Optimized otoacoustic reflection signal detection methodTechnical Field
The invention relates to the technical field of objective hearing detection based on otoacoustics, in particular to an optimized method for detecting reflected signals of otoacoustics.
Background
OAE (auditory evoked response) signals refer to audio signals spontaneously generated from within the external auditory meatus, which signals are often closely related to auditory function. Thus, accurate detection and analysis of OAE signals is of great importance for assessing auditory health. However, conventional OAE signal detection methods face many challenges in practical applications due to environmental noise and equipment limitations. Existing OAE signal detection methods mainly depend on the performance of filters, and different types of filters with different orders can lead to different results. In addition, in low signal-to-noise environments, it is also a great difficulty to accurately identify and quantify the linear relationship between two signals. These problems limit the accuracy and reliability of OAE signal detection. For this purpose, we propose an optimized otoacoustic reflection signal detection method.
Disclosure of Invention
The invention aims to provide an optimized otoacoustic reflection signal detection method so as to solve the problems in the background technology.
In order to achieve the above purpose, the invention provides the following technical scheme that the optimized otoacoustic reflection signal detection method comprises the following steps:
S11, eliminating a noise threshold, balancing detection accuracy and detection duration through a self-defined control program calculation process, setting parameters including threshold upper and lower limits, peak feature numbers, similarity weight ratio columns and a plurality of judgment standard parameter settings conforming to the weight numbers, and reserving better and reliable data blocks through signal preprocessing for next calculation;
S12, artifact elimination comprises the steps of filtering in the first 4.2ms, data length filling and band-pass filtering, wherein a data processing mode of a T-shaped window is combined with a filtering technology, coherent average method calculation is carried out, linear artifacts are filtered, and the signal to noise ratio of TEOAE is improved;
s13, performing similarity calculation, namely performing frequency domain conversion calculation, performing frequency domain spectrum interference calculation, performing group signal similarity calculation, performing signal matching by a coherent spectrum method, and enhancing quantitative evaluation of signal linearity under the condition of low signal-to-noise ratio;
S14, obtaining a reference signal with low bottom noise through wavelet decomposition and wavelet reconstruction, and extracting signal key frequency points according to signal properties for TEOAE signal-to-noise ratio calculation.
S15, continuous tone processing is achieved through data period interception, fourier transformation and audio breakpoint removal, noise interference is reduced, signal characteristics are highlighted, then frequency amplitude values at the third-order intermodulation points 2f1-f2 are detected and calculated, and average amplitude values of the three points about 2f1-f2 are calculated, so that the signal to noise ratio of the DPOAE is obtained.
S16, generating a sweep frequency signal and a sweep frequency detection threshold value, accurately identifying the type of the small cavity, and improving the identification precision of the small cavity.
Preferably, the S11 noise threshold removal algorithm includes the following steps:
S21, adopting a T-shaped window as a window function, and performing window function calculation on 1253 data lengths;
S22, filtering the stimulation signal of 4.2 milliseconds before the signal, and supplementing the 1254-2048 signal;
S23, removing noise interference outside a 500-6000HZ range by a band-pass filtering method;
s24, calculating coherent average of the signals to remove linear artifacts and improve the signal to noise ratio of the TEOAE.
Preferably, the step S12 is implemented by an artifact removal algorithm, and removes linear artifacts of the samples.
Preferably, the step S13 is implemented by similarity calculation, and a frequency domain coherent spectrum method is used to calculate the similarity between the two signals, and identify and quantify the linear relationship between the two signals.
Preferably, the S14 applies a specific decomposition order to stationary and non-stationary signals through wavelet noise reduction, and eliminates noise signals.
Preferably, the wavelet denoising and similarity calculation comprises the following steps:
S31, carrying out wavelet 3-order decomposition and reconstruction on the audio data to obtain a noise-reduced signal;
S32, performing frequency domain conversion on the reconstructed signal;
s33, calculating coherent spectrums of the two groups of signals in a frequency domain;
s34, calculating the similarity between the two signals according to the result of the coherent spectrum.
Preferably, the step S15 is implemented by DP continuous tone data processing and a characteristic signal detection algorithm, and each breakpoint is removed to change the audio into continuous tone, thereby reducing the overall rise of the background noise caused by the breakpoint and improving the accuracy and reliability of DP data processing.
Preferably, the DP detection algorithm comprises the steps of:
S41, removing one distortion point signal from each 1025 data lengths can reduce the frequency amplitude of the position of the power spectrum background noise highlighting third-order signal and improve the signal to noise ratio;
s42, detecting the frequency amplitude at the positions 2f1-f2, and calculating the strength of the signal.
S43, calculating the average amplitude of the three points about 2f1-f2 to obtain the signal to noise ratio.
Preferably, the step S16 is implemented by a small cavity identification algorithm, and the algorithm calculates the smoothness and the energy distribution interval of the data according to the integrity of the single sweep frequency test data, and distinguishes and identifies the type of the small cavity and the state of the earplug.
Preferably, the small cavity recognition algorithm comprises the steps of:
s51, generating a sweep frequency signal of 20-6000 (Hz) and sounding;
S52, calculating the smoothness and the energy distribution interval of the received data according to the distribution condition of the sweep frequency signals;
s53, setting different types of threshold ranges of the small cavities according to the distribution interval of the data.
Compared with the prior art, the invention has the beneficial effects that:
(1) The detection accuracy and the detection duration are weighed through a calculation process of a custom control program.
(2) The signal is first filtered using a set T-shaped window as a window function.
(3) The method adopts a frequency domain coherent spectrum method to calculate the similarity between two signals, can effectively identify and quantify the linear relation between the two signals, and can calculate the similarity even under the condition of low signal-to-noise ratio. In addition, frequency domain coherence spectroscopy can more effectively distinguish true correlations from occasional, nonsensical correlations.
(4) The frequency domain transformation is carried out on the noise-reduced audio data, so that the peak point of the signal can be found in a part of frequency bands. The signal to noise ratio calculation can be performed as 5 signal frequency points.
(5) The continuity of signal data is guaranteed by removing the break points, the audio frequency background noise is reduced, the reliability of signals is improved, the signal to noise ratio can be calculated better, and the detection efficiency of the ear sound is improved.
(6) The frequency sweep detection signal sound pressure level energy distribution interval has obvious distinction, a reasonable threshold range can be divided, and small cavity interference and partial space third-order intermodulation interference are eliminated.
Drawings
FIG. 1 is a flow chart of an implementation method of the present invention;
FIG. 2 is a graph of DPOAE embodiment case test results;
fig. 3 is a graph of TEOAE embodiment test results.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-3, the present invention provides a technical solution, an optimized otoacoustic reflection signal detection method, comprising the following steps:
S11, eliminating a noise threshold, balancing detection accuracy and detection duration through a self-defined control program calculation process, setting parameters including threshold upper and lower limits, peak feature numbers, similarity weight ratio columns and a plurality of judgment standard parameter settings conforming to the weight numbers, and reserving better and reliable data blocks through signal preprocessing for next calculation;
S12, artifact elimination comprises the steps of filtering in the first 4.2ms, data length filling and band-pass filtering, wherein a data processing mode of a T-shaped window is combined with a filtering technology, coherent average method calculation is carried out, linear artifacts are filtered, and the signal to noise ratio of TEOAE is improved;
s13, performing similarity calculation, namely performing frequency domain conversion calculation, performing frequency domain spectrum interference calculation, performing group signal similarity calculation, performing signal matching by a coherent spectrum method, and enhancing quantitative evaluation of signal linearity under the condition of low signal-to-noise ratio;
S14, obtaining a reference signal with low bottom noise through wavelet decomposition and wavelet reconstruction, and extracting signal key frequency points according to signal properties for TEOAE signal-to-noise ratio calculation.
S15, continuous tone processing is achieved through data period interception, fourier transformation and audio breakpoint removal, noise interference is reduced, signal characteristics are highlighted, then frequency amplitude values at the third-order intermodulation points 2f1-f2 are detected and calculated, and average amplitude values of the three points about 2f1-f2 are calculated, so that the signal to noise ratio of the DPOAE is obtained.
S16, generating a sweep frequency signal and a sweep frequency detection threshold value, accurately identifying the type of the small cavity, and improving the identification precision of the small cavity.
As a preferred embodiment, the S11 noise threshold removal algorithm includes the following steps:
S21, adopting a T-shaped window as a window function, and performing window function calculation on 1253 data lengths;
S22, filtering the stimulation signal of 4.2 milliseconds before the signal, and supplementing the 1254-2048 signal;
S23, removing noise interference outside a 500-6000HZ range by a band-pass filtering method;
s24, calculating coherent average of the signals to remove linear artifacts and improve the signal to noise ratio of the TEOAE.
As a preferred embodiment, the step S12 is implemented by an artifact removal algorithm, and removes linear artifacts of the samples.
As a preferred embodiment, the step S13 is implemented by similarity calculation, and a frequency domain coherence spectrum method is used to calculate the similarity between two signals, and identify and quantify the linear relationship between the two signals.
As a preferred embodiment, the S14 applies a specific decomposition order for stationary and non-stationary signals by wavelet noise reduction, excluding noise signals.
As a preferred embodiment, the wavelet denoising and similarity calculation includes the steps of:
S31, carrying out wavelet 3-order decomposition and reconstruction on the audio data to obtain a noise-reduced signal;
S32, performing frequency domain conversion on the reconstructed signal;
s33, calculating coherent spectrums of the two groups of signals in a frequency domain;
s34, calculating the similarity between the two signals according to the result of the coherent spectrum.
As a preferred implementation manner, the step S15 is realized through DP continuous sound data processing and a characteristic signal detection algorithm, and each breakpoint is removed so that the audio frequency becomes continuous sound, thereby reducing the overall elevation of the background noise caused by the breakpoint and improving the accuracy and reliability of DP data processing.
As a preferred embodiment, the DP detection algorithm comprises the steps of:
S41, removing one distortion point signal from each 1025 data lengths can reduce the frequency amplitude of the position of the power spectrum background noise highlighting third-order signal and improve the signal to noise ratio;
s42, detecting the frequency amplitude at the positions 2f1-f2, and calculating the strength of the signal.
S43, calculating the average amplitude of the three points about 2f1-f2 to obtain the signal to noise ratio.
As a preferred embodiment, the step S16 is implemented by a small cavity identification algorithm, which calculates the smoothness and energy distribution interval of the data according to the integrity of the single sweep test data, and distinguishes and identifies the type of small cavity and the state of the earplug.
As a preferred embodiment, the small cavity recognition algorithm comprises the steps of:
s51, generating a sweep frequency signal of 20-6000 (Hz) and sounding;
S52, calculating the smoothness and the energy distribution interval of the received data according to the distribution condition of the sweep frequency signals;
s53, setting different types of threshold ranges of the small cavities according to the distribution interval of the data.
The working principle of the invention is that the TEOAE algorithm is optimized, firstly, through collecting data and dynamically setting proper upper and lower limits of signal threshold values, periodic sample values and smooth step sizes, the peak characteristic number exceeding the noise threshold value is calculated and identified, useful signals and noise signals are effectively distinguished, important audio data are reserved, and then a T-shaped window function is applied to initialize the signals.
Eliminating the data of the first 4.2ms to avoid the influence of the stimulus signal on the result, then supplementing the data with different lengths, removing the signal outside the specific frequency range by band-pass filtering, calculating by a coherent average method to filter out linear artifacts and improve the signal to noise ratio of the TEOAE test, and finally calculating the similarity and the signal to noise ratio of the data after wavelet noise reduction and displaying the detection data.
The DPOAE time domain data processing comprises the steps of firstly generating a specific sweep frequency signal according to a preset threshold value, calculating the power spectrum density and peak energy output of an echo signal to distinguish different test environments, intercepting the obtained original sound data according to the period if the preset threshold value is met, carrying out Fourier transform, identifying and removing an audio breakpoint through an algorithm, converting the data into continuous sound, reducing noise interference introduced by the breakpoint, clearly highlighting signal characteristics, and finally combining the frequency amplitude calculation of third-order intermodulation points, focusing on analyzing key frequency points such as 2f1-f2, calculating the sound pressure level of signals and noise, and evaluating the signal-to-noise ratio of the DPOAE.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.