CN1263423C - Method and device for determining respiratory system condition by using respiratory system produced sound - Google Patents
Method and device for determining respiratory system condition by using respiratory system produced sound Download PDFInfo
- Publication number
- CN1263423C CN1263423C CNB011102497A CN01110249A CN1263423C CN 1263423 C CN1263423 C CN 1263423C CN B011102497 A CNB011102497 A CN B011102497A CN 01110249 A CN01110249 A CN 01110249A CN 1263423 C CN1263423 C CN 1263423C
- Authority
- CN
- China
- Prior art keywords
- respiratory system
- parameter
- sound
- situation
- send
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The present invention relates to a digital signal processing method and a device for determining the conditions of respiratory systems by applying sound. The method comprises the procedures: an audio signal formed according to sound sent by a respiratory system is received; the audio signal is sampled; the signal is ranged; a group of parameters are calculated from the ranged signal; the group of parameters is compared with a group of modules which can represent the condition of the respiratory system; one condition of the respiratory system is determined to be selected according to the result of the comparison procedure; finally, output which comprises a possible condition is provided.
Description
The present invention relates to the processing of a kind of data and digital signal, and particularly relate to the device that a kind of sound that uses the respiratory system situation to send determines the respiratory system situation.
The situation of respiratory system has good file to record and narrate, for instance, herein as " diagnosis of child respiratory diseases and the processing " of reference, G.M.Loughlin and H.Eigen, 1994.Most breathing sound all is undisturbedly, so become audible the time when breathing sound, will cause patient and family members' thereof care usually.For example stridulate, asthma, snoring and pulmonary block.Have the patient of outstanding severe chronic bronchitis, asthma or fibroid cyst can hear another kind of noise on one's body at some in addition, this noise overcast and the picture brattle at the incoming tide, outside the several chis of this patient, just can hear.
Asthma is a kind of seriality, almost as canorous sound, can hear when exhaling usually.Asthma is that air causes disturbance to produce because of flow restriction in large-scale and medium-sized intrathoracic airway.Gas flow rate by the small vent road is not enough to cause asthma.Stridulating is a kind of kenel of asthma, more loud and coarse usually, general mainly hear when air-breathing.It results from usually, and endotracheal blockage causes ever-increasing air agitation outside the thoracic cavity.Snoring is the sound that causes owing to the vibrations of throat soft tissue.Its normally a kind of overcast, the gutturophony that can only when air-breathing, hear; Except some extreme situations, can only hear in bed usually.Pulmonary blocks normally patient and is used for generalized a kind of saying.For different people, the different meaning of its representative, and it is used to describe the nonspecific sound that teenage sufferer caused by secretions that day by day increases and the main airway that slightly dwindles are arranged usually.
There is the patient of pulmonary sound can look for the internist to seek advice from.In this consulting, the first step is exactly to find the reason that produces noise, gives a suitable suggestion then and comes mitigation symptoms.Yet the option that can be used for assessing pulmonary sound is very few.In the situation of majority, appropriate assessment needs front and back and thoracic cavity, side X-ray photograph.Since under a lot of situations, the source of this noise is benign, patients just often will accept unnecessary radiograph outside unnecessary consulting so.
The objective of the invention is to propose a kind of device that is used for analyzing the respiratory system situation, it can be avoided fully because the restriction of known technology and the various problems that shortcoming is caused.
Some in the other features and advantages of the present invention can be exposed in the ensuing narration, and wherein some can obviously be learnt from narration, perhaps can obtain from embodiments of the invention.In order to reach above-mentioned and other advantage, the invention provides the method that noise that a kind of basis produced by certain situation is analyzed the respiratory system situation.The method comprising the steps of: the parameter of calculating the digital signal that the noise produced by unacknowledged situation still forms; Come relatively this parameter and the referrer module of representing some known condition in the respiratory system to calculate the probability consistent with referrer module; And whether the similarity of using this parameter and referrer module decide this parameter to coincide with this referrer module, if coincide, then selects this known condition to come as this still unacknowledged respiratory system situation.
In said method, further can comprise the following step respectively in each step: the step that receives the acoustical signal of expression respiratory system sound; This acoustical signal is converted to the step of digital signal; Comprise the step of this digital signal of format at least one block in the conversion of this acoustical signal, this block is made up of general 25 milliseconds numerical data and general 10 milliseconds overlapped data substantially; The step of the known condition of this respiratory system for output is provided, and the referrer module of this situation and this parameter have an acceptable similarity by contrast; Store and the step of the corresponding referrer module of this respiratory system known condition in a storage device; Calculate this parameter step and comprise the step of calculating a cheek coefficient of frequency from this digital signal; Calculate this cheek coefficient of frequency step and comprise that use one anti-high speed fourier transform forms this parameter; Calculate this parameter step and comprise about 39 parameters of calculating; This method also comprises the step of utilizing concealed Marko's module to calculate this referrer module; Relatively this parameter step comprises and utilizes a Viterbi decoder to calculate similarity between this parameter and this referrer module; Calculation procedure comprises one group of parameter calculating this sound; And comparison step comprises relatively this swarm parameter and one group of referrer module representing one group of known condition.
The present invention also provides a kind of sound that utilizes respiratory system to send to detect the device of respiratory system situation in addition, comprising: format the device of the digital signal of this sound at least one block; Calculate the device of a parameter from the formative digital signal of this sound; Utilize concealed Marko's module to calculate the device of this referrer module; To calculate the similarity between this parameter and referrer module, come the relatively device of the referrer module of this parameter and this representative respiratory system known condition; And utilize similarity between this parameter and this referrer module, determine whether this parameter is consistent with this referrer module, if select the device of this known condition as the situation of this respiratory system.
Analyze in the device of respiratory system situation at the noise that above-mentioned basis of the present invention is produced by certain situation, be preferably wherein and can also comprise the following device: the device that receives the acoustical signal of expression respiratory system sound; Change the device that this acoustical signal is a digital signal; This acoustical signal conversion equipment comprises that also this digital signal of format is the device of at least one block, and this block is made up of general 25 milliseconds numerical data and general 10 milliseconds overlapped data substantially; The device of the known condition of this respiratory system for output is provided, and the referrer module of this situation and this parameter have an acceptable similarity by contrast; Store device with the corresponding referrer module of this respiratory system known condition; The device that calculates this parameter also comprises the device that calculates a cheek coefficient of frequency from this digital signal; The device that calculates this cheek coefficient of frequency also comprises the device that uses an anti-high speed fourier transform to form this parameter; The device that calculates this parameter also comprises the device that calculates about 39 parameters; Utilize concealed Marko's module to calculate the device of this referrer module; Relatively the device of this parameter also comprises the device that utilizes Viterbi decoder to calculate the similarity between this parameter and this referrer module; Accountant comprises the device of one group of parameter calculating this sound of expression; And comparison means comprises the relatively device of one group of referrer module of one group of known condition of this group parameter and expression.
The present invention further provides the brain of can powering to read and comprised the medium of computer software, when this software is carried out on computers, the instrument that can make computer become to provide the noise of producing according to certain unknown situation to analyze the unknown situation of this respiratory system.This instrument has comprised the device that is still calculated parameter by expression in the format digital signal of the noise produced of unacknowledged situation; Come relatively this parameter and the device of representing the referrer module of some known condition in the respiratory system to calculate with the similarity of referrer module; And whether the similarity of using this parameter and referrer module decide this parameter to coincide with this referrer module, if coincide, then selects this corresponding known condition to come device as this still unacknowledged respiratory system situation.
Adopt the noise of producing by certain situation of the present invention to analyze the device of respiratory system situation, the patient is reduced when carrying out medical advice receive unnecessary and harmful radiograph, thereby be useful the health of human body.
It should be noted that above narration of doing and the ensuing both of being described in detail just demonstrate and also help the explanation, be not to a kind of restriction of the present invention.
Below in conjunction with accompanying drawing and one embodiment of the present of invention, by following narration purpose of the present invention, advantage and principle are done one and explain in more detail.
Shown in Figure 1 is human respiratory system's model;
The block chart of the demonstration system of a decision human respiratory system situation that is according to the present invention to be set up shown in Figure 2;
Shown in Figure 3 is a flow chart, and what this figure described is the step that relevant noise from a situation gained according to the present invention decides the respiratory system situation;
Shown in Figure 4 is a flow chart, and what this figure described is the relevant step of calculating the gained parameter according to the present invention;
Shown in Figure 5 is a flow chart, and this figure describes is about the step that at least one parameter and at least one referrer module are compared according to the present invention; And
Shown in Figure 6 is a flow chart, and what this figure described is the relevant step of selecting a kind of situation according to the present invention.
The preferred embodiment that will describe in detail now according to the present invention and come, these examples illustrate in the accompanying drawings.Wherein will use same reference number to part consistent or similar among the figure.
Shown in Figure 1 is human respiratory system's model 100.As can be seen from Figure 1, this human respiratory system 100 has comprised lung 105, vocal cords 110, trunnion 120, pharyngeal cavity 125, soft palate 127, tongue 129, nasal cavity 130 and oral cavity 140.It is typical cases in the respiratory system situation that respiratory system 100 creates unique sound 150.What table 1 was cited is according to the present invention, the situation that the respiratory system that can be diagnosed is possible.
Though a trained people may be able to distinguish the sound that some produce according to certain situation, most people is very difficult to only just can the diagnosis situation according to sound.The sound of the representative respiratory system situation that receives via utilization, the invention provides one can point out that respiratory system may situation diagnosis output, and in many examples, can eliminate the demand that unnecessary X-ray irradiation or medicine are consulted the merchant.
The situation of table 1. respiratory system
| 1 anaphylaxis | 19 hemosiderosis infringement pulmonary |
| 2 anaphylaxis bronchus and lung Koji Fungi disease | 20 exist foreign body (I) |
| 3 blood vessel lymph gland edema | 21 exist foreign body (II) |
| 4 breathing-function of deglutition are bad | 22 exist foreign body (III) |
| 5 asthma (I) | 23 burst pseudomembrane laryngitis |
| 6 asthma (II) | 24 times glottis is narrow |
| 7 fibroid cysts | 25 respiratory tracts are narrow |
| 8 epiglottitises | 26 tracheomalacias |
| 9 cysts of epiglottis | 27 psychic shocks |
| 10 foreign body obstructions | 28 blood vessels ring (vascular ring) |
| 11 ejection air-flow adverse currents | 29 viral bronchiolitises |
| The 12 throats capsule that swells, hemangioma | 30 vocal cords dysfunctions |
| 13 larynx trachea and bronchitis | 31 just send out snoring |
| 14 larynx esophageal intubation glands break | 32 obstacle sleep apneas |
| 15 larynxes are softening | 33 rhinophonia-choana is narrow |
| 16 larynx emulsus tumors | 34 rhinophonia-nasopharyngitis |
| 17 mediastinum node or blocks | 35 Bronchio-s are maked a whistling sound and are breathed heavily |
| 18 pharyngeal ulcer | 36 bronchitis (infectiousness is irritated blocks) |
Shown in Figure 2 is the example system 200 of institute's construction according to the present invention, this system situation that decides a human respiratory system.As seeing in Fig. 2, this system 200 has comprised 220, one storage modules 230 of 210, one data processors of 205, one analog/digitals of an input module (A/D) modular converter and an output module 240.
At the input module 205 shown in Fig. 2 system 200, reception is analyzed this situation from the sound 150 of the situation generation of respiratory system 100.This input module 205 can by various can replace easily and finish provide the device that is input to this analog to digital modular converter 210 to constitute.For instance, these device some of them have comprised networking interface module, the digital interface module and the mike of general utility functions.Though the A/D modular converter shown in the image pattern 2 can be used in some occasion, if sound 150 had just converted digital pattern to before being provided to system 200, that has not just needed the A/D modular converter.For the people who is familiar with this skill, conspicuous this data processor may comprise for example some extra hardware, software with and/or similar one group of CPU, auxiliary processor, memorizer, suitable devices such as the device of depositor and other processing and system.Storage module 230 can be made up of one group of changeable element or subsystem, for instance, comprises hard disk drive, optical disc apparatus, and the storage device of general utility functions, removable storage device and other can be used for finishing the device of memory function.Output module 230 can be made up of one group of changeable system, subsystem and element, for instance, comprises display device 242 and printer 244.Display device 242 can be made up of one group of changeable device, for instance, comprises cathode ray tube or liquid crystal display.In addition, this output device 240 can comprise other systems that output function can be provided, subsystem and device, for example digital interface of networking interface, storage device, general utility functions, parallel interface and serial line interface.In addition, this system 200 with and/or arbitrary above-mentioned module, device, element or function can be included in the random suitable combination of hardware, software or other suitable devices, and can be used as a standard autonomous system or integrate with other devices.
What Fig. 3 represented is a flow chart, is decorated with one group above about utilizing the sound 150 that is produced by certain situation to decide the correlation step of program 300 of the situation of respiratory system 100.As shown in Figure 3, this program has comprised the step that system 200 implements among Fig. 2.Program 300 starts from system 100 and produces the sound 150 that is transfused to the module reception.This input equipment 205 converts sound 150 to one group of analoging sound signal.
For this acoustical signal being converted to one group of formative digital signal, this A/D modular converter receives this analoging sound signal (step 310), this analoging sound signal that receives is taken a sample (step 320), and this analoging sound signal is converted to digital signal (step 330).In the embodiment of a demonstration, about 20 seconds sampling during this A/D modular converter 210 has carried out analoging sound signal with the sampling rate of 16000 samplings of about per second.In this example embodiment, this analoging sound signal of 20 seconds has produced one group of digital signal of enough making correct analysis via conversion.Yet inferior analoging sound signal may need more digital signal (that is to say, faster between sampling rate or longer sampling date).Though Fig. 3 has represented to receive one group of analoging sound signal (step 310), this system 200 can accept to represent the digital signal of the analoging sound signal of this situation easily via beginning from step 335.
In order to produce this formative digital signal, this data processor 220 has been handled this digital signal (step 335).A kind of suitable format digital signal structure can provide the most excellent processing performance and precise diagnosis.Though the structure of other kinds or form may be used, but in this example embodiment, this data processor 220 cuts into a series of block (step 335) with this digital signal, during each block be 25 milliseconds, the overlapped data that is inserted with 10 milliseconds between the consecutive block is to provide smooth data kenel.
This data processor 220 uses this formative digital signal to calculate at least one parameter (step 340), and Fig. 4 is the flow chart of a demonstration, and drawn is the step of coming the program 400 of calculating parameter (step 340) according to the present invention.As shown in Figure 4,220 pairs of formative digital signal blocks of this data processor carry out high speed fourier transform (step 410).The frequency domain value that obtains is through providing one group of coefficient as this at least one parameter to number conversion (step 420) and once anti-high speed fourier transform (FFT-1) (step 430).These coefficients are called cheek coefficient of frequency (mel-frequency) again.Though this example embodiment has been used the cheek coefficient of frequency, we still can calculate these coefficients with other technology, for instance, comprise time frequency band output (sub-band outputs), linear spectral increases the weight of linear predictor coefficient (perceptually weighted linearprediction) to (linespectrum pairs) and aesthesia.In the speech recognition field, we are " basis of speech recognition ", L.Rabiner and B.Juang, 1993 are added in this as reference, have wherein described the cheek coefficient of frequency, inferior frequency band output, linear spectral to and aesthesia increase the weight of the calculating of linear predictor coefficient.
Again with reference to figure 3, some these group coefficients are used to be used as this at least one parameter, and this parameter is brought referrer module that those that may experience with at least one expression system 100 in the storage module 230 that is stored in may breath states compare (step 350).
This calculation of parameter (step 340) produces at least one parameter (at least one value just).This data processor 220 needs compare with at least one referrer module (step 350) of at least one parameter.For comparing (step 350) required optimal parameter number is unfixed.If used parameter very little, then the final diagnosis that is provided by output module 130 will be compared inaccuracy.Opposite, if used too much parameter, then required processing and store to become and be difficult to burden in data processor 220.In an example embodiment, 39 parameters (39 values just) and this at least one referrer module are compared.
Shown in Figure 5 is an example flow chart, this figure describes be comprise relevant to the present invention, the program 500 of the secondary program that at least one parameter and at least one referrer module are compared.According to Fig. 5, this data processor 220 usefulness are calculated this at least one parameter and the probability of this at least one referrer module (step 520) or the value (step 510) of similarity and are made comparisons (step 350 of Fig. 3), and the value of this probability or similarity is exactly the similarity degree of this referrer module (step 520) and this at least one parameter of calculating from this formative digital signal.The value of the similarity that this calculates provides a coherent detection of the similarity degree between this at least one parameter and this corresponding referrer module.In an example embodiment of the present invention, this data processor 220 has used a Viterbi decoder (Viterbi Decoder) to calculate similar probability (step 510).In an example embodiment, this data processor 220 repeats (step 530) and is stored in 36 kinds of storage module 230 and calculates the value (step 510) of a probability or similarity with reference to module arbitrary, wherein these 36 kinds with reference to module each corresponding to the situation that is listed in 36 kinds of respiratory systems in the table 1.Though this embodiment has used Viterbi decoder, we still can use other technology to come relatively this at least one parameter and this at least one referrer module, for instance, comprise Dynamic Time Warping method (dynamic timewarping method), vector quantization (vector quantization), or artificial neuron networking (artificial neural networks).In the speech recognition field, " basis of speech recognition ", L.Rabiner and B.Juang, 1993, narrated Veterbi decoding (Viterbi Decode), Dynamic Time Warping, vectorial quantization and artificial neuron networking.
Again with reference to figure 3, in one embodiment of the invention, this data processor 220 has used a known condition training sample as the respiratory system that will analyze, calculate this respiratory system 100 of expression one to know a referrer module of situation.This training sample is made up of the sound that the known condition of this respiratory system produces substantially.In order to form this referrer module, this A/D modular converter converts this training sample to digital signal, and this data processor 220 with this digital signals formatization so that a formative digital signal to be provided, and calculate at least one parameter, make the corresponding referrer module (step 330 of the known case that will be diagnosed, 335,340).As a result, this referrer module has comprised at least one parameter, and this at least one parameter is corresponding to the training sample that is retained in the known condition that will be diagnosed in the storage module 230.For any other known condition that will be diagnosed, these data processor 220 uses one are used in other training samples of these other known condition, to calculate other referrer module of these other respiratory system known condition of expression.
In one embodiment of the invention, this storage device 230 has comprised the situation of being differentiated and diagnosing according to each, at least one referrer module of using concealed Marko's module (Hidden Markov model) to calculate.This data processor 220 utilizes the training sample of each respiratory system situation to calculate concealed Marko's module.For instance, this data processor 220 has utilized the format digital signal that obtains according to this training sample, calculates the referrer module of this situation corresponding to concealed Marko's module." basis of speech recognition ", L.Rabiner and B.Juang, 1993, narrated the concealed Marko's module that applies in the speech recognition.
With reference to figure 3, in order to determine whether this formative digital signal matches with this referrer module, 220 decisions (step 360) of this data processor are corresponding to this at least one parameter of one of this referrer module, and whether perhaps one group of parameter is in an acceptable similar value scope.In more detail, this data processor 220 is from being stored in storage module 230, makes a decision and selects (step 360) among respiratory system 100 situations of representing with at least one referrer module.As the some that determines and select a possibility situation, before an output is provided, 220 selections of this data processor, or say and prefer that the most similar situation can not be accepted or dissimilar situation and filter out.
What Fig. 6 represented is the example flow chart of a program 600, and this program has comprised the relevant step of selecting a kind of situation according to the present invention.With reference to figure 6, in an example embodiment according to the present invention, this data processor 220 uses the step that is illustrated in the program 600 to decide and choice situation (step 360).This data processor 220 has used the probability that receives from this comparison step (step 350) to assess the corresponding probability of each situation (step 610).This data processor 220 has been selected the value (step 620) of five maximum likelihoods.If the value (step 630) of an acceptable probability should five values of maximum be arranged, then this data processor 220 just provides this situation to come when output (step 650,660).If any in these five situations all can not be represented the real conditions of this problem, this data processor 220 is just only selected those acceptable situations (step 640), and provides a suitable output (step 650,660) to output module 240.This output can include the detailed pharmaceutical information (step 660) of selecting situation at this (a bit).As what in flow chart, mention, output can by this output module with one easily pattern be shown to a user (step 650).
With reference to figure 3, this output module 240 provides the indication of a situation to a user (step 370) again.In an example embodiment, this output module has comprised that for instance, a display 242 and a printer 244 are to provide the detailed medical information (step 660) of a user corresponding to this situation.
Therefore, the invention provides the situation that new device is diagnosed respiratory system.
Preferred embodiment of the present invention has disclosed as above, and purpose is expression and narrates its target.This is not the accurate form of wanting thorough detailed descriptionthe or limiting the invention to this exposure, and via above teaching enlightenment or implementation of the present invention, may be able to make and revising and variation.For instance, the present invention is not limited to the device shown in Fig. 1, but generally can the whole bag of tricks implements, scope definition of the present invention the scope of claims with and equivalence among.
Anyly have the knack of this operator, without departing from the spirit and scope of the present invention, should be able to make various changes and modification.Detailed description here and the example usefulness of all just demonstrating, and the real protection domain of the present invention should be as the criterion with the scope of accompanying claims.
Claims (13)
1. a sound that utilizes respiratory system to send detects the device of respiratory system situation, comprising:
Format the device of the digital signal of this sound at least one block;
Calculate the device of a parameter from the formative digital signal of this sound;
Utilize concealed Marko's module to calculate the device of this referrer module;
To calculate the similarity between this parameter and referrer module, come the relatively device of the referrer module of this parameter and this representative respiratory system known condition; And
Utilize the similarity between this parameter and this referrer module, determine whether this parameter is consistent with this referrer module, if select the device of this known condition as the situation of this respiratory system.
2. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that also comprising the device of the acoustical signal that receives expression respiratory system sound.
3. the sound that utilizes respiratory system to send as claimed in claim 2 detects the device of respiratory system situation, it is characterized in that also comprising that this acoustical signal of conversion is the device of digital signal.
4. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that this block is made up of general 25 milliseconds numerical data and general 10 milliseconds overlapped data substantially.
5. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that this calculating parameter device is that formative digital signal is carried out the high speed fourier transform, to obtain a frequency domain value, again this frequency domain value is carried out number conversion and anti-high speed fourier transform, with the device of the cheek coefficient of frequency that obtains to provide at least one this parameter.
6. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that this calculating parameter device be utilize time frequency band output, linear spectral to or aesthesia increase the weight of in the method for linear predictor coefficient any, to obtain the device of this parameter.
7. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that comprising the device of known condition for exporting that this respiratory system is provided, the referrer module of this situation and this parameter have an acceptable similarity by contrast.
8. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that also comprising the device that stores with the corresponding referrer module of this respiratory system known condition.
9. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that the device that calculates this parameter also comprises the device that calculates a cheek coefficient of frequency from this digital signal.
10. the sound that utilizes respiratory system to send as claimed in claim 9 detects the device of respiratory system situation, it is characterized in that the device that calculates this cheek coefficient of frequency comprises the device that uses an anti-high speed fourier transform to form this parameter.
11. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that relatively the device of this parameter also comprises the device that utilizes Viterbi decoder to calculate the similarity between this parameter and this referrer module.
12. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that this comparison means is any that utilize in the methods such as Dynamic Time Warping method, vectorial quantization or artificial neural network, to calculate the device of this similarity.
13. the sound that utilizes respiratory system to send as claimed in claim 1 detects the device of respiratory system situation, it is characterized in that:
Accountant comprises the device of one group of parameter calculating this sound of expression; And
Comparison means comprises the relatively device of one group of referrer module of one group of known condition of this group parameter and expression.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011102497A CN1263423C (en) | 2001-04-04 | 2001-04-04 | Method and device for determining respiratory system condition by using respiratory system produced sound |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011102497A CN1263423C (en) | 2001-04-04 | 2001-04-04 | Method and device for determining respiratory system condition by using respiratory system produced sound |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1377629A CN1377629A (en) | 2002-11-06 |
| CN1263423C true CN1263423C (en) | 2006-07-12 |
Family
ID=4658450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB011102497A Expired - Fee Related CN1263423C (en) | 2001-04-04 | 2001-04-04 | Method and device for determining respiratory system condition by using respiratory system produced sound |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1263423C (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103479335B (en) * | 2013-09-27 | 2016-02-17 | 深圳市倍轻松科技股份有限公司 | The recommend method that sound differentiates and system |
| US20170127977A1 (en) * | 2014-06-26 | 2017-05-11 | Koninklijke Philips N.V. | Non-invasive monitoring of pulmonary conditions |
| CN113729725B (en) * | 2021-09-30 | 2024-02-06 | 东南大学 | Method for extracting respiratory waveform from electrocardiosignal based on power spectrum first moment and smoother |
-
2001
- 2001-04-04 CN CNB011102497A patent/CN1263423C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1377629A (en) | 2002-11-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7457753B2 (en) | Telephone pathology assessment | |
| Zañartu et al. | Subglottal impedance-based inverse filtering of voiced sounds using neck surface acceleration | |
| AU2019356224B2 (en) | Estimating lung volume by speech analysis | |
| US11344225B2 (en) | Determining apnea-hypopnia index AHI from speech | |
| Sapienza et al. | Respiratory and laryngeal measures of children and women with bilateral vocal fold nodules | |
| US20220225930A1 (en) | Apnea monitoring method and apparatus | |
| WO2019216320A1 (en) | Machine learning apparatus, analysis apparatus, machine learning method, and analysis method | |
| Romero et al. | Deep learning features for robust detection of acoustic events in sleep-disordered breathing | |
| JP7291319B2 (en) | Evaluation method and apparatus for difficult airway based on speech technique by machine learning | |
| Vijayan et al. | Throat microphone speech recognition using mfcc | |
| Cohen et al. | Acoustic transmission of the respiratory system using speech stimulation | |
| CN1263423C (en) | Method and device for determining respiratory system condition by using respiratory system produced sound | |
| Lulich et al. | Semi-occluded vocal tract exercises in healthy young adults: Articulatory, acoustic, and aerodynamic measurements during phonation at threshold | |
| JP2023517175A (en) | Diagnosing medical conditions using voice recordings and listening to sounds from the body | |
| TW483747B (en) | Method and device for detecting status of breathing system by voice | |
| Howard | Speech fundamental period estimation using pattern classification | |
| Liu et al. | Breath Sounds Recognition and Classification for Respiration System Diseases | |
| JP7099615B2 (en) | State estimator, state estimation method, and program | |
| CN113158916B (en) | Cough sound automatic identification algorithm, device, medium and equipment | |
| CN115804568A (en) | Intelligent respiration signal processing method and device | |
| Malone et al. | Acoustic analysis of infantile stridor: a review | |
| CA1300267C (en) | Method and apparatus for determining articulatory parameters from speech data | |
| Harvill | Regularization for dysarthric speech recognition and telemedicine applications | |
| Soumya et al. | Hybrid Algorithm To Classify The Lung Abnormalities Using The Stethoscope Recordings | |
| Sivakumar et al. | Autoregressive Based Vocal Tract Shape Modelling of Vowels in Speech Processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060712 |