CN111820908B - Probe detector for abnormal ear sound cavity and detection method thereof - Google Patents

Abstract

The invention provides a probe detector of an otoacoustic abnormal cavity, which is characterized in that an electroacoustic transducer and an electroacoustic transducer for receiving sound are arranged in a probe shell, and a detected cavity is connected with the shell through a transmitting and receiving sound pipeline; a pulse stimulation sound system is arranged on the shell. The detection method is that the probe emits basic stimulating sound, and the stimulating short sound is played through a pulse stimulating sound system; the generated basic stimulation data are stored in a head file of the otoacoustic detection engineering by simulation software, so that the acquisition of samples, the processing of sample data, the statistics of stability indexes of the transmitted wave and the received wave and the judgment of probe indexes of the tested cavity are realized. The invention realizes the dynamic automatic calibration of the earphone probe by using a digital signal processing technology and a machine learning method, thereby achieving the purpose of distinguishing air from an abnormal coupling cavity; the detection of the probe in the air, the detection of the metal coupling cavity (0.2 cc to 0.7 cc) and the detection of the human auditory canal and the rough otoacoustic probe microphone calibration are automatically distinguished by a program.

Description

Probe detector for abnormal ear sound cavity and detection method thereof

Technical Field

The invention relates to a medical hearing detection system, in particular to a probe detector for an abnormal otoacoustic cavity and a detection method thereof.

Background

Otoacoustic emissions are an acoustic energy generated in the cochlea, through the ossicular chain and tympanic membrane conduction and released into the external auditory canal (Kemp, 1986). The sound intensity of otoacoustic emission signals is weak and does not lead to hearing awareness of the person, whereas high sensitivity microphones placed in the external auditory meatus can record sounds in the auditory meatus. Otoacoustic emission detection is an objective hearing detection method starting from infants, and usually, after a weak magnetic pole signal is given to cochlea, a probe, a transceiver and the like are used for detecting the weak acoustic energy, so that the purpose of hearing detection is achieved.

At present, few products with independent intellectual property rights are provided for domestic infant otoacoustic screening equipment, and most of the current markets use foreign product otoacoustic hearing screening equipment. The foreign products have larger difference in accuracy in clinical use, but the actual effect of the methods is obvious for microphone calibration schemes or acoustic measurement compensation methods which mainly use coupler transfer functions for abnormality detection of air and metal coupling cavities, but the treatment of products which are screened singly by depending on pressure sensors or additional external microphones and sound sources is complicated in actual calibration. In the process of detecting the ear sound, since the infant ear sound signal is very weak and is particularly easy to be influenced by environmental noise or in-ear body noise, whether the probe is normally placed or not and whether the probe is placed in the ear or not greatly influence the acquisition and analysis of the signal. The current ear sound probe detection technology mostly relies on the impedance of the detected cavity in the calibration compensation of the microphone, but in practical ear sound detection application, the impedance of the detected cavity is generally difficult to accurately obtain. And the hardware technology of the equipment has high cost, the calibration error is large under the condition of poor circuit noise control, and meanwhile, the technology implementation difficulty is relatively large for simple screening products.

Disclosure of Invention

Aiming at the defects existing in the existing otoacoustic detection, the invention aims to provide the probe detector for the abnormal cavity of the otoacoustic and the detection method thereof, and the digital signal processing technology and the machine learning method are used for realizing the dynamic automatic calibration of the earphone probe on the basis of not increasing the hardware cost, so that the aim of distinguishing the air and the abnormal coupling cavity is fulfilled. The detection of the probe in the air, the detection of the metal coupling cavity (0.2 cc to 0.7 cc) and the detection of the human auditory canal and the rough otoacoustic probe microphone calibration are automatically distinguished by a program.

In order to solve the technical problems, the invention adopts the following technical scheme: the probe detector of the abnormal cavity of sound, including the body of the probe, set up electroacoustic ring energy device and acoustic-electric transducer used for receiving the sound in the said body, the cavity to be measured is connected with body through transmitting and receiving the sound pipeline; a pulse stimulation sound system is arranged on the shell.

The pulse stimulus sound system is a simulation software Matlab2015 system, the simulation software Matlab2015 generates a pulse stimulus sound with a short sound (click) of 100 mu s, the intensity M is a stimulus pulse with 83dB, 25ms is selected at intervals, the energy is concentrated at 0.5-6kHz, and the sound construction function is that:

wherein the basic stimulus waveform is y (t) =sin (2pi×24000 t), t is more than or equal to 0 and less than or equal to 100 μs, the time is from 0, 602 values are calculated every other time, and the stimulus waveform is a group of stimulus waveforms; intensity M is assumed to be a fixed size; the generated basic stimulation data is stored in a header file of the otoacoustic detection project by using simulation software.

The detection method of the invention comprises the following steps:

placing a probe detector, starting a test, and entering a probe detection program;

the probe emits basic stimulus sounds;

playing the short stimulation sound through the pulse stimulation sound system; the generated basic stimulation data are stored into a header file of the otoacoustic detection project by simulation software, so that sample collection is realized;

processing the sample data, and calculating the stimulation parameters:

the noise level is calculated by receiving the reflected sound, and the Hamming window function is adopted to carry out filtering treatment

Wherein, alpha is 0.46, N is 602, the function is used for processing out the unnecessary noise, the fast FFT conversion is carried out in the effective frequency range, the energy response of the measured cavity of the reference frequency section is calculated respectively, and the frequency point is respectively: the energy values of 1khz,2khz,3khz,4khz,5khz are taken as the characteristic value α1, and the number of peaks occurring in the waveform of 6ms to 15ms of the time domain part is taken as the characteristic value α2;

fifthly, in the process of data acquisition and calculation of returned sound energy values, stability indexes of the transmitted wave and the received wave are counted;

according to the steps, whether the characteristic parameters alpha 1 and alpha 2 and the stability index in the step are satisfied: the characteristic parameter alpha 1 has 5 data parameter threshold settings, and the DP setting 65 judges the probe index of the tested cavity according to the requirements of the DP test or the TE test; if yes, the stimulus sound is emitted to meet the normal test condition, and normal test is carried out; and if the characteristic parameters and the stability indexes do not meet the test requirements, dynamically adjusting the probe, and turning to the step I.

The stability index in the detection method step is as follows:

stability of stimulus sound: the similarity between each emission wave and the previous emission wave is more than 90%;

stability of the received wave: the similarity between the received complete waveform and the waveform received last time is more than 80%;

similarity: euclidean distance using two sets of discrete data

To calculate the similarity, where N is a set of received wave lengths, where 602, data1, data2 are stimulus sounds or received reflected sound data, respectively, twice before and after.

According to the probe detector and the detection method of the acoustic anomaly cavity, which are designed by the technical scheme, short sound is selected for probe detection, short-time Fourier transform is carried out on sound signals received by the earphone according to the environment conditions of the test, and then characteristic extraction and classification detection are carried out on signals in different frequency domains. The dynamic automatic calibration of the earphone probe is realized by using a digital signal processing technology and a machine learning method on the basis of not increasing hardware cost, so that the aim of distinguishing air from an abnormal coupling cavity is fulfilled. The invention utilizes the program to automatically distinguish the detection of the probe in the air and the metal coupling cavity from the detection of the human auditory canal and the rough calibration of the auditory probe microphone. The invention has low hardware technology cost, small calibration error under the condition of poor circuit noise control, and small technology realization difficulty for simple screening products.

Drawings

FIG. 1 shows a schematic diagram of the probe detector of the otoacoustic anomaly cavity of the present invention;

FIG. 2 is a schematic diagram showing the structure of the probe detector of the present invention during detection;

FIG. 3 is a schematic flow chart of the detection method of the present invention.

Detailed Description

The probe detector and the detection method for the otoacoustic abnormal cavity are specifically described below with reference to the accompanying drawings.

The probe detector for the acoustic anomaly cavity comprises a probe shell 1, wherein an electroacoustic transducer 2 (micro-loudspeaker) and an electroacoustic transducer 3 (micro-microphone) are arranged in the shell 1, and the electroacoustic transducer 3 is used for receiving sound, as shown in fig. 1 and 2. The tested cavity 5 is connected with the shell 1 through the transmitting and receiving sound pipeline 4, and 0.1-0.7cc is taken in the experimental process of the tested cavity 5. A pulse stimulation sound system is arranged in the shell 1, so that the probe detects and selects the structure of short sound. The pulse stimulus sound system adopts simulation software Matlab2015, the simulation software Matlab2015 generates a pulse stimulus sound with a short sound (click) of 100 mu s, the intensity M is a stimulus pulse with 83dB, 25ms is selected as the interval between the stimulus sound system and the stimulus sound system, the energy is concentrated at 0.5-6kHz, and the sound construction function is that:

wherein the basic stimulus waveform is y (t) =sin (2pi×24000 t), t is more than or equal to 0 and less than or equal to 100 μs, the time starts from 0, 602 values are calculated every other time, and the stimulus waveform is a group of stimulus waveforms; intensity M is assumed to be a fixed size; the generated basic stimulation data is stored in a header file of the otoacoustic detection project by using simulation software. According to the invention, the earphone channel is selected to circularly transmit the short sound signal, and after time delay is 1 second, the microphone is started to receive the sound signal.

The detection method of the present invention, see fig. 3, comprises the steps of:

(1) Placing a probe detector, starting a test, and entering a probe detection program;

(2) Transmitting basic stimulus sound by the probe;

playing the short stimulation sound through the pulse stimulation sound system; the generated basic stimulation data are stored into a header file of the otoacoustic detection project by simulation software, so that sample collection is realized;

processing the sample data, and calculating the stimulation parameters:

the noise level is calculated by receiving the reflected sound, and the Hamming window function is adopted to carry out filtering treatment

Wherein, alpha is 0.46, N is 602, the function is used for processing out the unnecessary noise, the fast FFT conversion is carried out in the effective frequency range, the energy response of the measured cavity of the reference frequency section is calculated respectively, and the frequency point is respectively: the energy values of 1khz,2khz,3khz,4khz,5khz are taken as the characteristic value α1, and the number of peaks occurring in the waveform of 6ms to 15ms of the time domain part is taken as the characteristic value α2;

fifthly, in the process of data acquisition and calculation of returned sound energy values, stability indexes of the transmitted wave and the received wave are counted;

wherein, the stability index is:

stability of stimulus sound: the similarity between each emission wave and the previous emission wave is more than 90%;

stability of the received wave: the similarity between the received complete waveform and the waveform received last time is more than 80%;

similarity: euclidean distance using two sets of discrete data

To calculate the similarity, where N is a set of received wave lengths, where 602, data1, data2 are stimulus sounds or received reflected sound data, respectively, twice before and after.

According to the steps, whether the characteristic parameters alpha 1 and alpha 2 and the stability index in the step are satisfied: the characteristic parameter alpha 1 has 5 data parameter threshold settings, and the DP setting 65 judges the probe index of the tested cavity according to the requirements of the DP test or the TE test; if yes, the stimulus sound is emitted to meet the normal test condition, and normal test is carried out; and if the characteristic parameters and the stability indexes do not meet the test requirements, dynamically adjusting the probe, and turning to the step I.

Claims (2)

1. The detection method of the probe detector of the abnormal cavity of the ear sound is characterized in that the probe detector of the abnormal cavity of the ear sound comprises a probe shell, an electroacoustic transducer and an electroacoustic transducer for receiving sound are arranged in the shell, and the detected cavity is connected with the shell through a transmitting and receiving sound pipeline; a pulse stimulation sound system is arranged on the shell, the pulse stimulation sound system is a simulation software Matlab2015 system, the simulation software Matlab2015 generates a pulse stimulation sound with a short sound (click) of 100 mu s, the intensity M is a stimulation pulse with 83dB, 25ms is selected from the interval, the energy is concentrated at 0.5-6kHz, and the sound construction function is that:

wherein the basic stimulus waveform is y (t) =sin (2pi×24000 t), t is more than or equal to 0 and less than or equal to 100 μs, and the time is from 0, and 602 values are calculated as a group of stimulus waveforms; using simulation software to store the generated basic stimulation data into a head file of the otoacoustic detection project;

the detection method comprises the following steps:

(1) Placing a probe detector, starting a test, and entering a probe detection program;

(2) Transmitting basic stimulus sound by the probe;

(3) Playing the short stimulation sound through a pulse stimulation sound system; the generated basic stimulation data are stored into a header file of the otoacoustic detection project by simulation software, so that sample collection is realized;

(4) Processing the sample data, calculating a stimulation parameter:

the noise level is calculated by receiving the reflected sound, and the Hamming window function is adopted to carry out filtering treatment

Wherein, alpha is 0.46, N is 602, the function is used for processing out the unnecessary noise, the fast FFT conversion is carried out in the effective frequency range, the energy response of the measured cavity of the reference frequency section is calculated respectively, and the frequency point is respectively: the energy values of 1khz,2khz,3khz,4khz,5khz are taken as the characteristic value α1, and the number of peaks occurring in the waveform of 6ms to 15ms of the time domain part is taken as the characteristic value α2;

(5) In the process of data acquisition and calculation of the returned sound energy value, the stability indexes of the transmitted wave and the received wave are counted;

(6) According to the characteristic parameters alpha 1 and alpha 2 of the step (4), whether the stability index in the step (5) meets the following conditions: the characteristic parameter alpha 1 has 5 data parameter threshold settings, and the DP setting 65 judges the probe index of the tested cavity according to the requirements of the DP test or the TE test; if yes, the stimulus sound is emitted to meet the normal test condition, and normal test is carried out; and if not, dynamically adjusting the probe, and turning to the step (2) until the characteristic parameters and the stability index meet the test requirements.

2. The method for detecting a probe detector for an abnormal ear sound cavity according to claim 1, wherein the stability index in the step (5) of the detection method is as follows:

stability of stimulus sound: the similarity between each emission wave and the previous emission wave is more than 90%;

stability of the received wave: the similarity between the received complete waveform and the waveform received last time is more than 80%;

similarity: euclidean distance using two sets of discrete data

To calculate the similarity, where N is a set of received wave lengths, where 602, data1, data2 are stimulus sounds or received reflected sound data, respectively, twice before and after.

Images