CN108323158A - Heart sound identification method and cloud system - Google Patents
Heart sound identification method and cloud system Download PDFInfo
- Publication number
- CN108323158A CN108323158A CN201880000118.0A CN201880000118A CN108323158A CN 108323158 A CN108323158 A CN 108323158A CN 201880000118 A CN201880000118 A CN 201880000118A CN 108323158 A CN108323158 A CN 108323158A
- Authority
- CN
- China
- Prior art keywords
- heart sound
- data
- heart
- sound data
- network
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000000605 extraction Methods 0.000 claims description 30
- 238000004590 computer program Methods 0.000 claims description 18
- 230000002159 abnormal effect Effects 0.000 claims description 16
- 230000000747 cardiac effect Effects 0.000 claims description 16
- 230000000153 supplemental effect Effects 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 9
- 210000005036 nerve Anatomy 0.000 claims description 8
- 238000005192 partition Methods 0.000 claims description 8
- 230000007246 mechanism Effects 0.000 claims description 6
- 230000036541 health Effects 0.000 abstract description 10
- 238000012544 monitoring process Methods 0.000 abstract description 7
- 230000003862 health status Effects 0.000 abstract description 6
- 238000003745 diagnosis Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 10
- 238000013527 convolutional neural network Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 230000002612 cardiopulmonary effect Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 4
- 208000019622 heart disease Diseases 0.000 description 3
- 238000012952 Resampling Methods 0.000 description 2
- 206010003119 arrhythmia Diseases 0.000 description 2
- 230000006793 arrhythmia Effects 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 102100033620 Calponin-1 Human genes 0.000 description 1
- 102100033591 Calponin-2 Human genes 0.000 description 1
- 102100033592 Calponin-3 Human genes 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 101000945318 Homo sapiens Calponin-1 Proteins 0.000 description 1
- 101000945403 Homo sapiens Calponin-2 Proteins 0.000 description 1
- 101000945410 Homo sapiens Calponin-3 Proteins 0.000 description 1
- 208000037656 Respiratory Sounds Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005194 fractionation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Classifications
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The application provides a heart sound identification method and a cloud system, wherein the method comprises the following steps: identifying the collected sound data to obtain an identification result; and if the identification result is the heart sound data, identifying the heart sound data based on the heart sound frequency to obtain the heart sound type of the heart sound data. This application can be through telling the present health status of user to and whether the person of facilitating the use decides need doctor's remote help etc. according to the health status of self, realizes the diagnosis of the state of an illness of user, and health monitoring's intellectuality, and the recognition accuracy is preferred simultaneously.
Description
Technical field
This application involves intelligent medical technical fields, the more particularly to recognition methods of heart sound and cloud system.
Background technology
With internet+and portable medical technology development, intelligent medical equipment is more and more, stethoscope as doctor most
One of common clinical diagnosis tool, intelligence are that the research and development of intelligent medical equipment are preferred.But it in the prior art, intelligently listens
The intelligence degree for examining device is relatively low, only can be by recording to cardiopulmonary sound after, by doctor by terminal device to using
The cardiopulmonary sound of person carries out condition-inference either health monitoring or by the way that cardiopulmonary sound is uploaded to high in the clouds to realize doctor couple
The diagnosis of the state of an illness, and the monitoring etc. to health, i.e., only have basic data acquisition function, be related to condition-inference, Yi Jijian
There is still a need for could be realized by doctor is for further analysis to the data acquired for the part of health monitoring.
Invention content
The embodiment of the present application proposes recognition methods and the cloud system of heart sound, to solve existing Intelligent stethoscope intelligence journey
It spends relatively low, only has the technical issues of basic data acquisition function.
In one aspect, the embodiment of the present application provides a kind of recognition methods of heart sound, including：
Collected voice data is identified, recognition result is obtained；
If recognition result is heart sound data, the heart sound data is identified based on heart sound frequency, obtains the heart
The heart sound type of sound data.
On the other hand, the embodiment of the present application provides a kind of identification cloud system of heart sound, including：
First identification network obtains recognition result for collected voice data to be identified；
Second identification network, if for recognition result be heart sound data, based on heart sound frequency to the heart sound data into
Row identification, obtains the heart sound type of the heart sound data.
On the other hand, the embodiment of the present application provides a kind of electronic equipment, and the electronic equipment includes：
Transceiver, memory, one or more processors；And
One or more modules, one or more of modules are stored in the memory, and are configured to by institute
One or more processors execution is stated, one or more of modules include the finger for executing each step in the above method
It enables.
On the other hand, the embodiment of the present application provides a kind of computer program production being used in combination with electronic equipment
Product, the computer program product include computer-readable storage medium and are embedded in computer program mechanism therein, institute
It includes the instruction for executing each step in the above method to state computer program mechanism.
It has the beneficial effect that：
In the present embodiment, by the way that collected voice data is identified, it is determined whether be heart sound data, if being identified as
Heart sound data is then identified heart sound data based on heart sound frequency, obtains the heart sound type of heart sound data, and feed back to terminal
Equipment, to inform the current physical condition of user, and convenient for users to being decided whether according to the health status of itself
The remote assistance etc. for needing doctor, to realize the intelligence of user's condition-inference and health monitoring, while accuracy of identification
Preferably.
Description of the drawings
The specific embodiment of the application is described below with reference to accompanying drawings, wherein：
Fig. 1 is the method schematic of recognition of heart sound in the embodiment of the present application one；
Fig. 2 is the method flow schematic diagram of recognition of heart sound in the embodiment of the present application one；
Fig. 3 is the cloud system structure chart of recognition of heart sound in the embodiment of the present application two；
Fig. 4 is the structural schematic diagram of electronic equipment in the embodiment of the present application three.
Specific implementation mode
Below by way of specific example, the essence for embodiment technical solution that the present invention is furture elucidated.
In order to which the technical solution and advantage that make the application are more clearly understood, below in conjunction with attached drawing to the exemplary of the application
Embodiment is described in more detail, it is clear that and described embodiment is only a part of the embodiment of the application, rather than
The exhaustion of all embodiments.And in the absence of conflict, the feature in the embodiment and embodiment in this explanation can be mutual
It is combined.
Inventor notices during invention：
After existing Intelligent stethoscope only can be by recording to cardiopulmonary sound, it is still necessary to be set by terminal by doctor
The standby cardiopulmonary sound to user carries out condition-inference either health monitoring or by the way that cardiopulmonary sound is uploaded to high in the clouds, then
Long-range condition-inference or health monitoring are carried out by doctor, intelligence degree is relatively low.
Against the above deficiency/and it is based on this, the embodiment of the present application is proposed connects terminal device by Intelligent stethoscope, and will adopt
Whether the voice data from user collected is uploaded to cloud platform, be heart sound number using high in the clouds algorithm identification voice data
According to and heart sound data whether be normal cardiac sound, if normal cardiac sound, then send the normal recognition result of heart sound and set to terminal
It is standby, with the health for prompting user current；If abnormal heart sound, then specific heart sound Exception Type is identified, with convenient
User can decide whether the remote assistance of doctor according to the health status of itself, to realize in long-distance intelligent medical treatment
The extensive use in field.
For the ease of the implementation of the application, following Examples illustrates.
Embodiment 1
Fig. 1 shows the method schematic of recognition of heart sound in the embodiment of the present application one, and Fig. 2 shows the embodiment of the present application one
The method flow schematic diagram of middle recognition of heart sound, as shown in Figure 1 and Figure 2, this method includes：
Step 101：Collected voice data is identified, recognition result is obtained.
Step 102：If recognition result is heart sound data, the heart sound data is identified based on heart sound frequency, is obtained
To the heart sound type of the heart sound data.
In a step 101, Intelligent stethoscope acquires the voice data from user, and collected voice data is passed through
High in the clouds is uploaded to by terminal device and identifies network, collected voice data is identified by high in the clouds identification network, is known
Other result.
In a step 102, identification network in high in the clouds is for further processing according to recognition result, if recognition result is heart sound number
According to then further identifying whether the heart sound data is normal cardiac sound, if normal cardiac sound, then send the normal recognition result of heart sound
To terminal device, with the health for prompting user current；If abnormal heart sound, then specific heart sound exception class is identified
Type, with convenient for users to determining whether user needs the remote assistance of doctor according to the health status of itself.
In the present embodiment, described that collected voice data is identified, recognition result is obtained, including：
The voice data is identified using preset first nerves network, determines whether the voice data is the heart
Sound data.
In implementation, the first preset convolutional neural networks (CNN is utilized：Convolutional Neural Network) it is right
Collected voice data is identified, and is not heart sound data by the voice data if it not is heart sound data that recognition result, which is,
Recognition result be sent to terminal device, with prompt user to acquire sound position be adjusted；If recognition result is heart sound
Data are then for further processing to heart sound data by cloud system.
In the present embodiment, described that the heart sound data is identified based on heart sound frequency, obtain the heart sound data
Heart sound type, including：
The heart sound data is identified using preset nervus opticus network, obtains the heart sound class of the heart sound data
Type；
The preset nervus opticus network includes feature extraction network and sorter network, and the feature extraction network is base
It is trained in heart sound frequecy characteristic.
In the present embodiment, described that Classification and Identification is carried out to the heart sound data using preset nervus opticus network, it obtains
To the heart sound type of the heart sound data, including：
Classification and Identification is carried out to the heart sound data using preset nervus opticus network, whether determines the heart sound data
For normal cardiac sound；
If the heart sound data is normal cardiac sound, the normal recognition result of heart sound is sent；
If the heart sound data is abnormal heart sound, the abnormal heart sound is divided using preset third nerve network
Class identifies, obtains heart sound Exception Type.
In implementation, Classification and Identification is carried out to heart sound data using the second preset convolutional neural networks, that is, judges user
Heart sound (for example, palmic rate) it is whether normal, if heart sound data is normal cardiac sound, the normal recognition result of the heart sound is sent out
Terminal device is given, to prompt user's heart sound normal, while preserving the normal cardiac sound, for use in user's health shape
The statistical analysis of condition；If heart sound data is abnormal heart sound, using preset third convolutional neural networks to the exception heart sound into
Row Classification and Identification, determines specific heart disease type, such as arrhythmia cordis, valve disorder etc., and by heart disease class
The recognition result of type is sent to terminal device, with convenient for users to according to the health status of itself decide whether request doctor
Remote assistance.
In the present embodiment, the feature extraction network trains to obtain based on heart sound frequecy characteristic, including：
By the heart sound data of initialization according to heart sound frequency partition be multiple heart sound data samples；
It is respectively trained to obtain multiple feature extraction networks according to multiple heart sound data samples.
In implementation, specifically include：
1) heart sound data of initialization is pre-processed：
A) resampling, sample frequency 1000Hz, and to the heart sound number after resampling are carried out to the heart sound data of initialization
According to bandpass filtering treatment is carried out, heart sound data of the heart sound frequency within the scope of 25Hz-400Hz is obtained；
B) denoising is carried out to the spike noise in the heart sound data after bandpass filtering treatment；
C) mean value computation is carried out to the heart sound data after denoising and standard deviation calculates, and by subtracting mean value and divided by mark
The heart sound data after denoising is normalized in the mode of quasi- difference；
D) heart sound data after normalized is divided according to the heart sound period (for example, heart beat cycle), i.e., by institute
There is heart sound cycle duration to be uniformly extended to the longest heart sound period in all heart sound datas, for example, setting heart beat cycle as 2.5
Second；
E) heart sound data in each heart sound period after extension is split into 4 parts according to heart sound frequency, for example, by the heart
Sound data split into 4 hearts of the heart sound frequency within the scope of 25Hz-45Hz, 45Hz-80Hz, 80Hz-200Hz, 200Hz-400Hz
Phone data；Alternatively, heart sound data is split into heart sound frequency in 25Hz-45Hz, 45Hz-80Hz, 80Hz-200Hz, 200Hz-
5 heart sound subdatas within the scope of 400Hz, 400Hz-500Hz, fractionation quantity herein can be set according to actual conditions,
This implementation is not defined to splitting quantity.
2) training process of feature extraction network specifically includes：
4 heart sound subdatas that pretreatment obtains are inputted respectively in 4 feature extraction networks, to realize to 4 features
The training for extracting network, obtains trained 4 feature extraction networks；Alternatively, 5 heart sound subdatas point that pretreatment is obtained
To realize the training to 5 feature extraction networks, trained 5 feature extractions Shu Ru not be obtained in 5 feature extraction networks
The quantity of network, the feature extraction network trained herein can be set according to actual conditions, this implementation is not to feature extraction net
The quantity of network is defined.
3) training process of sorter network specifically includes：
Using trained 4 either 5 feature extraction networks extract to obtain the heart sound of 4 or 5 heart sound subdatas
Frequecy characteristic inputs the heart sound frequecy characteristic of 4 or 5 heart sound subdatas in the sorter network based on Sigmoid functions,
To realize the training to sorter network, so as to utilize trained sorter network, according to 4 or 5 heart sound subdatas
Heart sound frequecy characteristic output heart sound data each heart sound type prediction result, the prediction result be 2 and be 1 decimal,
The confidence level of normal cardiac sound and abnormal heart sound is corresponded to respectively.
Wherein, the heart sound frequecy characteristic and supplemental characteristic of multiple heart sound subdatas can also be passed through to the training of sorter network
It realizes, supplemental characteristic can be first heart sound, second heart sound, mel-frequency cepstrum coefficient (MFCC：Mel Frequency
Cepstrum Coefficient), the mark of the standard deviation of the amplitude kurtosis of first heart sound, the mean value of first heart sound, first heart sound
Accurate poor, second heart sound mean value, the standard deviation of second heart sound, first heart sound and the mean value at second heart sound interval, first heart sound and
One or more of the standard deviation at second heart sound interval, to improve the accuracy of identification of sorter network.
In the present embodiment, described that Classification and Identification is carried out to the heart sound data using preset nervus opticus network, it obtains
To the heart sound type of the heart sound data, including：
By the heart sound data according to heart sound frequency partition be multiple heart sound subdatas；
Using multiple feature extraction networks, the heart sound frequecy characteristic of the multiple heart sound subdata is obtained；
Using sorter network, identify to obtain the heart sound data according to the heart sound frequecy characteristic of the multiple heart sound subdata
Heart sound type.
In the present embodiment, the heart sound frequecy characteristic according to the multiple heart sound subdata identifies to obtain the heart sound
The heart sound type of data, including：
According to the heart sound frequecy characteristic of the multiple heart sound subdata, setting for each heart sound type in the heart sound data is determined
Reliability；
According to the confidence level of each heart sound type, the heart sound type of the heart sound data is determined.
In implementation, the detailed process packet of Classification and Identification is carried out to the heart sound data using preset nervus opticus network
It includes：
1) collected heart sound data is pre-processed, wherein by the heart sound data in N number of heart sound period according to heart sound frequency
Rate splits into 4*N or 5*N heart sound subdata；
2) by 4 of each heart sound period either 5 heart sound subdatas input training obtained 4 or 5 spies respectively
Sign extraction network obtains 4 either heart sound frequecy characteristics of 5 heart sound subdatas and by obtain 4 or 5 heart sound
In the sorter network that the heart sound frequecy characteristic input training of data obtains, each heart sound of the heart sound data in each heart sound period is obtained
The prediction result (being, for example, confidence level) of type, heart sound type includes normal cardiac sound and abnormal heart sound, the prediction of each heart sound type
As a result the decimal for being 2 and being 1, to obtain N number of heart sound period heart sound data normal cardiac sound and abnormal heart sound prediction
As a result.
3) be directed to N number of heart sound period heart sound data each heart sound type prediction result, respectively add up normal cardiac sound and
The confidence level summation of abnormal heart sound, and determine that heart sound type of the confidence level summation more than 0.5*N is final recognition result；Or
Person, using the higher heart sound type of confidence level summation in the prediction result of the heart sound data in N number of heart sound period as final identification
As a result, fixed condition can be set final recognition result according to actual conditions really herein, this implementation is not to determining specifically
Condition is defined.
In the present embodiment, the heart sound frequecy characteristic according to the multiple heart sound subdata identifies to obtain the heart sound
The heart sound type of data, including：
According to the heart sound frequecy characteristic and supplemental characteristic of the multiple heart sound subdata, the heart sound of the heart sound data is determined
Type；
The supplemental characteristic is the amplitude kurtosis of first heart sound, second heart sound, mel-frequency cepstrum coefficient, first heart sound
Standard deviation, the mean value of first heart sound, the standard deviation of first heart sound, the mean value of second heart sound, the standard deviation of second heart sound, first heart
Mean value, first heart sound and one or more of the standard deviation at second heart sound interval of sound and second heart sound interval.
In implementation, the heart sound frequecy characteristic of the input terminal of trained sorter network in addition to including multiple heart sound subdatas
Outside, the accuracy of identification of heart sound type can also be improved by increasing supplemental characteristic, for example, extracting the in each heart sound period
The standard deviation of the amplitude kurtosis of one heart sound, second heart sound, the mel-frequency cepstrum coefficient of diastole, first heart sound, first heart sound
Mean value, the standard deviation of first heart sound, the mean value of second heart sound, the standard deviation of second heart sound, between first heart sound and second heart sound
Every mean value, first heart sound and one or more of the standard deviation at second heart sound interval, and be added to trained classification net
The input terminal of network.Wherein, first heart sound and second heart sound can be the systole phase and diastole of corresponding heartbeat, can also be correspondence
The heart sounds and lung sounds of cardiopulmonary sound, herein first heart sound and second heart sound can be set according to actual conditions, this
Implementation does not limit first heart sound and second heart sound specifically.
The application is described in detail the embodiment of the present application 1, detailed process is as follows by taking concrete scene as an example：
Step 201：Collected voice data is inputted into the first convolutional neural networks, i.e. CNN1, and is identified whether as the heart
Sound data, if not heart sound data, then be sent to terminal device by the recognition result that the voice data is not heart sound data.
Step 202：If heart sound data, then heart sound data is inputted into the second convolutional neural networks, i.e. CNN2 identifies heart sound
Whether data (for example, palmic rate) are normal, if heart sound data is normal, the normal recognition result of the heart sound data is sent to
Terminal device, while preserving heart sound data.
Step 203：If heart sound data is abnormal, heart sound data is inputted into third convolutional neural networks, i.e. CNN3, identification
Go out specific heart disease type, such as arrhythmia cordis, valve disorder etc., and whether needs doctor remotely to help user
The result helped is sent to terminal device.
Embodiment 2
Based on same inventive concept, a kind of cloud system of recognition of heart sound is additionally provided in the embodiment of the present application, due to these
The principle that equipment solves the problems, such as is similar to a kind of method of recognition of heart sound, therefore the implementation of these equipment may refer to the reality of method
It applies, overlaps will not be repeated.
Fig. 3 shows the cloud system structure chart of recognition of heart sound in the embodiment of the present application two, as shown in figure 3, recognition of heart sound
Cloud system 300 may include：Terminal device 301, first identifies that network 302 and second identifies network 303.
Terminal device 301, for acquiring voice data.
First identification network 302 obtains recognition result for collected voice data to be identified.
Second identification network 303, if being heart sound data for recognition result, based on heart sound frequency to the heart sound data
It is identified, obtains the heart sound type of the heart sound data.
In the present embodiment, the first identification network includes：
The voice data is identified using preset first nerves network, determines whether the voice data is the heart
Sound data.
In the present embodiment, the second identification network includes：
The heart sound data is identified using preset nervus opticus network, obtains the heart sound class of the heart sound data
Type；
The preset nervus opticus network includes feature extraction network and sorter network, and the feature extraction network is base
It is trained in heart sound frequecy characteristic.
In the present embodiment, described that Classification and Identification is carried out to the heart sound data using preset nervus opticus network, it obtains
To the heart sound type of the heart sound data, including：
Classification and Identification is carried out to the heart sound data using preset nervus opticus network, whether determines the heart sound data
For normal cardiac sound；
If the heart sound data is normal cardiac sound, the normal recognition result of heart sound is sent；
If the heart sound data is abnormal heart sound, the abnormal heart sound is divided using preset third nerve network
Class identifies, obtains heart sound Exception Type.
In the present embodiment, the feature extraction network trains to obtain based on heart sound frequecy characteristic, including：
By the heart sound data of initialization according to heart sound frequency partition be multiple heart sound data samples；
It is respectively trained to obtain multiple feature extraction networks according to multiple heart sound data samples.
In the present embodiment, described that Classification and Identification is carried out to the heart sound data using preset nervus opticus network, it obtains
To the heart sound type of the heart sound data, including：
By the heart sound data according to heart sound frequency partition be multiple heart sound subdatas；
Using multiple feature extraction networks, the heart sound frequecy characteristic of the multiple heart sound subdata is obtained；
Using sorter network, identify to obtain the heart sound data according to the heart sound frequecy characteristic of the multiple heart sound subdata
Heart sound type.
In the present embodiment, the heart sound frequecy characteristic according to the multiple heart sound subdata identifies to obtain the heart sound
The heart sound type of data, including：
According to the heart sound frequecy characteristic of the multiple heart sound subdata, setting for each heart sound type in the heart sound data is determined
Reliability；
According to the confidence level of each heart sound type, the heart sound type of the heart sound data is determined.
In the present embodiment, the heart sound frequecy characteristic according to the multiple heart sound subdata identifies to obtain the heart sound
The heart sound type of data, including：
According to the heart sound frequecy characteristic and supplemental characteristic of the multiple heart sound subdata, the heart sound of the heart sound data is determined
Type；
The supplemental characteristic is the amplitude kurtosis of first heart sound, second heart sound, mel-frequency cepstrum coefficient, first heart sound
Standard deviation, the mean value of first heart sound, the standard deviation of first heart sound, the mean value of second heart sound, the standard deviation of second heart sound, first heart
Mean value, first heart sound and one or more of the standard deviation at second heart sound interval of sound and second heart sound interval.
Embodiment 3
Based on same inventive concept, a kind of electronic equipment is additionally provided in the embodiment of the present application, due to its principle and one kind
The method of recognition of heart sound is similar, therefore its implementation may refer to the implementation of method, and overlaps will not be repeated.
Fig. 4 shows the structural schematic diagram of electronic equipment in the embodiment of the present application three, as shown in figure 4, the electronic equipment
Including：Transceiver 401, memory 402, one or more processors 403；And one or more modules, it is one or
Multiple modules are stored in the memory, and are configured to be executed by one or more of processors, it is one or
Multiple modules include the instruction for executing each step in any above method.
Embodiment 4
Based on same inventive concept, the embodiment of the present application also provides a kind of computer journeys being used in combination with electronic equipment
Sequence product implements the implementation that may refer to method since its principle is similar to a kind of method of recognition of heart sound, repetition
Place repeats no more.The computer program product includes computer-readable storage medium and is embedded in computer program therein
Mechanism, the computer program mechanism include the instruction for executing each step in any above method.
For convenience of description, each section of apparatus described above is divided into various modules with function and describes respectively.Certainly, exist
Implement each module or the function of unit can be realized in same or multiple softwares or hardware when the application.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application
Apply the form of example.Moreover, the application can be used in one or more wherein include computer usable program code computer
The computer program production implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The application is with reference to method, the flow of equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram describe.It should be understood that can be realized by computer program instructions every first-class in flowchart and/or the block diagram
The combination of flow and/or box in journey and/or box and flowchart and/or the block diagram.These computer programs can be provided
Instruct the processor of all-purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine so that the instruction executed by computer or the processor of other programmable data processing devices is generated for real
The device for the function of being specified in present one flow of flow chart or one box of multiple flows and/or block diagram or multiple boxes.
These computer program instructions, which may also be stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that instruction generation stored in the computer readable memory includes referring to
Enable the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one box of block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device so that count
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, in computer or
The instruction executed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although the preferred embodiment of the application has been described, created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the application range.
Claims (18)
1. a kind of recognition methods of heart sound, which is characterized in that including：
Collected voice data is identified, recognition result is obtained；
If recognition result is heart sound data, the heart sound data is identified based on heart sound frequency, obtains the heart sound number
According to heart sound type.
2. the method as described in claim 1, which is characterized in that it is described that collected voice data is identified, known
Not as a result, including：
The voice data is identified using preset first nerves network, determines whether the voice data is heart sound number
According to.
3. the method as described in claim 1, which is characterized in that described to be known to the heart sound data based on heart sound frequency
Not, the heart sound type of the heart sound data is obtained, including：
The heart sound data is identified using preset nervus opticus network, obtains the heart sound type of the heart sound data；
The preset nervus opticus network includes feature extraction network and sorter network, and the feature extraction network is based on the heart
Voice frequency feature is trained.
4. method as claimed in claim 3, which is characterized in that it is described using preset nervus opticus network to the heart sound number
According to Classification and Identification is carried out, the heart sound type of the heart sound data is obtained, including：
Classification and Identification is carried out to the heart sound data using preset nervus opticus network, determines whether the heart sound data is just
Chang Xinyin；
If the heart sound data is normal cardiac sound, the normal recognition result of heart sound is sent；
If the heart sound data is abnormal heart sound, classification knowledge is carried out to the abnormal heart sound using preset third nerve network
Not, heart sound Exception Type is obtained.
5. method as claimed in claim 3, which is characterized in that the feature extraction network is to be trained based on heart sound frequecy characteristic
It obtains, including：
By the heart sound data of initialization according to heart sound frequency partition be multiple heart sound data samples；
It is respectively trained to obtain multiple feature extraction networks according to multiple heart sound data samples.
6. the method as described in claim 3 or 5, which is characterized in that it is described using preset nervus opticus network to the heart
Sound data carry out Classification and Identification, obtain the heart sound type of the heart sound data, including：
By the heart sound data according to heart sound frequency partition be multiple heart sound subdatas；
Using multiple feature extraction networks, the heart sound frequecy characteristic of the multiple heart sound subdata is obtained；
Using sorter network, identify to obtain the heart of the heart sound data according to the heart sound frequecy characteristic of the multiple heart sound subdata
Sound type.
7. method as claimed in claim 6, which is characterized in that described special according to the heart sound frequency of the multiple heart sound subdata
Sign identification obtains the heart sound type of the heart sound data, including：
According to the heart sound frequecy characteristic of the multiple heart sound subdata, the confidence of each heart sound type in the heart sound data is determined
Degree；
According to the confidence level of each heart sound type, the heart sound type of the heart sound data is determined.
8. method as claimed in claim 6, which is characterized in that described special according to the heart sound frequency of the multiple heart sound subdata
Sign identification obtains the heart sound type of the heart sound data, including：
According to the heart sound frequecy characteristic and supplemental characteristic of the multiple heart sound subdata, the heart sound class of the heart sound data is determined
Type；
The supplemental characteristic is the standard of the amplitude kurtosis of first heart sound, second heart sound, mel-frequency cepstrum coefficient, first heart sound
Difference, the mean value of first heart sound, the standard deviation of first heart sound, the mean value of second heart sound, the standard deviation of second heart sound, first heart sound with
Mean value, first heart sound and one or more of the standard deviation at second heart sound interval at second heart sound interval.
9. a kind of identification cloud system of heart sound, which is characterized in that including：
Terminal device, for acquiring voice data；
First identification network obtains recognition result for collected voice data to be identified；
Second identification network knows the heart sound data based on heart sound frequency if being heart sound data for recognition result
Not, the heart sound type of the heart sound data is obtained.
10. cloud system as claimed in claim 9, which is characterized in that described first identifies that network includes：
The voice data is identified using preset first nerves network, determines whether the voice data is heart sound number
According to.
11. cloud system as claimed in claim 9, which is characterized in that described second identifies that network includes：
The heart sound data is identified using preset nervus opticus network, obtains the heart sound type of the heart sound data；
The preset nervus opticus network includes feature extraction network and sorter network, and the feature extraction network is based on the heart
Voice frequency feature is trained.
12. cloud system as claimed in claim 11, which is characterized in that it is described using preset nervus opticus network to the heart
Sound data carry out Classification and Identification, obtain the heart sound type of the heart sound data, including：
Classification and Identification is carried out to the heart sound data using preset nervus opticus network, determines whether the heart sound data is just
Chang Xinyin；
If the heart sound data is normal cardiac sound, the normal recognition result of heart sound is sent；
If the heart sound data is abnormal heart sound, classification knowledge is carried out to the abnormal heart sound using preset third nerve network
Not, heart sound Exception Type is obtained.
13. cloud system as claimed in claim 11, which is characterized in that the feature extraction network is based on heart sound frequecy characteristic
What training obtained, including：
By the heart sound data of initialization according to heart sound frequency partition be multiple heart sound data samples；
It is respectively trained to obtain multiple feature extraction networks according to multiple heart sound data samples.
14. the cloud system as described in claim 11 or 13, which is characterized in that it is described using preset nervus opticus network to institute
It states heart sound data and carries out Classification and Identification, obtain the heart sound type of the heart sound data, including：
By the heart sound data according to heart sound frequency partition be multiple heart sound subdatas；
Using multiple feature extraction networks, the heart sound frequecy characteristic of the multiple heart sound subdata is obtained；
Using sorter network, identify to obtain the heart of the heart sound data according to the heart sound frequecy characteristic of the multiple heart sound subdata
Sound type.
15. cloud system as claimed in claim 14, which is characterized in that the heart sound frequency according to the multiple heart sound subdata
Rate feature recognition obtains the heart sound type of the heart sound data, including：
According to the heart sound frequecy characteristic of the multiple heart sound subdata, the confidence of each heart sound type in the heart sound data is determined
Degree；
According to the confidence level of each heart sound type, the heart sound type of the heart sound data is determined.
16. cloud system as claimed in claim 14, which is characterized in that the heart sound frequency according to the multiple heart sound subdata
Rate feature recognition obtains the heart sound type of the heart sound data, including：
According to the heart sound frequecy characteristic and supplemental characteristic of the multiple heart sound subdata, the heart sound class of the heart sound data is determined
Type；
The supplemental characteristic is the standard of the amplitude kurtosis of first heart sound, second heart sound, mel-frequency cepstrum coefficient, first heart sound
Difference, the mean value of first heart sound, the standard deviation of first heart sound, the mean value of second heart sound, the standard deviation of second heart sound, first heart sound with
Mean value, first heart sound and one or more of the standard deviation at second heart sound interval at second heart sound interval.
17. a kind of electronic equipment, which is characterized in that the electronic equipment includes：
Transceiver, memory, one or more processors；And
One or more modules, one or more of modules are stored in the memory, and are configured to by described one
A or multiple processors execute, and one or more of modules include being required in 1-8 in any the method for perform claim
The instruction of each step.
18. a kind of computer program product being used in combination with electronic equipment, the computer program product includes that computer can
The storage medium of reading and it is embedded in computer program mechanism therein, the computer program mechanism includes being wanted for perform claim
Ask the instruction of each step in any the method in 1-8.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2018/073237 WO2019140600A1 (en) | 2018-01-18 | 2018-01-18 | Method for identifying cardiac sound, and cloud system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108323158A true CN108323158A (en) | 2018-07-24 |
Family
ID=62895870
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201880000118.0A Pending CN108323158A (en) | 2018-01-18 | 2018-01-18 | Heart sound identification method and cloud system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108323158A (en) |
| WO (1) | WO2019140600A1 (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109107016A (en) * | 2018-08-17 | 2019-01-01 | 贵州优品睡眠健康产业有限公司 | Body-sensing vibration music aid-sleeping system |
| CN109330622A (en) * | 2018-11-21 | 2019-02-15 | 英华达（上海）科技有限公司 | Intelligent human-body monitors system and its abdomen sound monitoring device |
| CN110123367A (en) * | 2019-04-04 | 2019-08-16 | 平安科技（深圳）有限公司 | Computer equipment, recognition of heart sound device, method, model training apparatus and storage medium |
| CN110558944A (en) * | 2019-09-09 | 2019-12-13 | 成都智能迭迦科技合伙企业(有限合伙) | Heart sound processing method and device, electronic equipment and computer readable storage medium |
| CN111904459A (en) * | 2020-08-27 | 2020-11-10 | 广东汉泓医疗科技有限公司 | Cardiopulmonary sound auscultation detector for guiding rapid auscultation, auscultation system and auscultation method |
| CN112598086A (en) * | 2021-03-04 | 2021-04-02 | 四川大学 | Deep neural network-based common colon disease classification method and auxiliary system |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001022883A1 (en) * | 1999-09-29 | 2001-04-05 | Siemens Corporate Research, Inc. | Multi-modal cardiac diagnostic decision support system and method |
| CN1850007A (en) * | 2006-05-16 | 2006-10-25 | 清华大学深圳研究生院 | Heart disease automatic classification system based on heart sound analysis and heart sound segmentation method |
| CN101478921A (en) * | 2006-06-26 | 2009-07-08 | 科洛普拉斯特公司 | Multi parametric classification of cardiovascular sound |
| CN102334985A (en) * | 2010-07-16 | 2012-02-01 | 香港理工大学 | Method for detecting pulmonary artery blood pressure by using heart sound analysis method of multilayer feedforward network |
| CN103479383A (en) * | 2013-09-25 | 2014-01-01 | 清华大学 | Method and device for analyzing heart sound signals, and intelligent heart stethoscope provided with device for analyzing heart sound signals |
| CN105212960A (en) * | 2015-08-19 | 2016-01-06 | 四川长虹电器股份有限公司 | Cardiechema signals method for evaluating quality |
| CN105982692A (en) * | 2015-04-29 | 2016-10-05 | 广东医学院附属医院 | Multifunctional stethoscope for broadband parameter monitoring and implementation method of multifunctional stethoscope |
| CN106308845A (en) * | 2015-06-23 | 2017-01-11 | 黄楚 | Intelligent remote stethoscope and using method thereof |
| CN106983519A (en) * | 2017-05-11 | 2017-07-28 | 戊星务(大连)医疗器械有限公司 | Wireless electron auscultation display and audio and electrocardiographic oscillogram diagnostic analysis method |
| CN107529645A (en) * | 2017-06-29 | 2018-01-02 | 重庆邮电大学 | A kind of heart sound intelligent diagnosis system and method based on deep learning |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7351207B2 (en) * | 2003-07-18 | 2008-04-01 | The Board Of Trustees Of The University Of Illinois | Extraction of one or more discrete heart sounds from heart sound information |
| CN102697520B (en) * | 2012-05-08 | 2014-10-29 | 天津沃康科技有限公司 | Electronic stethoscope based on intelligent distinguishing function |
| CN105662457A (en) * | 2016-03-22 | 2016-06-15 | 宁波元鼎电子科技有限公司 | Intelligent echometer |
-
2018
- 2018-01-18 WO PCT/CN2018/073237 patent/WO2019140600A1/en active Application Filing
- 2018-01-18 CN CN201880000118.0A patent/CN108323158A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001022883A1 (en) * | 1999-09-29 | 2001-04-05 | Siemens Corporate Research, Inc. | Multi-modal cardiac diagnostic decision support system and method |
| CN1850007A (en) * | 2006-05-16 | 2006-10-25 | 清华大学深圳研究生院 | Heart disease automatic classification system based on heart sound analysis and heart sound segmentation method |
| CN101478921A (en) * | 2006-06-26 | 2009-07-08 | 科洛普拉斯特公司 | Multi parametric classification of cardiovascular sound |
| CN102334985A (en) * | 2010-07-16 | 2012-02-01 | 香港理工大学 | Method for detecting pulmonary artery blood pressure by using heart sound analysis method of multilayer feedforward network |
| CN103479383A (en) * | 2013-09-25 | 2014-01-01 | 清华大学 | Method and device for analyzing heart sound signals, and intelligent heart stethoscope provided with device for analyzing heart sound signals |
| CN105982692A (en) * | 2015-04-29 | 2016-10-05 | 广东医学院附属医院 | Multifunctional stethoscope for broadband parameter monitoring and implementation method of multifunctional stethoscope |
| CN106308845A (en) * | 2015-06-23 | 2017-01-11 | 黄楚 | Intelligent remote stethoscope and using method thereof |
| CN105212960A (en) * | 2015-08-19 | 2016-01-06 | 四川长虹电器股份有限公司 | Cardiechema signals method for evaluating quality |
| CN106983519A (en) * | 2017-05-11 | 2017-07-28 | 戊星务(大连)医疗器械有限公司 | Wireless electron auscultation display and audio and electrocardiographic oscillogram diagnostic analysis method |
| CN107529645A (en) * | 2017-06-29 | 2018-01-02 | 重庆邮电大学 | A kind of heart sound intelligent diagnosis system and method based on deep learning |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109107016A (en) * | 2018-08-17 | 2019-01-01 | 贵州优品睡眠健康产业有限公司 | Body-sensing vibration music aid-sleeping system |
| CN109330622A (en) * | 2018-11-21 | 2019-02-15 | 英华达（上海）科技有限公司 | Intelligent human-body monitors system and its abdomen sound monitoring device |
| CN110123367A (en) * | 2019-04-04 | 2019-08-16 | 平安科技（深圳）有限公司 | Computer equipment, recognition of heart sound device, method, model training apparatus and storage medium |
| CN110123367B (en) * | 2019-04-04 | 2022-11-15 | 平安科技（深圳）有限公司 | Computer device, heart sound recognition method, model training device, and storage medium |
| CN110558944A (en) * | 2019-09-09 | 2019-12-13 | 成都智能迭迦科技合伙企业(有限合伙) | Heart sound processing method and device, electronic equipment and computer readable storage medium |
| CN111904459A (en) * | 2020-08-27 | 2020-11-10 | 广东汉泓医疗科技有限公司 | Cardiopulmonary sound auscultation detector for guiding rapid auscultation, auscultation system and auscultation method |
| CN112598086A (en) * | 2021-03-04 | 2021-04-02 | 四川大学 | Deep neural network-based common colon disease classification method and auxiliary system |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019140600A1 (en) | 2019-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108323158A (en) | Heart sound identification method and cloud system | |
| Yang et al. | Using PPG signals and wearable devices for atrial fibrillation screening | |
| Tazarv et al. | A deep learning approach to predict blood pressure from ppg signals | |
| TW201642803A (en) | Wearable pulse sensing device signal quality estimation | |
| WO2021017313A1 (en) | Atrial fibrillation detection method and apparatus, computer device, and storage medium | |
| CN111772619A (en) | Electrocardiogram heart beat identification method, terminal device and storage medium | |
| CA3069891A1 (en) | Discovering novel features to use in machine learning techniques, such as machine learning techniques for diagnosing medical conditions | |
| US8798726B2 (en) | Method and apparatus for eliminating motion artifacts of bio signal using personalized bio signal pattern | |
| RU2657384C2 (en) | Method and system for noninvasive screening physiological parameters and pathology | |
| CN109602410A (en) | A kind of wearable device and its monitoring of pulse method | |
| CN105701196A (en) | Intelligent robot oriented audio processing method and intelligent robot | |
| Chen et al. | Edge2Analysis: a novel AIoT platform for atrial fibrillation recognition and detection | |
| CN108461130A (en) | Medical task intelligent dispatching method and system | |
| JP2023505111A (en) | Systems and methods for physiological measurements from optical data | |
| CN115778403A (en) | Electrocardiogram analysis method, electrocardiogram analysis device, electronic equipment and storage medium | |
| CN114027804A (en) | Pulse condition diagnosis method, device and readable storage medium | |
| CN113693610B (en) | Method and device for processing few-lead electrocardiogram data, storage medium and computer equipment | |
| CN117349600A (en) | Heart sound and heart electricity combined diagnosis method and system based on dual-mode dual input | |
| TWI792333B (en) | Prediction method and system of low blood pressure | |
| Hamza et al. | A Comprehensive Overview of Heart Sound Analysis Using Machine Learning Methods | |
| CN114469040A (en) | Heart rate detection method and device, storage medium and electronic equipment | |
| JP2022016356A (en) | System and method to determine location of catheter | |
| CN109192315B (en) | Comprehensive age detection system based on weighted kernel regression and packaged deviation search | |
| TWI795219B (en) | Method and system of detecting specific physiological syndrome related to blood circulation and deep sleep based on hemodynamic analysis | |
| Kamath et al. | Modeling QRS complex in dECG |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210210 Address after: 201111 2nd floor, building 2, no.1508, Kunyang Road, Minhang District, Shanghai Applicant after: Dalu Robot Co.,Ltd. Address before: 518000 Room 201, building A, No. 1, Qian Wan Road, Qianhai Shenzhen Hong Kong cooperation zone, Shenzhen, Guangdong (Shenzhen Qianhai business secretary Co., Ltd.) Applicant before: CLOUDMINDS (SHENZHEN) ROBOTICS SYSTEMS Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 201111 Building 8, No. 207, Zhongqing Road, Minhang District, Shanghai Applicant after: Dayu robot Co.,Ltd. Address before: 201111 2nd floor, building 2, no.1508, Kunyang Road, Minhang District, Shanghai Applicant before: Dalu Robot Co.,Ltd. |
|
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180724 |