is enormously reduced with the invention, in that several measurements are interleaved in the time-frequency space, i.e., they are presented alternately time-delayed. Thus, for example, within a sufficiently often-repeated block, 7 time-delayed frequencies can be stimulated and analyzed.

(41) The simplest implementation is a fixed arrangement of excitation pulses inside a block; this is called a block-fixed procedure. When arranging the pulse pair within a block, the frequency and time sequence must be chosen in way that allows the signals to interfere with each other only within a negligible range.

(42) This procedure functions almost optimally, especially when the noise background and the occurrence of artefacts is similar at all the frequencies used.

(43) Further optimization is achieved when the stimulating pulse pair is arranged, not in a block-fixed, but in a block-flexible way. This leads to additional time-saving in situations where the measurement at one frequency-level combination has already achieved a sufficient signal-to-noise ratio, while another still requires significantly more time.

(44) In this case, in the block-flexible pulse interleaving procedure, before the completion of one of the single measurements currently taking place in the block, one or several pulse pairs that already have achieved the required signal-to-noise ratio are replaced with those of another frequency or level, which are still being measured.

(45) For this purpose, a standard rate is produced, which sets the minimum distance in the time-frequency-level space.

(46) In detail:

(47) 2.1) Measurement of an Individual DPOAE

(48) Suitable procedures for determining criteria for inclusion of measurement valueshere, the DPOAEin the averaging process, and for determining an average value that returns a preset signal-to-noise ratio (SNR), are sufficiently known from the state of the art; see, for example, Mller and Specht, Sorted averagingprinciple and application to auditory brainstem responses, in Scand. Audiol. 1999, 28: 145-9.

(49) The inventors, however, have determined that the decay time of pulse responses can present a problem, when the measurement time is to be reduced by decreasing the block length.

(50) During typical measurements in the frequency range of around f2=2 kHz, a block length of T=70 ms is sufficient to make sure that the pulse response has already decayed within one block, so that it does not produce any appreciable measuring error due to interference with the pulse response from the repeated simulation by another pulse pair.

(51) This, however, is not true when the frequency of the DPOAE in question (as a rule, fdp=2f1f2) is close to a spontaneous otoacoustic emission (SOAE).

(52) In order to limit the measuring error produced by these problems, the inventors employ various procedures: A) Adjusting the block length to the decay time of the pulse under a certain level. This procedure has the drawback that, under certain circumstances, a significant prolongation of the measurement time must be accepted, and that a certain averaging with associated time costs is already necessary to determine the decay time. Thus, the procedure is only useful if the necessary measurement is at a precisely maintained frequency. B) In the clinical practice, the measurement of a precisely defined frequency is, however, not necessary. In the general view, a sufficient representation of the state of the cochlea can be obtained by measuring frequencies at octave intervalswith higher requirements, at half-octave intervalsthus, for example, at f2=0.75, 1, 1.5, 2, 3, 4, 6, 8 kHz.

(53) In such applications, measuring at an f2 frequency that is within 50-150 Hz of the target frequency might be sufficient. Thus, if the presence of SOAE's needs to be checked, either a) in an a priori measurement, whose assessment will be used to offset the actual f2 frequency of the stimulus to maintain a sufficient safety distance between SOAE's and fdp, or b) when measuring DPOAE's, to be able to recognize the problem with too high decay time at its f2 target frequencies, via an algorithm that is executed during the measurement, with a repetition of the measurement with offset f2 stimulus frequencies, if necessary.

(54) The choice between these two variants can be decided on the probability of occurrence of SOAE's for the given application. If, for application with severely hearing impaired, SOAE's are only rarely to be expected, the time costs can be saved by omitting SOAE's (e.g., hearing screening for older adults); in contrast, with newborns, SOAE's are almost guaranteed to be present, and the probability of their distorting the DPOAE measurement is higher.

(55) For A), an SOAE measurement typically takes 40 s to obtain a noise background of 30 dB SPL at f=2 kHz. The necessary frequency offset can be adjusted, depending upon the strength of the SOAE's found.

(56) For B), the optimized averaging algorithm is to be used. It predicts that when blocks with unusually high noise levels are excluded, the noise follows a clear rule. If it does not, or if, after a certain time, practically every block is excluded, it is a sign that the chosen noise value no longer accords with the assumption of a random, uncorrelated process.

(57) If, for example, the noise figure is related to the amplitude of the envelopment of the time signal immediately before another pulse response is expected, the averaging process can function properly for a certain number of blocks. However, as soon as the noise amplitude drops to the value of the decayed pulse response to the previous stimulus pulse, the amplitude will not be reduced by averaging anymore, if the decaying pulse response is highly reproducible, and, thus, shows a high correlation.

(58) A scheme for the measuring start of a growth function assumes that the measurement begins with a starting level L2 for the second primary tone at f2. The level L1 for the first primary tone f1 is determined for the frequency in question according to a standard curve, e.g., according to the level range as described by P. Kummer et al. in The level and growth behavior of the 2f1f2 distortion product otoacoustic emission and its relationship to auditory sensitivity in normal hearing and cochlear hearing loss, in J. Acoust. Soc. Am., 103(6): 3431-3444, 06 1998.

(59) If no DPOAE can be measured, the level L2 of the second primary tone is increased by L2, until the maximum achievable primary tone level is produced. If no DPOAE is found even then, another optimization of the first primary tone level L1 is attempted.

(60) If a DPOAE is measured, the threshold value approximation procedure described below follows.

(61) 2.2) Threshold Value Approximation

(62) 2.2.1 Determination of the First Measuring Point

(63) Since experience with normal hearers shows that the DPOAE growth functions saturate partially relatively early, the first measured value should be picked so that it corresponds with the highest possible excitation level, at which, as a rule, no saturation is to be expected. In the case of a normal hearer, a point thus measured would still be usable and measured without loss of time. It would effectively be the threshold-furthest point of the growth function.

(64) Experience thus far suggests the excitation level of L2=45 dB SPL. If this point is selected to be too low, the number of cases where the maximum permissible measurement time has elapsed without obtaining a usable single measurement value grows.

(65) The strategy can optionally be varied as follows: The maximum permissible measurement time is reduced for the first measuring point. This results in less time being lost, if the DPOAE threshold is above L2. If it is only slightly lower, then, in a later step, when L2 will be measured with a longer measurement time, another attempt at obtaining the same measuring point can be made. This last method uses the advantage of the comparably quickly obtained information of the upper points of the growth function, in order to be able to adjust time-optimally the level of the lowest point of the growth function, whereby the highest time efficiency is achieved.

(66) 2.2.2 Determination of the Second Measuring Point

(67) Hereinafter, it will be assumed that there is a measurement of three pulse pairs for a given f2, i.e., of three different noise levels L2, for each second excitation frequency f2. This corresponds to the minimal number of points that can be used in the extrapolation procedure with quality control (determining the correlation coefficients and the standard deviation of the estimated threshold).

(68) The measurements of normal hearers show a high correlation between the slope of the growth function and the amplitude of the DPOAE at an average excitation levelaround L2=45 dB SPL. It is thus advisable to choose the second excitation level so that it falls between the first and the third excitation levels. For this purpose, the slope of the growth function is estimated with the help of population average values (see Dalhoff et al., 2013, Hear. Res. Bd. 296, Table 2, p. 77) that are best determined independently of frequency.

(69) The slope can then be used for determining the time-optimal excitation level L2 for the last, threshold-closest measuring point. For this purpose, Section 2.2.3 provides a procedure, which then in equation (2) is used as m for the population average value mentioned above. With the first (threshold-furthest) measuring point and the population data, the threshold-closest (third) measuring point is estimated. The second measuring point is then placed in the middle, between the first and the third point.

(70) 2.2.3 Determining the Last Measuring Point at the Lowest L2 Excitation Level

(71) It is assumed that two or more measuring points for a growth function are already available at higher excitation levels L2. In accordance with Section 2.2.2, for example, both the higher measuring points are available, so that now, the third (threshold-closest) measuring point can be determined againthis time, however, with an individually defined slope.

(72) This requires solving the problem of how to choose the last and lowest L2 excitation level that maintains, with the lowest possible time costs, minimal estimation error for the threshold of the growth function in the final extrapolation.

(73) Two questions must be answered for this purpose. First: What influence does the measuring error for the last, threshold-closest measuring point resulting from limited averaging time have on the regression, and thereby, on the estimation error of the extrapolated threshold? It is generally clear here that a measuring point close to the threshold (with the same measuring error) leads to lower extrapolation error. Second: The last point of the growth function must, however, be measured at a certain signal-to-noise ratio in order to be viewed as valid; this means that the closer it is selected to be to the threshold of the growth function, the longer the required averaging time will be. Both these considerations combined produce a conclusive optimality criterion.

(74) The points of the growth function should be numbered from the lowest to the highest excitation level, i.e., opposite to their time order. Next, we search for point P.sub.1 with excitation level x.sub.1, while points P.sub.2, . . . , P.sub.n are already available. Based upon the points available, a preliminary regression line (or, at n=3, a connecting line) can be found, with y(x)=m.sub.e,n1x+b.sub.e,n1, where e stands for estimated, and n1 represents the number of points involved in the estimation.

(75) It will be assumed in what follows that y(x) describes the actual growth function without any error, while the additional new measuring point produces a measuring error N.sub.1 due to the final averaging time, so that the measured (or estimated from the measurement) value has an amplitude of y.sub.1,e=y.sub.1+N.sub.1, if y.sub.1 is the true amplitude.

(76) The slope of the regression lines based upon the measurement of P.sub.1 (which includes error) is calculated with:

(77) $\begin{array}{cc}{m}_e=\frac{\stackrel{\_}{{\mathrm{xy}}_e}-\stackrel{\_}{x}\stackrel{\_}{y_e}}{\stackrel{\_}{x^2}-{\stackrel{\_}{x}}^2}& \left(1\right)\end{array}$ The line here means that the averaging process is variable. Therefore:

(78) $\begin{array}{cc}m=m_e-m=\frac{\stackrel{\_}{{\mathrm{xy}}_e}-\stackrel{\_}{x}{\stackrel{\_}{y}}_e}{\stackrel{\_}{x^2}-{\stackrel{\_}{x}}^2}-m& \left(2\right)\end{array}$ The intercept is calculated with:
b.sub.e=y.sub.em.sub.e{tilde over (x)}(3) and b=b.sub.eb becomes

(79) $\begin{array}{cc}b=\frac{N_1}{n}-m\stackrel{\_}{x}& \left(4\right)\end{array}$ The extrapolation error x.sub.edpt present due to the measuring error is determined with the quotient rule, provided the error is not too big:

(80) $\begin{array}{cc}{x}_{\mathrm{edpt}}=\frac{bm-mb}{m^2}& \left(5\right)\\ \mathrm{Therefore}\text{:}& \\ \mathrm{nm}{x}_{\mathrm{edpt}}=N_1\frac{\left(x_1-\stackrel{\_}{x}\right)\left(\stackrel{\_}{x}-x_{\mathrm{edpt}}\right)-}{}& \left(6\right)\\ \mathrm{abbreviated}\text{:}& \\ =\stackrel{\_}{x^2}-{\stackrel{\_}{x}}^2& \left(7\right)\end{array}$ After separating the required point x.sub.1, on the basis of the average value, we find

(81) $\begin{array}{cc}{N}_{1,\mathrm{reg}}=\frac{\mathrm{nm}{x}_{\mathrm{edpt}}\left(x_1^2\left(\frac{n-1}{n}\right)-\frac{2x_1}{n}+-\frac{{}^2}{n}\right)}{x_1\left(x_{\mathrm{edpt}}\left(1-n\right)+\right)+x_{\mathrm{edpt}}+}& \left(8\right)\end{array}$ with =.sub.2.sup.nx.sub.i and =.sub.2.sup.nx.sub.i.sup.2. The subscript in N.sub.1,reg, here represents the acceptable noise for the regression criterion. This equation is used to calculate the permissible noise at point P.sub.1, if a maximum extrapolation error x.sub.edpt is provided. The requirement for the signal-to-noise ratio criterion is:

(82) $\begin{array}{cc}{N}_{1,\mathrm{snr}}=\frac{m\left(x_1-x_{\mathrm{edpt}}\right)}{\mathrm{SNR}}& \left(9\right)\end{array}$ Thus, the signal-to-noise ratio is the linear ratio of the amplitudes. The two requirements for the noise as defined in the equations (8) and (9) allow the required averaging time to be determined, if a noise amplitude density {dot over (N)} is known:

(83) $\begin{array}{cc}={\left(\frac{\stackrel{.}{N}}{N_1}\right)}^2& \left(10\right)\end{array}$ Since both conditions are to be satisfied simultaneously, the respective longer measuring time must be observed. Equalizing the two noise criteria results in a quadratic equation with the following solution:

(84) $\begin{array}{cc}{x}_1=-\frac{p}{2}-\sqrt{\frac{p^2}{4}-q}& \left(11\right)\\ \mathrm{with}& \\ p=-\frac{\left(2x_{\mathrm{edpt}}\mathrm{SNR}+-x_{\mathrm{edpt}}^2\left(1-n\right)\right)}{}& \left(12\right)\\ q=\frac{(x_{\mathrm{edpt}}\mathrm{SNR}\left(n-\right)+x_{\mathrm{edpt}}\left(x_{\mathrm{edpt}}+\right)}{}& \left(13\right)\\ =\left(x_{\mathrm{edpt}}\mathrm{SNR}+x_{\mathrm{edpt}}\right)\left(n-1\right)-& \left(14\right)\end{array}$

(85) It may be necessary to deviate from the recommendation in practice. It is thus known that the growth function for normal hearing persons no longer follows the ideal linear course in the semi-logarithmic representation below a level of approximately L.sub.2=25 dB SPL, since, strictly speaking, there is no distortion product threshold. If an excitation level is recommended in this area due to the process described above, the estimation accuracy is systematically impaired. It is possible to determine, for example, a minimum value for L.sub.2.sup.(1).

(86) Alternatively, a set of L2 noise levels can be deposited in the computer in a simplified process, said set to be used for every excitation frequency f2. It is further provided that a separate set of L2 noise levels be deposited for each excitation frequency f2.

(87) 2.3) Temporal Pulse Interlacing Procedure

(88) 2.3.1 Time-frequency Pulse Interlacing Procedure

(89) 2.3.1.1 Rigid Block Time-frequency Pulse Interlacing Procedure

(90) We are assuming that a panel with several pulse pairs of different frequencies f1, f2 and levels L1, L2 are presented in a block, and that, nevertheless, so-called ensembles are formed from, for example, 4 such blocks each in PTPV process (see, for example, Zelle et al., Extraction of otoacoustic distortion product sources using pulse basis functions, in J. Acoust. Soc. Am., 134(1): EL64-EL69, 07 20139), said ensembles allowing the extraction of the time signal of a desired distortion component, e.g., at fdp=2f1f2.

(91) We call this presentation mode a rigid block mode if the sequence of the pulse within a block is given and/or cannot be modified during the measurement.

(92) One possible approach is a delayed arrangement of different excitation frequencies f2(i), i=1, 2, 3, . . . n of n pulse pairs in a block. The block length is selected, for example, to T (and/or T.sub.B)=160 ms; the starting times of the pulse pairs are evenly distributed across the blockin this example, with n=4 and f2=1.5; 3; 2; 4 kHz. The main interest is then the extraction of the distortion component at fdp=2f1f2.

(93) FIG. 3 shows the envelopes 31 for four different f2 excitation pulses with the frequencies 1.5; 4; 2; 3 kHz for the rigid block time-frequency pulse interlacing procedures. Those envelopes are used to switch the excitation sounds on or off. The pulse form is represented in this image only by cosine-shaped ramps without a steady state. FIG. 3 shows a snapshot, as it were, of a measuring block for a panel.

(94) FIG. 4 shows the distribution of the four pulse pairs combined in a Panel A to the four slots of the block.

(95) FIG. 3b shows the envelopes 31 for four different f2 excitation impulses with the frequencies 1.5; 4; 2; 3 kHz, as well as the envelopes 32 for the four different f1 excitation pulses with the frequencies 1.25; 3.33; 1.67; 2.5 kHz corresponding in the pulse pairs for the rigid block time-frequency pulse interlacing procedure. FIG. 3b therefore shows a block with four pulse pairs that are presented in four slots (Slot 1 through Slot 4) of 40 ms duration each. As can be seen, the excitation frequencies f1 of the pulse pairs are switched off at the start of a slot and switched on again at the end of a slot. Due to the slot length and/or the switch-on period of the excitation pulses of the f1 frequencies, these each develop plateaus, as can be seen from their envelopes. The four different f2 excitation frequencies that are presented in four excitation pulses within the four slots are only switched on respectively once the f1 excitation pulses have been switched on. The f1 excitation pulses are only switched off once the f2 excitation pulses have been switched off and have faded. The processes of switching the individual excitation pulses f1 and f2 on and off are represented by cosine-shaped ramps. On the other hand, the plateaus for the f1 pulses do form a steady state that is between the cosine-shaped ramps of the processes of switching on and off. Block time T.sub.B in the depicted block consisting of four pulse pairs f1, f2 is 160 ms, corresponding to the sum of the four slot times T.sub.S (440 ms). As shown, the presentation of short f2 pulses (and/or the presentation of short pulses for at least one of the two sounds of the pulse pairs) allows the separation of two signal components in the time signal of the distortion product, which may lead to falsified measuring results, in case of the normal continuous presentation, through interference.

(96) Panel A of the four f2 pulse pairs is presented in a block of 160 ms, in which each f2 pulse pair takes up a slot of 40 ms. This block is repeated until a pre-selected signal-to-noise ratio (here also referred to as SNR) is reached for all four measured DPOAE.

(97) This selection of frequencies and their arrangement is appropriate for two reasons:

(98) A) Masking and Interference

(99) Frequencies in semi-octave steps, i.e., here, for slots i=1; 3 and i=2; 4, are arranged in the block at a maximum distance, since the frequency ratio f2/fdp is approx. 1.5. The consequence is that basically two interferences dominatein this example, for i=1;3: I) The presentation of the f2(i=3) pulse can mask the answer fdp(i=1) on the cochlea, if it has not yet faded sufficiently. This, however, applies only to the second source, which is often not in the foreground of the diagnostic procedure. II) The presentation of the f2(I=3) pulse can mask the f2(I=1) pulse on the cochlea, since they are only a minor third apart. This is also where, usually, the primary contribution to the DPOAE's that is more important for the diagnosis is affected. The time lag of T/2=80 ms here ensures that the interfering and/or masking signal components have usually faded to an acceptable level.
B) Time Needed

(100) In this rigid block process, one has to average until the frequency that delivers the worst signal-to-noise ration and/or the highest artifact ratio has reached the required minimum signal-to-noise ratio. In this example, this is the measurement at f2=1.5 kHz. Its noise background is, however, usually no more than 30% worse than the frequencies f2=2; 3; 4 kHz; so, the time loss is manageable.

(101) FIG. 5 shows the distribution of seven excitation frequencies f2 on two panels A and B in two blocks with four slots of 40 ms each. First, panel A is processed, i.e., for each f2, the average of the noise level Ldp of the respective DPOAE with the desired SNR is determined.

(102) Then, panel B is processed in the same way. The block with panel B also has a block duration of 160 ms each, but could be reduced to 120 ms, since the frequency spacings between the f2 and the fdp are of sufficient size.

(103) 2.3.1.2 Flexible Block Time-frequency Pulse Interlacing Procedure with Fixed Pulse Arrangement

(104) If the frequency area to be covered has a massive variation in signal-to-noise background or of artifact frequency, and/or if the process is to run perfectly and individually for each patient, the variability with respect to the required averaging time for the respective frequency f2 must be expected to be large.

(105) One solution for this problem is a flexible block method for the arrangement of pulse pairs. The frequencies used are, for example, f2=1; 1.5; 2; 3; 4; 6; 8 kHz. In normal conditions, the measurement at f2=1 kHz shall take four times longer than at f2=2; 3; 4 kHz, since the noise background is doubled, at least if one is aiming for the same noise background.

(106) There are two basic strategies for reacting to this situation. Firstly, one can accept the higher noise background. The consequence of this is that the estimation accuracy is less at lower frequencies, since this noise is taken into consideration over the course of the threshold approximation when the point that is closest to the threshold in the growth function is determined, and a higher extrapolation error is accepted. Secondly, one can try to provide a respectively longer measuring time for the frequency with the increased noise background.

(107) The frequencies are therefore divided into three panels A, B, and C that are each presented in a block with 4 slots: slot 1 permanently presents the pulse pair at f2=1 KHz for all three panels; slot 3 presents the pulse pair for of the time at f2=1.5 kHz, and the pulse pair at f2=2 kHz for the rest of the time; in slot 2 the pulse pair at f2=3 kHz is presented for the first half of the measuring time and the pulse pair at f2=8 kHz for the second half of the measuring time; and in slot 4 the pulse pair at f2=6 kHz is presented for the first half of the measuring time, and the pulse pair at f2=8 KHz for the second half of the measuring time.

(108) FIG. 6 shows the distribution of the seven f2 pulse pairs on the total of three panels A, B, and C. Once again, the panels A, B, and C are processed consecutively. In panel A, the average values for the Ldp at f2=3 kHz and f2=6 kHz are already determined with sufficient SNR. In panel B, the measurement is continued for f2=1 kHz, completed for f2=1.5 kHz, and started for f2=4 kHz and f2=8 kHz. The measuring time for panel A is greater than the one for panel B. In panel C, the measurement for f2=1 kHz, f2=4 kHz, and f2=8 kHz is completed and recorded and concluded for f2=2 kHz.

(109) In such an arrangement, the various noise and/or artifact conditions for the various excitation frequencies are taken into account. It is ensured that a presentation of an f2(i+1) a semi-octave higher than f2(i) never immediately follows the latter, and, additionally, the total measuring time in slots 2 and 4 is asymmetrically distributed on the three panels so that the higher frequencies, f2=6; 8 kHz have more averaging time assigned to them when the noise background in them rises slightly again.

(110) Here, the first excitation level is processed first, followed by the second excitation level in the same arrangement according to the processing of the threshold approximation, etc.

(111) This flexible block procedure may also work with less than an ideal outcome, especially if the DPOAE's are very diverse, i.e., there is a strong probability of a massive hearing impairment. A higher level for the pulse interlacing procedure can thus only be reached with a free, situation-adapted pulse arrangement.

(112) 2.3.1.3 Flexible Block Time-frequency Pulse Interlacing Procedure with Free Pulse Arrangement

(113) In this case, jobs are assigned to the measurement of growth functions according to the threshold approximation procedure. If DPOAE's are measured for seven frequencies f2 in semi-octave steps, seven jobs are processed accordingly. It is expedient to conduct the measurements with very similar excitation levels L2, to reduce masking issues. Consequently, all jobs start with the level that would be the one farthest from the threshold for a normal hearing person.

(114) However, those seven jobs are not necessarily started simultaneously. Rather, measurement starts with panel A. Panels B through D have not been determined yet; they are the result of the automatically conducted new allocation of a slot once the computer unit 15 has determined that a respective DPOAE for a pulse pair measured in panel A has a sufficient SNR.

(115) FIGS. 7 and 8 show an exemplary representation of how the new allocation takes place. The jobs are distributed to four panels A, B, C, and D with up to four slots in a block, according to FIG. 7. The current arrangement of jobsi.e., pulse pairs of different f2in the four slots is referred to as a panel and has to correspond to a set of criteria that, for example, take masking effects into consideration.

(116) Estimated runtimes are entered into a matrix for the slots. When the measurement starts (panel A), the estimated values are based upon empirical data of the population that dominates in terms of diagnostic tasks. All jobs are monitored during the presentation of the individual blocks. As soon as a job has reached its termination criteria (sufficient SNR or reaching the maximum presentation time), the estimated processing times for the slots are recalculated, and the freed slot is reallocated to a pulse pair that has yet to be processed.

(117) In the example, the start setting consists of the same arrangement as given above in 2.3.1.2, but is subsequently varied. For the start setting, it was assumed that measurements at f2=1 kHz require four times as much time as those at f22 kHz due to the double noise level, and those at f2=1.5 kHz still require twice the time.

(118) Then, the selected arrangement is almost ideal; theoretically, time is wasted in panel C, since the measurement would finish early, at 2 kHz. The change from panel B and C is crucial. Here, the number of slots was reduced to two, since the required distance of one octave is maintained for the two remaining frequencies so that the double number of pulses per time can be presented, compared to a 4-slot system. Alternatively, the same measuring time may be obtained, theoretically, by applying a consistent 2-slot system.

(119) If it becomes obvious after processing jobs 2 and 7 (f2=3 and/or 8 kHz) that they are clearly earlier, or that the jobs in the remaining slots are running with a delay, a check is conducted to see whether an as yet unprocessed job can be accepted. In this case, this could be job 3 (f2=2 kHz). If, however, it is processed, only job 1 remains in the end, so that, probably, no time can be saved. However, as soon as the jobs in slot 4 are also finished early, the switch to the 2-slot system can occur at an earlier point in time, and the measurement can be completed with an individual measurement at 1 kHz once job 3 (f2=2 kHz) has been processed.

(120) While the distribution of a total of seven f2 pulse pairs to four panels and four blocks with a variable number of slots under the simplified assumption of uniform measuring times for selected f2 pulse pairs is shown in FIG. 7 for the flexible block time-frequency pulse interlacing procedure with free pulse pair arrangement, FIG. 8 shows a modification in which no uniform measuring times are provided.

(121) As soon as a slot is freed because the average value of the Ldp was determined with sufficient SNR for the f2 measured there until that point, it is allocated to a new f2, resulting in a new panel. This may result in a reduced measuring time for the individual panels. This particularly applies to hearing-impaired patients, since the new procedure does not require any a priori assumptions regarding the expected SNR and/or the required measuring times of the respective DPOAE's.