CN112767970A - Abnormal lung sound detection method and system - Google Patents
Abnormal lung sound detection method and system Download PDFInfo
- Publication number
- CN112767970A CN112767970A CN202110088144.7A CN202110088144A CN112767970A CN 112767970 A CN112767970 A CN 112767970A CN 202110088144 A CN202110088144 A CN 202110088144A CN 112767970 A CN112767970 A CN 112767970A
- Authority
- CN
- China
- Prior art keywords
- sound signal
- signal
- lung sound
- original
- signals
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/66—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for extracting parameters related to health condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B7/00—Instruments for auscultation
- A61B7/003—Detecting lung or respiration noise
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B7/00—Instruments for auscultation
- A61B7/02—Stethoscopes
- A61B7/04—Electric stethoscopes
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/18—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being spectral information of each sub-band
Abstract
The invention discloses a method and a system for detecting abnormal lung sounds, wherein the method comprises the following steps: collecting original breath sound signals, and separating original lung sound signals from the original breath sound signals based on a wavelet transform method; inserting N Luo-Yin reference signals into the original lung sound signal to obtain a lung sound signal to be detected; performing time-frequency feature extraction on the lung sound signal to be detected to obtain an overall time-domain spectrogram signal corresponding to the lung sound signal to be detected; and identifying all peak information contained in the overall time domain spectrogram signal, and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information. In the embodiment of the invention, the abnormal lung sound signal characteristics are taken as a reference basis, and a signal processing technology is combined to provide a more accurate diagnosis result for preliminary automatic diagnosis research of the lung.
Description
Technical Field
The invention relates to the technical field of biomedical signal detection, in particular to an abnormal lung sound detection method and system.
Background
With the rapid development of modern medical technology, lung sound signal analysis has become a research hotspot gradually as one of the most important methods for diagnosing lung diseases. The lung sound signal is a sound signal generated in the respiratory process of human beings, and the change of the sound signal can directly reflect the physiological and pathological changes of the lungs. The traditional lung sound auscultation examination result is easily influenced by the subjectivity of a doctor, and the accuracy of the diagnosis result mainly depends on the clinical auscultation experience and the skilled degree of the skill of the doctor, so that the clinical requirement on the abnormal detection of the lung sound cannot be met, and further diagnosis and treatment are achieved.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a method and a system for detecting abnormal lung sounds, which can provide more accurate diagnosis results for preliminary automatic diagnosis research of lungs by taking the characteristics of abnormal lung sound signals as reference bases and combining a signal processing technology.
In order to solve the above problem, the present invention provides an abnormal lung sound detection method, including:
collecting original breath sound signals, and separating original lung sound signals from the original breath sound signals based on a wavelet transform method;
inserting N Luo-Yin reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
performing time-frequency feature extraction on the lung sound signal to be detected to obtain an overall time-domain spectrogram signal corresponding to the lung sound signal to be detected;
and identifying all peak information contained in the overall time domain spectrogram signal, and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information.
Optionally, the separating the original lung sound signal from the original breathing sound signal based on the wavelet transform method includes:
performing five-level wavelet decomposition on the original breathing sound signal based on the frequency range of the original breathing sound signal to form a five-level breathing sound signal;
determining a threshold line utilized by each layer of breathing sound signal based on all sampling data contained in each layer of breathing sound signal in the five layers of breathing sound signals;
based on a threshold line utilized by each layer of breathing sound signal, removing unstable sampling data contained in each layer of breathing sound signal;
and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
Optionally, the inserting N rhoone reference signals into the original lung sound signal to obtain the lung sound signal to be detected includes:
dividing the original lung sound signal into N sections of lung sound signals, and correspondingly inserting each of N pieces of Luo sound reference signals into the initial end of one section of lung sound signal in the N sections of lung sound signals to obtain a lung sound signal to be detected, namely N sections of secondary lung sound signals;
and recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
Optionally, the extracting time-frequency features of the lung sound signal to be detected to obtain an overall time-domain spectrogram signal corresponding to the lung sound signal to be detected includes:
performing short-time Fourier transform processing on each secondary lung sound signal in the N sections of secondary lung sound signals to obtain N time-frequency spectrogram signals;
and performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, and extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal.
Optionally, the identifying all peak information included in the overall time domain spectrogram signal, and determining the current state of the original lung sound signal according to the statistical number and the occurrence time of all peak information includes:
judging that the original lung sound signal is in a normal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is equal to N and the occurrence time of the N pieces of peak information is within the range of the original occurrence time;
judging that the original lung sound signal is in an abnormal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the range of the original occurrence time;
and returning to insert N Luo-Yin reference signals into the original lung sound signal to obtain the lung sound signal to be detected based on the fact that the statistical quantity of all peak value information contained in the overall time domain spectrogram signal is smaller than N.
In addition, an embodiment of the present invention further provides a system for detecting abnormal lung sounds, where the system includes:
the signal acquisition module is used for acquiring an original breath sound signal and separating an original lung sound signal from the original breath sound signal based on a wavelet transform method;
the signal processing module is used for inserting N Roots reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
the characteristic extraction module is used for extracting time-frequency characteristics of the lung sound signal to be detected to obtain an overall time domain spectrogram signal corresponding to the lung sound signal to be detected;
and the state identification module is used for identifying all peak information contained in the overall time domain spectrogram signal and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information.
Optionally, the signal acquisition module is configured to perform five-level wavelet decomposition on the original breathing sound signal based on a frequency range of the original breathing sound signal to form a five-level breathing sound signal; determining a threshold line utilized by each layer of breathing sound signal based on all sampling data contained in each layer of breathing sound signal in the five layers of breathing sound signals; based on a threshold line utilized by each layer of breathing sound signal, removing unstable sampling data contained in each layer of breathing sound signal; and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
Optionally, the signal processing module is configured to divide the original lung sound signal into N segments of lung sound signals, and correspondingly insert each of N compass sound reference signals into a start end of one of the N segments of lung sound signals to obtain a lung sound signal to be detected, that is, N segments of secondary lung sound signals; and recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
Optionally, the feature extraction module is configured to perform short-time fourier transform processing on each of the N segments of secondary lung sound signals to obtain N time-frequency spectrogram signals; and performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, and extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal.
Optionally, the state identification module is configured to determine that the original lung sound signal is in a normal state based on that it is identified that the statistical number of all peak information included in the overall time domain spectrogram signal is equal to N and the occurrence time of N peak information falls within the original occurrence time range; judging that the original lung sound signal is in an abnormal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the range of the original occurrence time; and returning to operate the signal processing module based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is smaller than N.
In the embodiment of the invention, the pre-denoising of the patient lung sound signal can be realized by utilizing a wavelet transform method, so that more accurate basic data is provided for the subsequent abnormal lung sound identification process; by taking the abnormal lung sound signal characteristics as a reference basis and combining a signal processing technology, a relatively accurate diagnosis result can be provided for preliminary automated diagnosis and research of the lung so as to assist a doctor in early warning the health condition of the lung of a patient.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of an abnormal lung sound detection method according to an embodiment of the present invention;
fig. 2 is a schematic composition diagram of an abnormal lung sound detection system according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
Referring to fig. 1, fig. 1 is a flow chart illustrating an abnormal lung sound detection method according to an embodiment of the present invention.
As shown in fig. 1, a method for detecting abnormal lung sounds, the method comprising:
s101, collecting original breath sound signals, and separating original lung sound signals from the original breath sound signals based on a wavelet transform method;
the implementation process of the invention comprises the following steps:
(1) based on the frequency range of the original breathing sound signal, five-level wavelet decomposition is carried out on the original breathing sound signal to form five-level breathing sound signals, which are specifically represented as follows:
based on the frequency range of the original breathing sound signal being 0-2000 Hz, performing first-layer decomposition on the original breathing sound signal to obtain a wavelet coefficient a1(frequency band of 0 to 1000Hz) and d1(the frequency band is 1000 Hz-2000 Hz); for the wavelet coefficient a1Performing second-layer decomposition to obtain wavelet coefficient a2(frequency band of 0 to 500Hz) and d2(the frequency band is 500 Hz-1000 Hz); for the wavelet coefficient a2Performing third-layer decomposition to obtain wavelet coefficient a3(frequency band of 0 to 250Hz) and d3(the frequency band is 250 Hz-500 Hz); for the wavelet coefficient a3Performing fourth-layer decomposition to obtain wavelet coefficient a4(frequency band of 0 to 125Hz) and d4(the frequency band is 125 Hz-250 Hz); for the wavelet coefficient a4Performing fifth-layer decomposition to obtain wavelet coefficient a5(frequency band of 0 to 63Hz) and d5(the frequency band is 63 Hz-125 Hz);
five layers of breath sound signals can be directly decomposed from the original breath sound signals, and the five layers of breath sound signals respectively comprise: the first layer of breathing sound signal (the frequency band is 1000 Hz-2000 Hz), the second layer of breathing sound signal (the frequency band is 500 Hz-1000 Hz), the third layer of breathing sound signal (the frequency band is 250 Hz-500 Hz), the fourth layer of breathing sound signal (the frequency band is 125 Hz-250 Hz) and the fifth layer of breathing sound signal (the frequency band is 63 Hz-125 Hz);
(2) based on all the sampling data contained in each layer of the five layers of breathing sound signals, determining a threshold line utilized by each layer of breathing sound signal, which is specifically represented as follows:
firstly, counting the number M of all sampling data contained in the j layer breathing sound signaljDetermining a fixed threshold T utilized for the layer j breathing sound signaljComprises the following steps:
secondly, based on the fixed threshold T utilized by the jth layer breathing sound signaljWavelet coefficient d corresponding to j-th layer breathing sound signaljThe comparison relationship between the first layer and the second layer, and the threshold line utilized for determining the jth layer breathing sound signal is as follows:
wherein, media _ j is the median of all the sampling data contained in the j-th layer breathing sound signal, sigmajIs the noise standard deviation of the j-th layer breathing sound signal, and alpha is the quantized coefficient (0)<α<1)；
(3) Based on a threshold line utilized by each layer of respiratory sound signal, removing unstable sampling data contained in each layer of respiratory sound signal, wherein the processing process is carried out according to the distinguishing characteristics of the heart sound signal and the lung sound signal;
(4) and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
S102, inserting N Roots reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
the specific implementation process comprises the following steps: firstly, dividing the original lung sound signal into N sections of lung sound signals, and correspondingly inserting each of N sections of the Luo-sound reference signals into the initial end of one section of the lung sound signals in the N sections of the lung sound signals to obtain lung sound signals to be detected, namely N sections of the secondary lung sound signals; secondly, recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
Dividing the original lung sound signal into N sections of lung sound signals to execute subsequent feature extraction and judgment, and improving the operation accuracy of the original lung sound signal; furthermore, a section of the luo-yin reference signal is added at the starting end of a section of the lung sound signal, and the characteristic value of the section of the luo-yin reference signal is used as a comparison on the basis, so that small peak interference information generated when the section of the lung sound signal is subjected to subsequent characteristic extraction can be eliminated.
S103, extracting time-frequency characteristics of the lung sound signal to be detected to obtain an overall time domain spectrogram signal corresponding to the lung sound signal to be detected;
the implementation process of the invention comprises the following steps:
(1) performing short-time Fourier transform processing on each secondary lung sound signal in the N sections of secondary lung sound signals to obtain N time-frequency spectrogram signals, wherein the ith section of secondary lung sound signal xi(t) corresponding time-frequency spectrogram signal Si(τ, f) is:
in the formula: tau is a time shift parameter, f is frequency, t is time, and W is a variable Gaussian window function;
(2) performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal, wherein the ith time-frequency spectrogram signal SiThe time domain spectrogram signal corresponding to (tau, f) is: si(t)＝∫Si(τ，f)df。
And S104, identifying all peak information contained in the overall time domain spectrogram signal, and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information.
In the implementation process of the present invention, all peak information contained in the overall time domain spectrogram signal is extracted by using a time domain peak method, and each extracted peak information carries a corresponding occurrence time, and at this time, the process of determining the original lung sound signal includes:
judging that the original lung sound signal is in a normal state if the N peak information are corresponding to the characteristic values of the N Luo-Yin reference signals based on the recognition that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is equal to N and the appearance time of the N peak information is within the range of the original appearance time;
based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the original occurrence time range, it is indicated that except that N pieces of peak information are corresponding to the characteristic values of the N pieces of Roots reference signals, other pieces of peak information are similar to the characteristic values of the N pieces of Roots reference signals, and then the original lung sound signal is judged to be in an abnormal state;
and returning to the re-operation step S102 if it is determined that an error occurs in the process of performing the feature identification on the original lung sound signal based on the fact that the statistical number of all peak information included in the overall time domain spectrogram signal is smaller than N.
In the embodiment of the invention, the pre-denoising of the patient lung sound signal can be realized by utilizing a wavelet transform method, so that more accurate basic data is provided for the subsequent abnormal lung sound identification process; by taking the abnormal lung sound signal characteristics as a reference basis and combining a signal processing technology, a relatively accurate diagnosis result can be provided for preliminary automated diagnosis and research of the lung so as to assist a doctor in early warning the health condition of the lung of a patient.
Examples
Referring to fig. 2, fig. 2 is a schematic diagram illustrating a composition of an abnormal lung sound detection system according to an embodiment of the present invention.
As shown in fig. 2, an abnormal lung sound detection system includes:
the signal acquisition module 201 is configured to acquire an original breath sound signal and separate an original lung sound signal from the original breath sound signal based on a wavelet transform method;
the implementation process of the invention comprises the following steps:
(1) based on the frequency range of the original breathing sound signal, five-level wavelet decomposition is carried out on the original breathing sound signal to form five-level breathing sound signals, which are specifically represented as follows:
based on the frequency range of the original breathing sound signal being 0-2000 Hz, performing first-layer decomposition on the original breathing sound signal to obtain a wavelet coefficient a1(frequency band of 0 to 1000Hz) and d1(the frequency band is 1000 Hz-2000 Hz); for the wavelet coefficient a1Performing second-layer decomposition to obtain wavelet coefficient a2(frequency band of 0 to 500Hz) and d2(the frequency band is 500 Hz-1000 Hz); for the wavelet coefficient a2Performing third-layer decomposition to obtain wavelet coefficient a3(frequency band of 0 to 250Hz) and d3(the frequency band is 250 Hz-500 Hz); for the wavelet coefficient a3Performing fourth-layer decomposition to obtain wavelet coefficient a4(frequency band of 0 to 125Hz) and d4(the frequency band is 125 Hz-250 Hz); for the wavelet coefficient a4Performing fifth-layer decomposition to obtain wavelet coefficient a5(frequency band of 0 to 63Hz) and d5(the frequency band is 63 Hz-125 Hz);
five layers of breath sound signals can be directly decomposed from the original breath sound signals, and the five layers of breath sound signals respectively comprise: the first layer of breathing sound signal (the frequency band is 1000 Hz-2000 Hz), the second layer of breathing sound signal (the frequency band is 500 Hz-1000 Hz), the third layer of breathing sound signal (the frequency band is 250 Hz-500 Hz), the fourth layer of breathing sound signal (the frequency band is 125 Hz-250 Hz) and the fifth layer of breathing sound signal (the frequency band is 63 Hz-125 Hz);
(2) based on all the sampling data contained in each layer of the five layers of breathing sound signals, determining a threshold line utilized by each layer of breathing sound signal, which is specifically represented as follows:
firstly, counting the number M of all sampling data contained in the j layer breathing sound signaljDetermining a fixed threshold T utilized for the layer j breathing sound signaljComprises the following steps:
secondly, based on the fixed threshold T utilized by the jth layer breathing sound signaljWavelet coefficient d corresponding to j-th layer breathing sound signaljThe comparison relationship between the first layer and the second layer, and the threshold line utilized for determining the jth layer breathing sound signal is as follows:
wherein, media _ j is the median of all the sampling data contained in the j-th layer breathing sound signal, sigmajIs the noise standard deviation of the j-th layer breathing sound signal, and alpha is the quantized coefficient (0)<α<1)；
(3) Based on a threshold line utilized by each layer of respiratory sound signal, removing unstable sampling data contained in each layer of respiratory sound signal, wherein the processing process is carried out according to the distinguishing characteristics of the heart sound signal and the lung sound signal;
(4) and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
The signal processing module 202 is configured to insert N rhoone reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
the specific implementation process comprises the following steps: firstly, dividing the original lung sound signal into N sections of lung sound signals, and correspondingly inserting each of N sections of the Luo-sound reference signals into the initial end of one section of the lung sound signals in the N sections of the lung sound signals to obtain lung sound signals to be detected, namely N sections of the secondary lung sound signals; secondly, recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
Dividing the original lung sound signal into N sections of lung sound signals to execute subsequent feature extraction and judgment, and improving the operation accuracy of the original lung sound signal; furthermore, a section of the luo-yin reference signal is added at the starting end of a section of the lung sound signal, and the characteristic value of the section of the luo-yin reference signal is used as a comparison on the basis, so that small peak interference information generated when the section of the lung sound signal is subjected to subsequent characteristic extraction can be eliminated.
The feature extraction module 203 is configured to perform time-frequency feature extraction on the lung sound signal to be detected, and acquire an overall time-domain spectrogram signal corresponding to the lung sound signal to be detected;
the implementation process of the invention comprises the following steps:
(1) performing short-time Fourier transform processing on each secondary lung sound signal in the N sections of secondary lung sound signals to obtain N time-frequency spectrogram signals, wherein the ith section of secondary lung sound signal xi(t) corresponding time-frequency spectrogram signal Si(τ, f) is:
in the formula: tau is a time shift parameter, f is frequency, t is time, and W is a variable Gaussian window function;
(2) performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal, wherein the ith time-frequency spectrogram signal SiTime domain spectrogram signal corresponding to (tau, f)Comprises the following steps: si(t)＝∫Si(τ，f)df。
The state identification module 204 is configured to identify all peak information included in the overall time domain spectrogram signal, and determine the current state of the original lung sound signal according to the statistical number and the occurrence time of all peak information.
In the implementation process of the present invention, all peak information contained in the overall time domain spectrogram signal is extracted by using a time domain peak method, and each extracted peak information carries a corresponding occurrence time, and at this time, the process of determining the original lung sound signal includes:
judging that the original lung sound signal is in a normal state if the N peak information are corresponding to the characteristic values of the N Luo-Yin reference signals based on the recognition that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is equal to N and the appearance time of the N peak information is within the range of the original appearance time;
based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the original occurrence time range, it is indicated that except that N pieces of peak information are corresponding to the characteristic values of the N pieces of Roots reference signals, other pieces of peak information are similar to the characteristic values of the N pieces of Roots reference signals, and then the original lung sound signal is judged to be in an abnormal state;
based on that the statistical number of all peak information included in the overall time domain spectrogram signal is identified to be less than N, which indicates that an error occurs in the process of performing feature identification on the original lung sound signal, the signal processing module 202 returns to re-run.
In the embodiment of the invention, the pre-denoising of the patient lung sound signal can be realized by utilizing a wavelet transform method, so that more accurate basic data is provided for the subsequent abnormal lung sound identification process; by taking the abnormal lung sound signal characteristics as a reference basis and combining a signal processing technology, a relatively accurate diagnosis result can be provided for preliminary automated diagnosis and research of the lung so as to assist a doctor in early warning the health condition of the lung of a patient.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer-readable storage medium, and the storage medium may include: read Only Memory (ROM), Random Access Memory (RAM), magnetic or optical disks, and the like.
The method and the system for detecting abnormal lung sounds provided by the embodiment of the invention are described in detail, a specific example is adopted in the text to explain the principle and the implementation of the invention, and the description of the embodiment is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. An abnormal lung sound detection method, comprising:
collecting original breath sound signals, and separating original lung sound signals from the original breath sound signals based on a wavelet transform method;
inserting N Luo-Yin reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
performing time-frequency feature extraction on the lung sound signal to be detected to obtain an overall time-domain spectrogram signal corresponding to the lung sound signal to be detected;
and identifying all peak information contained in the overall time domain spectrogram signal, and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information.
2. The abnormal lung sound detection method according to claim 1, wherein the separating the original lung sound signal from the original breathing sound signal based on the wavelet transform method comprises:
performing five-level wavelet decomposition on the original breathing sound signal based on the frequency range of the original breathing sound signal to form a five-level breathing sound signal;
determining a threshold line utilized by each layer of breathing sound signal based on all sampling data contained in each layer of breathing sound signal in the five layers of breathing sound signals;
based on a threshold line utilized by each layer of breathing sound signal, removing unstable sampling data contained in each layer of breathing sound signal;
and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
3. The abnormal lung sound detection method according to claim 2, wherein the inserting N Luo-Yin reference signals into the original lung sound signal to obtain the lung sound signal to be detected comprises:
dividing the original lung sound signal into N sections of lung sound signals, and correspondingly inserting each of N pieces of Luo sound reference signals into the initial end of one section of lung sound signal in the N sections of lung sound signals to obtain a lung sound signal to be detected, namely N sections of secondary lung sound signals;
and recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
4. The abnormal lung sound detection method according to claim 3, wherein the extracting the time-frequency feature of the lung sound signal to be detected to obtain the overall time-domain spectrogram signal corresponding to the lung sound signal to be detected comprises:
performing short-time Fourier transform processing on each secondary lung sound signal in the N sections of secondary lung sound signals to obtain N time-frequency spectrogram signals;
and performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, and extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal.
5. The abnormal lung sound detection method according to claim 4, wherein the identifying all peak information contained in the whole time domain spectrogram signal, and determining the current state of the original lung sound signal according to the statistical number and the occurrence time of all peak information comprises:
judging that the original lung sound signal is in a normal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is equal to N and the occurrence time of the N pieces of peak information is within the range of the original occurrence time;
judging that the original lung sound signal is in an abnormal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the range of the original occurrence time;
and returning to insert N Luo-Yin reference signals into the original lung sound signal to obtain the lung sound signal to be detected based on the fact that the statistical quantity of all peak value information contained in the overall time domain spectrogram signal is smaller than N.
6. An abnormal lung sound detection system, the system comprising:
the signal acquisition module is used for acquiring an original breath sound signal and separating an original lung sound signal from the original breath sound signal based on a wavelet transform method;
the signal processing module is used for inserting N Roots reference signals into the original lung sound signal to obtain a lung sound signal to be detected;
the characteristic extraction module is used for extracting time-frequency characteristics of the lung sound signal to be detected to obtain an overall time domain spectrogram signal corresponding to the lung sound signal to be detected;
and the state identification module is used for identifying all peak information contained in the overall time domain spectrogram signal and judging the current state of the original lung sound signal according to the statistical quantity and the occurrence time of all peak information.
7. The abnormal lung sound detection system of claim 6, wherein the signal acquisition module is configured to perform five-level wavelet decomposition on the original breathing sound signal based on the frequency range of the original breathing sound signal to form five-level breathing sound signals; determining a threshold line utilized by each layer of breathing sound signal based on all sampling data contained in each layer of breathing sound signal in the five layers of breathing sound signals; based on a threshold line utilized by each layer of breathing sound signal, removing unstable sampling data contained in each layer of breathing sound signal; and reconstructing the processed five-layer breath sound signals to obtain original lung sound signals.
8. The abnormal lung sound detection system of claim 7, wherein the signal processing module is configured to divide the original lung sound signal into N segments of lung sound signals, and insert each of N compass sound reference signals into a start end of one of the N segments of lung sound signals, so as to obtain a lung sound signal to be detected, that is, N segments of secondary lung sound signals; and recording the original occurrence time range of the N Luo-Yin reference signals contained in the lung sound signal to be detected.
9. The abnormal lung sound detection system of claim 8, wherein the feature extraction module is configured to perform short-time fourier transform processing on each of the N secondary lung sound signals to obtain N time-frequency spectrogram signals; and performing frequency domain integration processing on each time-frequency spectrogram signal in the N time-frequency spectrogram signals, and extracting N time-domain spectrogram signals corresponding to the N time-frequency spectrogram signals to form an integral time-domain spectrogram signal.
10. The abnormal lung sound detection system of claim 9, wherein the state identification module is configured to determine that the original lung sound signal is in a normal state based on identifying that a statistical number of all peak information included in the overall time domain spectrogram signal is equal to N and occurrence times of N peak information fall within the original occurrence time range; judging that the original lung sound signal is in an abnormal state based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is larger than N and the occurrence time of N pieces of peak information is within the range of the original occurrence time; and returning to operate the signal processing module based on the fact that the statistical quantity of all peak information contained in the overall time domain spectrogram signal is smaller than N.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088144.7A CN112767970A (en) | 2021-01-22 | 2021-01-22 | Abnormal lung sound detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088144.7A CN112767970A (en) | 2021-01-22 | 2021-01-22 | Abnormal lung sound detection method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112767970A true CN112767970A (en) | 2021-05-07 |
Family
ID=75705654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110088144.7A Pending CN112767970A (en) | 2021-01-22 | 2021-01-22 | Abnormal lung sound detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112767970A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100249611A1 (en) * | 2009-03-26 | 2010-09-30 | Edan Instruments. Inc. | Respiratory Signal Processing Method |
| CN106022258A (en) * | 2016-05-18 | 2016-10-12 | 成都济森科技有限公司 | Digital stethoscope and method for filtering heart sounds and extracting lung sounds |
| CN107798350A (en) * | 2017-11-08 | 2018-03-13 | 华南师范大学 | A kind of heart and lung sounds signal recognition methods and system |
| WO2018124173A1 (en) * | 2016-12-27 | 2018-07-05 | Ａｍｉ株式会社 | Biological monitoring device |
| CN108764184A (en) * | 2018-06-01 | 2018-11-06 | 广东工业大学 | A kind of separation method of heart and lung sounds signal, device, equipment and storage medium |
| CN111317499A (en) * | 2018-12-17 | 2020-06-23 | 天津光电通信技术有限公司 | Heart sound signal processing method based on wavelet technology |
-
2021
- 2021-01-22 CN CN202110088144.7A patent/CN112767970A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100249611A1 (en) * | 2009-03-26 | 2010-09-30 | Edan Instruments. Inc. | Respiratory Signal Processing Method |
| CN106022258A (en) * | 2016-05-18 | 2016-10-12 | 成都济森科技有限公司 | Digital stethoscope and method for filtering heart sounds and extracting lung sounds |
| WO2018124173A1 (en) * | 2016-12-27 | 2018-07-05 | Ａｍｉ株式会社 | Biological monitoring device |
| CN107798350A (en) * | 2017-11-08 | 2018-03-13 | 华南师范大学 | A kind of heart and lung sounds signal recognition methods and system |
| CN108764184A (en) * | 2018-06-01 | 2018-11-06 | 广东工业大学 | A kind of separation method of heart and lung sounds signal, device, equipment and storage medium |
| CN111317499A (en) * | 2018-12-17 | 2020-06-23 | 天津光电通信技术有限公司 | Heart sound signal processing method based on wavelet technology |
Non-Patent Citations (2)
| Title |
|---|
| 姚小静 等: "肺音信号分析及其识别方法的研究进展", 《重庆理工大学学报( 自然科学)》 * |
| 李真真，吴效明: "基于 S 变换的罗音信号检测算法", 《华南理工大学学报 ( 自 然 科 学 版)》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Windmon et al. | Tussiswatch: A smart-phone system to identify cough episodes as early symptoms of chronic obstructive pulmonary disease and congestive heart failure | |
| US9198634B2 (en) | Medical decision support system | |
| US8690789B2 (en) | Categorizing automatically generated physiological data based on industry guidelines | |
| CN110970042B (en) | Pulmonary ralated artificial intelligence real-time classification method, system and device of electronic stethoscope and readable storage medium | |
| CN110731778B (en) | Method and system for recognizing breathing sound signal based on visualization | |
| CN112949639B (en) | Congenital heart disease heart sound intelligent segmentation classification algorithm, device and storage medium | |
| Yamashita et al. | Robust classification between normal and abnormal lung sounds using adventitious-sound and heart-sound models | |
| CN113057648A (en) | ECG signal classification method based on composite LSTM structure | |
| CN111317499A (en) | Heart sound signal processing method based on wavelet technology | |
| Shukla et al. | An efficient heart sound segmentation approach using kurtosis and zero frequency filter features | |
| Abramov et al. | Information system for diagnosis of respiratory system diseases | |
| Abdollahpur et al. | Cycle selection and neuro-voting system for classifying heart sound recordings | |
| CN111938691B (en) | Basic heart sound identification method and equipment | |
| JP2023531464A (en) | A method and system for screening for obstructive sleep apnea during wakefulness using anthropometric information and tracheal breath sounds | |
| Zhong et al. | Heart murmur recognition based on hidden Markov model | |
| CN110507299B (en) | Heart rate signal detection device and method | |
| Jin et al. | Automatic wheeze detection using histograms of sample entropy | |
| CN112767970A (en) | Abnormal lung sound detection method and system | |
| Paul et al. | Noise reduction for heart sounds using a modified minimum-mean squared error estimator with ECG gating | |
| CN116246654A (en) | Breathing sound automatic classification method based on improved Swin-transducer | |
| Ari et al. | On a robust algorithm for heart sound segmentation | |
| Zia et al. | Noise detection and elimination for improved acoustic detection of coronary artery disease | |
| Ozkan et al. | Improved segmentation with dynamic threshold adjustment for phonocardiography recordings | |
| Aras et al. | Endpoint detection of lung sounds for electronic auscultation | |
| Courtemanche et al. | Heart sound segmentation based on mel-scaled wavelet transform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |