CN110473616A - A kind of audio signal processing method, apparatus and system - Google Patents
A kind of audio signal processing method, apparatus and system Download PDFInfo
- Publication number
- CN110473616A CN110473616A CN201910757443.8A CN201910757443A CN110473616A CN 110473616 A CN110473616 A CN 110473616A CN 201910757443 A CN201910757443 A CN 201910757443A CN 110473616 A CN110473616 A CN 110473616A
- Authority
- CN
- China
- Prior art keywords
- patient
- department
- voice messaging
- preset
- information
- 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.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 12
- 230000005236 sound signal Effects 0.000 title claims abstract description 12
- 208000024891 symptom Diseases 0.000 claims abstract description 79
- 238000010276 construction Methods 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 19
- 230000001755 vocal effect Effects 0.000 claims description 40
- 230000006870 function Effects 0.000 claims description 24
- 230000036760 body temperature Effects 0.000 claims description 18
- 230000001815 facial effect Effects 0.000 claims description 16
- 210000002345 respiratory system Anatomy 0.000 claims description 12
- 230000009467 reduction Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 description 10
- 238000003062 neural network model Methods 0.000 description 9
- 206010015150 Erythema Diseases 0.000 description 5
- 206010061218 Inflammation Diseases 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 231100000321 erythema Toxicity 0.000 description 5
- 230000004054 inflammatory process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 210000005036 nerve Anatomy 0.000 description 2
- 235000006041 Prunus persica f compressa Nutrition 0.000 description 1
- 240000006522 Prunus persica f. compressa Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/66—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for extracting parameters related to health condition
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- General Business, Economics & Management (AREA)
- Multimedia (AREA)
- Primary Health Care (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Databases & Information Systems (AREA)
- Quality & Reliability (AREA)
- Pathology (AREA)
- Data Mining & Analysis (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application provides a kind of audio signal processing methods, apparatus and system, wherein method includes: in the case where receiving the first preset instructions, acquire the first voice messaging, first voice messaging includes the symptom information of patient, is identified to the first voice messaging, and the symptom information of patient is obtained.The history information of the patient is read from the social security card of patient, at least by the symptom information of patient and history information construction feature vector, feature vector is inputted into preset model, obtain registering department needed for patient as first object department, preset model is at least with the symptom information of historic patient and history information, and the department that registers of mark is that training sample training obtains, and obtains the number that patient goes to a doctor in first object department, output number.The application can accurately determine the department that registers needed for patient.
Description
Technical field
This application involves electronic information field more particularly to a kind of audio signal processing methods, apparatus and system.
Background technique
During hospital registers, the department to register needed for patient is indefinite in some cases, for example, with the epoch
The division for developing hospital department is more and more careful, alternatively, the symptom of patient is unobvious etc., causes to register needed for patient is indefinite
Department.Currently, can be by registering on machine of independently registering.
But the accuracy of registering for machine of independently registering is low, i.e. the user department that uses machine of independently registering to be hung is uses
The probability for the department that family should actually register is lower.
Summary of the invention
This application provides a kind of audio signal processing methods, apparatus and system, it is therefore intended that solves machine of independently registering
The low problem of accuracy of registering.
To achieve the goals above, this application provides following technical schemes:
This application provides a kind of audio signal processing methods, comprising:
In the case where receiving the first preset instructions, the first voice messaging is acquired；First voice messaging includes suffering from
The symptom information of person；
First voice messaging is identified, the symptom information of the patient is obtained；
The history information of the patient is read from the social security card of the patient；
At least by the symptom information of the patient and history information construction feature vector；
Described eigenvector is inputted into preset model, first object department is in the department that obtains registering needed for the patient；
The preset model is at least instructed with the symptom information of historic patient and history information, and the department that registers of mark for training sample
It gets；
Obtain the number that the patient goes to a doctor in the first object department；
Export the number.
Optionally, after the first voice messaging of the acquisition, further includes:
Vocal print is identified from first voice messaging；
In the case where the vocal print is the vocal print of the patient, identify that the first of the patient is pre- from the vocal print
If the current value of vital sign；The first default vital sign is the default sign for indicating respiratory system；Also, described in obtaining
The current value of the default vital sign of the second of patient；The second default vital sign is to indicate body temperature and preset facial characteristics
Sign；
It is described at least by the symptom information of the patient and history information construction feature vector, comprising:
By the symptom information of the patient, history information, the patient the first default vital sign current value, and
The current value of the default vital sign of the second of the patient, construction feature vector.
Optionally, the output layer of the preset model includes preset regression function；The regression function output is preset
The department that respectively registers is respectively the probability of department of registering needed for the patient, and probability is greater than predetermined probabilities threshold value by the output layer
And it is the first object department that the number of having registered, which is not up to the department that registers of preset threshold,.
Optionally, described that first voice messaging is identified, obtain the symptom information of the patient, comprising:
Gain and/or noise reduction are carried out to first voice messaging, first voice messaging that obtains that treated；
Treated that the first voice messaging is identified to described, obtains the symptom information of the patient.
Optionally, further includes:
In the case where receiving the second preset instructions, the second voice messaging is acquired；Include in second voice messaging
The information of the department to register needed for the patient；
Second voice messaging is identified, obtaining the department to register needed for the patient is the second target department；
Obtain the number that the patient goes to a doctor in second target department；
Export the number.
Present invention also provides a kind of speech signal processing devices, comprising:
Acquisition module, for acquiring the first voice messaging in the case where receiving the first preset instructions；First language
Message breath includes the symptom information of patient；
Identification module obtains the symptom information of the patient for identifying to first voice messaging；
Read module, for reading the history information of the patient from the social security card of the patient；
Module is constructed, at least by the symptom information of the patient and history information construction feature vector；
Input module, for described eigenvector to be inputted preset model, the department that obtains registering needed for the patient is
First object department；The preset model is at least with the symptom information of historic patient and history information, and the section that registers of mark
Room is that training sample training obtains；
Module is obtained, the number gone to a doctor for obtaining the patient in the first object department；
Output module, for exporting the number.
Optionally, the acquisition module is also used to after the first voice messaging of the acquisition, is believed from first voice
Vocal print is identified in breath；
The identification module is also used to know from the vocal print in the case where the vocal print is the vocal print of the patient
Not Chu the patient the first default vital sign current value；The first default vital sign is the pre- of expression respiratory system
If sign；Also, obtain the current value of the second default vital sign of the patient；The second default vital sign is to indicate
The sign of body temperature and preset facial characteristics；
The building module, at least by the symptom information of the patient and history information construction feature vector, comprising:
The building module, specifically for first presetting the symptom information of the patient, history information, the patient
The current value of the second default vital sign of the current value of vital sign and the patient, construction feature vector.
Optionally, the output layer of the preset model includes preset regression function；The regression function output is preset
The department that respectively registers is respectively the probability of department of registering needed for the patient, and probability is greater than predetermined probabilities threshold value by the output layer
And it is the first object department that the number of having registered, which is not up to the department that registers of preset threshold,.
Optionally, the identification module obtains the symptom of the patient for identifying to first voice messaging
Information, comprising:
The identification module, specifically for obtaining to first voice messaging progress gain and/or noise reduction, treated
First voice messaging；Treated that the first voice messaging is identified to described, obtains the symptom information of the patient.
Optionally, further includes: processing module, for acquiring the second voice in the case where receiving the second preset instructions
Information；It include the information of the department to register needed for the patient in second voice messaging；
Second voice messaging is identified, obtaining the department to register needed for the patient is the second target department；
Obtain the number that the patient goes to a doctor in second target department；
Export the number.
Present invention also provides a kind of speech signal processing systems, comprising: processor；
The processor, for acquiring the first voice messaging in the case where receiving the first preset instructions；Described first
Voice messaging includes the symptom information of patient；
First voice messaging is identified, the symptom information of the patient is obtained；
The history information of the patient is read from the social security card of the patient；
At least by the symptom information of the patient and history information construction feature vector；
Input preset model is read from the social security card of the patient, the department that obtains registering needed for the patient is first
Target department；The preset model is at least with the symptom information of historic patient and history information, and the department that registers marked is
Training sample training obtains；
Obtain the number that the patient goes to a doctor in the first object department；
Export the number.
Optionally, the system also includes remote temperature sensor and face sensors；The remote temperature sensor with
The face sensor is connected with the processor respectively；
The remote temperature sensor, for measuring body temperature；
The face sensor, for extracting facial characteristics；
The processor is also used to after receiving the first voice messaging, identifies vocal print from first voice messaging,
In the case where the vocal print is the vocal print of the patient, the first default life entity of the patient is identified from the vocal print
The current value of sign；The first default vital sign is the default sign for indicating respiratory system；And it is passed from the remote temperature
In sensor and the face sensor, the current value of the second default vital sign of the patient is extracted；Described second default life
Life sign is to indicate the sign of body temperature and preset facial characteristics；
The processor, at least by the symptom information of the patient and history information construction feature vector, comprising:
The processor, specifically for by the first default life of the symptom information of the patient, history information, the patient
Order the current value of the current value of sign and the second default vital sign of the patient, construction feature vector.
Optionally, the output layer of the preset model includes preset regression function；The regression function output is preset
The department that respectively registers is respectively the probability of department of registering needed for the patient, and probability is greater than predetermined probabilities threshold value by the output layer
And it is the first object department that the number of having registered, which is not up to the department that registers of preset threshold,.
Optionally, the processor is also used in the case where receiving the second preset instructions, acquisition the second voice letter
Breath；It include the information of the department to register needed for the patient in second voice messaging；Second voice messaging is carried out
Identification, obtaining the department to register needed for the patient is the second target department；For the patient second target department into
Row registration obtains the number that the patient goes to a doctor in second target department；Export the number.
In audio signal processing method described herein, apparatus and system, the case where receiving the first preset instructions
Under, acquire the first voice messaging, wherein include the symptom information of patient in the first voice messaging, know from the first voice messaging
Not Chu patient symptom information, the history information of patient is read from the social security card of patient, at least by the symptom information of patient and
The feature vector of building is inputted preset model by history information construction feature vector, and the department that obtains registering needed for patient is first
Target department obtains the number that patient goes to a doctor in first object department, exports the number.Wherein, preset model is at least with history
The symptom information and history information of patient, and the department that registers of mark is that training sample training obtains.
The symptom information of patient is obtained by way of speech recognition since the application gives, and combines the medical history of patient
Information determines the department that registers needed for patient, i.e. first object department using preset model.Also, patient is obtained in the first mesh
The medical number of department is marked, and exports the number.Therefore, the application can determine register needed for patient department and model
Department is registered needed for determining patient with certain accuracy.And patient is obtained in the medical of the required department that registers
Number, as patient registers in the required department that registers, and therefore, reduces the indefinite actually required department that registers of user
Caused by misplaced department the accuracy that machine registers of independently registering using application scheme can be improved in turn.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of training process schematic diagram of preset model disclosed in the embodiment of the present application；
Fig. 2 is a kind of flow chart of audio signal processing method disclosed in the embodiment of the present application；
Fig. 3 is a kind of structural schematic diagram of speech signal processing device disclosed in the embodiment of the present application；
Fig. 4 is a kind of structural schematic diagram of speech signal processing system disclosed in the embodiment of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on
Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall in the protection scope of this application.
Inventor has found under study for action, the low reason of the existing accuracy that machine is registered of independently registering are as follows: it is existing from
Main machine of registering can not accurately predict the department (department that registers for meeting patient symptom) that patient should actually register.Therefore,
In the embodiment of the present application, the integrated speech signal processing apparatus in machine of independently registering, so that the autonomous registration machine after integrated
Device determines the symptom information of patient according to collected voice, and the history information of patient is obtained from the social security card of patient, will be suffered from
The symptom information and history information of person inputs preset model, obtains the department that registers needed for patient, wherein preset model is by going through
The symptom information and history information of history patient, and the department that registers of mark is that training sample training obtains, so that guaranteeing default
Model determines that the department that registers needed for patient has certain accuracy, thus, improve the accuracy registered.
Fig. 1 is a kind of training process schematic diagram of preset model provided by the embodiments of the present application, comprising the following steps:
S101, training sample is obtained.
In the present embodiment, training sample includes at least: the history information of the symptom information of historic patient, historic patient,
And the department that registers of historic patient.
In order to enable the accuracy for the department that registers needed for the patient of the model output after training is higher, in the present embodiment
In, training sample can also include: the booking time of historic patient, the place of registering of historic patient, historic patient it is first pre-
If the value and historic patient of the default vital sign of the second of the value of vital sign, historic patient are preset each when registering
It registers the situation of registering of department.
Wherein, locating season when booking time can register for user, for example, summer.Place of registering can be user
The area at place, for example, county XX, the city XX, XX province etc..
First default vital sign is the default sign for indicating respiratory system, such as: whether catch a cold with whether tonsillotome is deposited
In inflammation.The value of first default vital sign is the value of the default sign of respiratory system, such as, if the result of flu
(for example, flu) and tonsillotome whether the result (for example, there are inflammation for tonsillotome) etc. of inflammation.
Second default vital sign is the sign for indicating body temperature and preset facial characteristics, wherein preset facial characteristics
It can be skin color and skin with the presence or absence of erythema etc..The value of second default vital sign is the second default vital sign
End value, for example, the result (for example, body temperature be 37 degrees Celsius) of body temperature, colour of skin result (for example, partially yellow) and skin whether
There are the result of erythema (for example, without erythema on skins).
The situation of registering of the preset department that respectively registers may include: that the preset department that respectively registers has registered the quantity of personnel,
And whether the quantity for the personnel that registered reaches preset threshold.Wherein, the preset department that registers can register for hospital is all
Department, specifically, the particular content of the preset department that registers can be set according to actual needs, the present embodiment is not to default
The particular content of the department that registers limit.
Preset threshold refers to the upper limit value of acceptable personnel amount of registering in the department that registers, specifically, preset threshold
Specific value can be configured according to the actual situation, and the present embodiment does not limit the specific value of preset threshold.
S102, training sample input preset model is trained, the preset model after being trained.
In the present embodiment, preset model can be deep neural network model, it is of course also possible to be other models, this
Embodiment does not limit the particular content of preset model.
In the case where preset model is deep neural network model, deep neural network model includes output layer, also,
Output layer includes preset regression function, is being inputted into the deep neural network model for predicting the department that registers needed for patient
Feature vector after, the regression function export patient preset each department respectively needed for the probability registered, depth nerve
The output layer of network model by probability be greater than predetermined probabilities threshold value and the number of having registered be not up to preset threshold register department into
Row output.
For example, training sample is the symptom information of historic patient, the history information of historic patient, when registering of historic patient
Between, the place of registering of historic patient, the value of the first default vital sign of historic patient, historic patient the second default life
The department that registers of the value of sign, historic patient register situation and the historic patient of the preset department that respectively registers when registering.
May include: to the process that deep neural network model is trained using the training sample
In the present embodiment, first, by the symptom information of historic patient in training sample, historic patient history information,
The booking time of historic patient, the place of registering of historic patient, the value of the first default vital sign of historic patient, history are suffered from
The value of the default vital sign of the second of person, historic patient are respectively registered the situation of registering of department when registering, and are configured to feature vector.
Second, the feature vector of building is inputted into depth nerve neural network, the recurrence of the deep neural network output layer
It is the probability of department of registering needed for the historic patient that function, which exports the preset department that respectively registers, that is, is determined in the preset department that registers
Each probability of the department as the department that registers needed for the historic patient of registering.
Third, also by each department that registers of regression function output, probability is greater than the output layer of the deep neural network
The department that registers that predetermined probabilities threshold value and personnel amount of having registered are not up to preset threshold is exported.
4th, after the output layer of the deep neural network exports the department that registers needed for the historic patient, by default
Loss function, calculate between the department that registers of the historic patient marked in register department and training sample of output layer output
Loss function value.And the value of each parameter matrix in the deep neural network model is adjusted according to the loss function value.
5th, using the training method of above-mentioned first to fourth this four step, by multiple training samples to depth nerve net
Network model carries out circuit training, the deep neural network model after being trained.
The beneficial effect of the present embodiment includes:
The output layer of neural network model predicts the department that registers using classification function in the prior art, in the extension of prediction
In the case that the personnel amount registered in number department reaches preset threshold, then it represents that currently without the registrable department of patient,
But in practice, patient can register department in addition to what is predicted registers other than section room, be likely present others and register department, by
The department that registers only is exported in classification function, causes to miss the available department that registers of patient, the i.e. prediction result of model output
Accuracy it is low.
In the present embodiment, the output layer of deep neural network model includes preset regression function, and the regression function is defeated
Preset each probability of the department as the department that registers needed for patient of registering out, also, output layer output is that probability is greater than
Predetermined probabilities threshold value and the number of having registered are not up to the department that registers of preset threshold, so that probability is greater than the section that registers of preset threshold
The quantity of room may be it is multiple, the quantity of the personnel that registered of some department that registers reaches default threshold in multiple department that registers
In the case where value, output layer can be greater than with output probability predetermined probabilities threshold value and the number of having registered be not up to preset threshold its
He registers department, so as to avoid the registrable department of patient is missed in the prior art, and then improves deep neural network mould
The accuracy of the prediction result of type output.
Fig. 2 is a kind of audio signal processing method provided by the embodiments of the present application, comprising the following steps:
S201, judge whether to receive the first preset instructions, if it is, S202 is executed, if it is not, then executing S213.
In the present embodiment, the first preset instructions indicate the department that registers needed for the indefinite patient of user, wherein Yong Huke
To be sufferers themselves, it is also possible to the people for helping patient to register.In the present embodiment, two can be shown on machine of independently registering
A button mark, a button indication are clearly registered department, and the department that registers is not known in another button indication.User is according to reality
Border situation, the button clicked needed for selecting.In this step, the instruction that the triggering of some button is clicked by user, judges whether
Receive the first preset instructions.It, in practice, can be with it should be noted that this step has been merely given as a kind of implementation
It realizes by other means, the present embodiment do not limit concrete implementation mode.
S202, the first voice messaging of acquisition.
It in the present embodiment, include the symptom information of patient in the first voice messaging.
S203, the first voice messaging is identified, obtains the symptom information of patient.
Specifically, in order to guarantee the disease for identifying patient from the first voice messaging collected in noisy environment
Shape information is in the present embodiment handled collected first voice messaging, first voice messaging that obtains that treated,
And the symptom information of patient is identified from treated the first voice messaging.
It wherein, include: that speech gain is oriented to the first voice messaging to the processing of the first voice messaging of acquisition,
And/or it is oriented noise reduction gain.Wherein, orientation noise reduction gain is specifically as follows DOA orientation noise reduction gain.Specifically, orientation
Speech gain and the realization process for orienting noise reduction gain are all the prior art, are not discussed here.
S204, the history information for obtaining patient.
In the present embodiment, the equipment of audio signal processing method provided by the invention is integrated (for example, self-service extension in use
Number machine) during, user can first be inserted into the social security card of patient.In this step, patient can be read from social security card
History information.
S205, vocal print is identified from the first voice messaging.
Specifically, vocal print can be identified from the first voice messaging using MFCC voiceprint recognition algorithm, specifically, identification sound
The process of line is the prior art, and which is not described herein again.
Whether the vocal print that S206, judgement identify is the vocal print of patient, if it is, S207 is executed, if it is not, then executing
S210。
In the present embodiment, it is hung in the equipment that social security card is inserted into integrated the application audio signal processing method for the first time
In the case where number, acquire the voice of the holder of the social security card, and identify vocal print from the voice of acquisition, by the social security card with
The vocal print identified is saved, that is, vocal print of the vocal print that will identify that as the holder of the social security card.
In this step, it in the case where the vocal print identified from the first voice messaging is the vocal print of patient, executes
Otherwise S207 and S208 executes S210.
S207, identified from vocal print patient the first default vital sign current value.
In this step, the first default vital sign be indicate respiratory system default sign, such as: whether catch a cold with it is flat
Peach body whether there is inflammation.The current value of first default vital sign is the value of the default sign of respiratory system, such as, if
Flu as a result, and tonsillotome whether the result etc. of inflammation.When people's flu or respiratory system related disease, sound
Vocal print feature will change, and relevant vocal print training set is labeled, and by machine learning or deep learning, can obtain
One intelligentized identification model.Related training below is similar, after obtaining a reasonable data set, could be formed with effect
Intelligent recognition model.
S208, obtain patient the second default vital sign current value.
In this step, the second default vital sign can be the sign of expression body temperature and preset facial characteristics, wherein
Preset facial characteristics can be skin color and skin with the presence or absence of erythema etc..The current value of second default vital sign indicates
The current results value of second default vital sign, for example, body temperature as a result, the colour of skin as a result, and skin with the presence or absence of erythema
As a result.
This step can obtain the current value of the second default vital sign from default equipment, wherein default equipment can be
Temperature sensor and face sensor.Wherein, temperature sensor can be remote temperature sensor.Specifically, being passed by body temperature
Sensor measures body temperature, measures facial characteristics by face sensor.In this step, patient can be obtained from temperature sensor
Body temperature current value, from face sensor obtain patient facial region's feature current value, obtain the second default life entity of patient
The current value of sign.
S209, it is given birth to using the symptom information of patient, history information, the current value of the first default vital sign and second are default
Order the current value construction feature vector of sign.
Specifically, the process of construction feature vector is the prior art, which is not described herein again.
It determines to register needed for patient the accuracy of department to improve preset model, it in the present embodiment, can be with
By the symptom information of patient, history information, the current value of the first default vital sign, the current value of the second default vital sign,
Whether time, place, each department number of having registered reach the end value construction feature vector of preset threshold.
It should be noted that the meaning that the items for construction feature vector respectively indicate, can refer to the corresponding reality of Fig. 1
The meaning provided in example is applied, which is not described herein again.
After having executed this step, then execute S211.
S210, by the symptom information of patient and history information construction feature vector.
Specifically, the process of construction feature vector is the prior art, which is not described herein again.
It determines to register needed for patient the accuracy of department to improve preset model, it in the present embodiment, can be with
Whether the symptom information of patient, history information, time, place, preset each department number of having registered are reached into preset threshold
End value construction feature vector.
After having executed this step, then execute S211.
S211, feature vector is inputted into preset model, first object department is in the department that obtains registering needed for patient.
In this step, preset model is the model that the corresponding embodiment training of Fig. 1 obtains.
S212, the number that patient goes to a doctor in first object department is obtained.
In this step, message can be sent into the message queue of the system of first object department, and obtains current institute
Serial number of the message of transmission in message queue, the serial number are exactly the number that patient goes to a doctor in first object department.
After having executed this step, then execute S216.
S213, the second voice messaging of acquisition.
In the case where being not received by the first preset instructions, this step is executed, in the present embodiment, default is receiving
In the case where second preset instructions, this step is executed.Specifically, including the department to register needed for patient in the second voice messaging
Information.Specifically, the process of the second voice messaging of acquisition is the prior art, which is not described herein again.
S214, the second voice messaging is identified, obtaining the department to register needed for patient is the second target department.
Specifically, this step handles the second voice messaging, second voice messaging that obtains that treated, and from processing
The department to register needed for patient is identified in the second voice messaging afterwards.Specifically, to the treatment process of the second voice messaging with it is right
The treatment process of first voice messaging is identical, and which is not described herein again.
S215, the number that patient goes to a doctor in the second target department is obtained.
In this step, message can be sent into the message queue of the system for the department that registers needed for patient, and obtained
Serial number of the currently transmitted message in message queue, the serial number are exactly that patient is medical in the required department that registers
Number.
S216, output number.
Display number, can also print List of Documents.
The embodiment of the present application has the advantages that
Beneficial effect one,
The embodiment of the present application gives the symptom information that patient is obtained by way of speech recognition, and combines the disease of patient
History information determines the department that registers needed for patient, i.e. first object department using preset model.Also, in first object department
It registers for patient.Therefore, the application can determine the department that registers needed for patient, and be patient in the required section that registers
Room is registered, and therefore, user only needs once to be lined up can be completed to register, to avoid bright by being lined up twice in the prior art
Really required register and is registered at department, and then reduces queuing time, is registered the spent time to reduce user.
Beneficial effect two,
In the embodiment of the present application, the voice of acquisition is handled, specifically, processing include orientation speech gain and/
Or orientation noise reduction gain, the voice that obtains that treated, so that identifying required information from treated voice.So that in hospital
In environment noisy in this way, increase the probability that required information is recognized accurately from the voice of user.
Fig. 3 is a kind of processing unit of voice signal provided by the embodiments of the present application, comprising: acquisition module 301, identification mould
Block 302, building module 304, input module 305, obtains module 306 and output module 307 at read module 303.
Wherein, acquisition module 301 is used in the case where receiving the first preset instructions, acquires the first voice messaging, the
One voice messaging includes the symptom information of patient.Identification module 302 obtains patient's for identifying to the first voice messaging
Symptom information.Read module 303 from the social security card of patient for reading the history information of patient.Module 304 is constructed for extremely
The symptom information and history information construction feature vector of major general patient.Input module 305 is used to inputting feature vector into default mould
Type, first object department is in the department that obtains registering needed for patient, and preset model is at least with the symptom information of historic patient and disease
History information, and the department that registers of mark is that training sample training obtains.Module 306 is obtained for obtaining patient in first object
The medical number of department.Output module 307 is for exporting number.
Optionally, acquisition module 301 is also used to after acquiring the first voice messaging, the identification sound from the first voice messaging
Line.Identification module 302 is also used in the case where vocal print is the vocal print of patient, and the first default life of patient is identified from vocal print
The current value of sign is ordered, the first default vital sign is the default sign for indicating respiratory system, also, the second of acquisition patient is pre-
If the current value of vital sign, the second default vital sign is the sign for indicating body temperature and preset facial characteristics.
Module 304 is constructed at least by the symptom information of patient and history information construction feature vector, comprising:
Construct module 304 be specifically used for by the symptom information of patient, history information, patient the first default vital sign
The current value of the second default vital sign of current value and patient, construction feature vector.
Optionally, the output layer of preset model includes preset regression function, and regression function exports the preset section that respectively registers
Room is respectively the probability of department of registering needed for patient, and probability is greater than predetermined probabilities threshold value to output layer and number of having registered does not reach
The department that registers to preset threshold is first object department.
Optionally, identification module 302 obtains the symptom information of patient for identifying to the first voice messaging, comprising:
Identification module 302 is specifically used for carrying out gain and/or noise reduction to the first voice messaging, first language that obtains that treated
Message breath, to treated, the first voice messaging is identified, obtains the symptom information of patient.
Optionally, device further include: processing module 308 is used in the case where receiving the second preset instructions, acquisition
Second voice messaging includes the information of the department to register needed for patient in the second voice messaging, knows to the second voice messaging
Not, obtaining the department to register needed for patient is the second target department, obtains the number that patient goes to a doctor in the second target department, output
Number.
Fig. 4 is a kind of speech signal processing system provided by the embodiments of the present application, comprising: processor.
Processor is used in the case where receiving the first preset instructions, acquires the first voice messaging, the first voice messaging
Symptom information including patient identifies the first voice messaging, obtains the symptom information of patient, from the social security card of patient
The history information for reading patient, at least by the symptom information of patient and history information construction feature vector, from the social security card of patient
Middle reading inputs preset model, and for first object department, preset model is at least suffered from history for the department that obtains registering needed for patient
The symptom information and history information of person, and the department that registers of mark is that training sample training obtains, and obtains patient in the first mesh
Mark the medical number of department, output number.
Optionally, system further include: remote temperature sensor and face sensor, remote temperature sensor and face sense
Device is connected with processor respectively.
Remote temperature sensor, for measuring body temperature.Face sensor, for extracting facial characteristics.Processor is also used to
After receiving the first voice messaging, vocal print is identified from the first voice messaging, in the case where vocal print is the vocal print of patient, from
The current value of the first default vital sign of patient is identified in vocal print, the first default vital sign is the pre- of expression respiratory system
If sign, and, from remote temperature sensor and face sensor, extract the current of the second default vital sign of patient
Value, the second default vital sign are the sign for indicating body temperature and preset facial characteristics.
Processor, at least by the symptom information of patient and history information construction feature vector, comprising:
Processor, specifically for by the symptom information of patient, history information, patient the first default vital sign it is current
The current value of the second default vital sign of value and patient, construction feature vector.
Optionally, the output layer of preset model includes preset regression function, and regression function exports the preset section that respectively registers
Room is respectively the probability of department of registering needed for patient, and probability is greater than predetermined probabilities threshold value to output layer and number of having registered does not reach
The department that registers to preset threshold is first object department.
Optionally, processor is also used in the case where receiving the second preset instructions, the second voice messaging of acquisition, and second
Include the information of the department to register needed for patient in voice messaging, the second voice messaging is identified, obtains hanging needed for patient
Number department be the second target department, registered in second target department for patient, obtain patient in the second target family
The medical number in room, output number.
If function described in the embodiment of the present application method is realized in the form of SFU software functional unit and as independent production
Product when selling or using, can store in a storage medium readable by a compute device.Based on this understanding, the application is real
The part for applying a part that contributes to existing technology or the technical solution can be embodied in the form of software products,
The software product is stored in a storage medium, including some instructions are used so that a calculating equipment (can be personal meter
Calculation machine, server, mobile computing device or network equipment etc.) execute each embodiment the method for the application whole or portion
Step by step.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), with
Machine accesses various Jie that can store program code such as memory (RAM, Random Access Memory), magnetic or disk
Matter.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other
The difference of embodiment, same or similar part may refer to each other between each embodiment.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the application
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (10)
1. a kind of audio signal processing method characterized by comprising
In the case where receiving the first preset instructions, the first voice messaging is acquired；First voice messaging includes patient's
Symptom information；
First voice messaging is identified, the symptom information of the patient is obtained；
The history information of the patient is read from the social security card of the patient；
At least by the symptom information of the patient and history information construction feature vector；
Described eigenvector is inputted into preset model, first object department is in the department that obtains registering needed for the patient；It is described
Preset model is at least with the symptom information of historic patient and history information, and the department that registers of mark is that training sample is trained
It arrives；
Obtain the number that the patient goes to a doctor in the first object department；
Export the number.
2. the method according to claim 1, wherein after the first voice messaging of the acquisition, further includes:
Vocal print is identified from first voice messaging；
In the case where the vocal print is the vocal print of the patient, the first default life of the patient is identified from the vocal print
Order the current value of sign；The first default vital sign is the default sign for indicating respiratory system；Also, obtain the patient
The second default vital sign current value；The second default vital sign is the body for indicating body temperature and preset facial characteristics
Sign；
It is described at least by the symptom information of the patient and history information construction feature vector, comprising:
By the symptom information of the patient, history information, the current value of the first default vital sign of the patient and described
The current value of the default vital sign of the second of patient, construction feature vector.
3. the method according to claim 1, wherein the output layer of the preset model includes preset recurrence letter
Number；It is respectively the probability of department of registering needed for the patient that the regression function, which exports the preset department that respectively registers, described defeated
Probability is greater than predetermined probabilities threshold value by layer out and the number of having registered is not up to the department that registers of preset threshold as the first object
Department.
4. being obtained the method according to claim 1, wherein described identify first voice messaging
The symptom information of the patient, comprising:
Gain and/or noise reduction are carried out to first voice messaging, first voice messaging that obtains that treated；
Treated that the first voice messaging is identified to described, obtains the symptom information of the patient.
5. the method according to claim 1, wherein further include:
In the case where receiving the second preset instructions, the second voice messaging is acquired；It include described in second voice messaging
The information of the department to register needed for patient；
Second voice messaging is identified, obtaining the department to register needed for the patient is the second target department；
Obtain the number that the patient goes to a doctor in second target department；
Export the number.
6. a kind of speech signal processing device characterized by comprising
Acquisition module, for acquiring the first voice messaging in the case where receiving the first preset instructions；The first voice letter
Breath includes the symptom information of patient；
Identification module obtains the symptom information of the patient for identifying to first voice messaging；
Read module, for reading the history information of the patient from the social security card of the patient；
Module is constructed, at least by the symptom information of the patient and history information construction feature vector；
Input module, for described eigenvector to be inputted preset model, the department that obtains registering needed for the patient is first
Target department；The preset model is at least with the symptom information of historic patient and history information, and the department that registers marked is
Training sample training obtains；
Module is obtained, the number gone to a doctor for obtaining the patient in the first object department；
Output module, for exporting the number.
7. a kind of speech signal processing system characterized by comprising processor；
The processor, for acquiring the first voice messaging in the case where receiving the first preset instructions；First voice
Information includes the symptom information of patient；
First voice messaging is identified, the symptom information of the patient is obtained；
The history information of the patient is read from the social security card of the patient；
At least by the symptom information of the patient and history information construction feature vector；
Input preset model is read from the social security card of the patient, first object is in the department that obtains registering needed for the patient
Department；The preset model is at least with the symptom information of historic patient and history information, and the department that registers of mark is training
Sample training obtains；
Obtain the number that the patient goes to a doctor in the first object department；
Export the number.
8. system according to claim 7, which is characterized in that the system also includes: remote temperature sensor and face
Sensor；The remote temperature sensor is connected with the processor respectively with the face sensor；
The remote temperature sensor, for measuring body temperature；
The face sensor, for extracting facial characteristics；
The processor is also used to after receiving the first voice messaging, vocal print is identified from first voice messaging, in institute
In the case where stating the vocal print that vocal print is the patient, the first default vital sign of the patient is identified from the vocal print
Current value；The first default vital sign is the default sign for indicating respiratory system；And from the remote temperature sensor
In the face sensor, the current value of the second default vital sign of the patient is extracted；The second default life entity
Sign is to indicate the sign of body temperature and preset facial characteristics；
The processor, at least by the symptom information of the patient and history information construction feature vector, comprising:
The processor, specifically for by the symptom information of the patient, history information, the patient the first default life entity
The current value of the second default vital sign of the current value of sign and the patient, construction feature vector.
9. system according to claim 7, which is characterized in that the output layer of the preset model includes preset recurrence letter
Number；It is respectively the probability of department of registering needed for the patient that the regression function, which exports the preset department that respectively registers, described defeated
Probability is greater than predetermined probabilities threshold value by layer out and the number of having registered is not up to the department that registers of preset threshold as the first object
Department.
10. system according to claim 7, which is characterized in that
The processor is also used in the case where receiving the second preset instructions, acquires the second voice messaging；Second language
It include the information of the department to register needed for the patient in message breath；Second voice messaging is identified, is obtained described
The department to register needed for patient is the second target department；It is registered for the patient in second target department, obtains institute
State the number that patient goes to a doctor in second target department；Export the number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910757443.8A CN110473616B (en) | 2019-08-16 | 2019-08-16 | Voice signal processing method, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910757443.8A CN110473616B (en) | 2019-08-16 | 2019-08-16 | Voice signal processing method, device and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110473616A true CN110473616A (en) | 2019-11-19 |
| CN110473616B CN110473616B (en) | 2022-08-23 |
Family
ID=68511790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910757443.8A Active CN110473616B (en) | 2019-08-16 | 2019-08-16 | Voice signal processing method, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110473616B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111415449A (en) * | 2020-03-17 | 2020-07-14 | 北京声智科技有限公司 | Registration method, electronic equipment, server and registration system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105913846A (en) * | 2016-05-25 | 2016-08-31 | 北京云知声信息技术有限公司 | Method, device and system for realizing voice registration |
| CN108053841A (en) * | 2017-10-23 | 2018-05-18 | 平安科技（深圳）有限公司 | The method and application server of disease forecasting are carried out using voice |
| CN108550394A (en) * | 2018-03-12 | 2018-09-18 | 广州势必可赢网络科技有限公司 | A kind of method and device of diagnosing a disease based on Application on Voiceprint Recognition |
| CN108735234A (en) * | 2018-07-05 | 2018-11-02 | 浙江中点人工智能科技有限公司 | A kind of device monitoring health status using voice messaging |
| CN109171644A (en) * | 2018-06-22 | 2019-01-11 | 平安科技（深圳）有限公司 | Health control method, device, computer equipment and storage medium based on voice recognition |
| CN109318239A (en) * | 2018-10-09 | 2019-02-12 | 深圳市三宝创新智能有限公司 | A kind of hospital guide robot, hospital guide's method and device |
| CN109585001A (en) * | 2017-09-29 | 2019-04-05 | 北京搜狗科技发展有限公司 | A kind of data analysing method, device, electronic equipment and storage medium |
| CN109830306A (en) * | 2019-03-13 | 2019-05-31 | 天津红康云健康科技有限公司 | A kind of intelligent sound evolution system based on sound groove recognition technology in e |
-
2019
- 2019-08-16 CN CN201910757443.8A patent/CN110473616B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105913846A (en) * | 2016-05-25 | 2016-08-31 | 北京云知声信息技术有限公司 | Method, device and system for realizing voice registration |
| CN109585001A (en) * | 2017-09-29 | 2019-04-05 | 北京搜狗科技发展有限公司 | A kind of data analysing method, device, electronic equipment and storage medium |
| CN108053841A (en) * | 2017-10-23 | 2018-05-18 | 平安科技（深圳）有限公司 | The method and application server of disease forecasting are carried out using voice |
| CN108550394A (en) * | 2018-03-12 | 2018-09-18 | 广州势必可赢网络科技有限公司 | A kind of method and device of diagnosing a disease based on Application on Voiceprint Recognition |
| CN109171644A (en) * | 2018-06-22 | 2019-01-11 | 平安科技（深圳）有限公司 | Health control method, device, computer equipment and storage medium based on voice recognition |
| CN108735234A (en) * | 2018-07-05 | 2018-11-02 | 浙江中点人工智能科技有限公司 | A kind of device monitoring health status using voice messaging |
| CN109318239A (en) * | 2018-10-09 | 2019-02-12 | 深圳市三宝创新智能有限公司 | A kind of hospital guide robot, hospital guide's method and device |
| CN109830306A (en) * | 2019-03-13 | 2019-05-31 | 天津红康云健康科技有限公司 | A kind of intelligent sound evolution system based on sound groove recognition technology in e |
Non-Patent Citations (1)
| Title |
|---|
| 孙俊菲: "基于O2O的医疗服务模式创新研究", 《中国优秀硕士学位论文全文数据库 (医药卫生科技辑)》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111415449A (en) * | 2020-03-17 | 2020-07-14 | 北京声智科技有限公司 | Registration method, electronic equipment, server and registration system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110473616B (en) | 2022-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pandey | Machine learning and IoT for prediction and detection of stress | |
| JP6909078B2 (en) | Disease onset prediction device, disease onset prediction method and program | |
| Ansarullah et al. | Significance of visible non-invasive risk attributes for the initial prediction of heart disease using different machine learning techniques | |
| KR20170061222A (en) | The method for prediction health data value through generation of health data pattern and the apparatus thereof | |
| CN109817312A (en) | A kind of medical bootstrap technique and computer equipment | |
| CN110473615B (en) | Information processing method and device | |
| WO2019242155A1 (en) | Voice recognition-based health management method and apparatus, and computer device | |
| CN112382407A (en) | Risk management and control method and device, electronic equipment and storage medium | |
| US20220165417A1 (en) | Population-level gaussian processes for clinical time series forecasting | |
| CN107590473A (en) | A kind of human face in-vivo detection method, medium and relevant apparatus | |
| CN111985584A (en) | Disease auxiliary detection equipment, method, device and medium based on multi-mode data | |
| Singla et al. | A novel fuzzy logic-based medical expert system for diagnosis of chronic kidney disease | |
| CN114783580B (en) | Medical data quality evaluation method and system | |
| CN108875509A (en) | Biopsy method, device and system and storage medium | |
| CN116092681A (en) | Method, system, electronic device and storage medium for determining health index score | |
| WO2023097780A1 (en) | Classification method and device for classifying patient‑ventilator asynchrony phenomenon in mechanical ventilation process | |
| CN110473616A (en) | A kind of audio signal processing method, apparatus and system | |
| Sim et al. | Improving the accuracy of erroneous-plan recognition system for Activities of Daily Living | |
| Yang et al. | More to less (M2L): Enhanced health recognition in the wild with reduced modality of wearable sensors | |
| JP2024513618A (en) | Methods and systems for personalized prediction of infections and sepsis | |
| Arowolo et al. | Detection of COVID-19 from Chest X-Ray Images using CNN and ANN Approach | |
| Minkin et al. | Behavioral parameters as COVID-19 signs. New opportunities and old problems of medical diagnostics | |
| Srinivas et al. | Artificial intelligence based forecasting techniques for the covid-19 pandemic | |
| Joshi et al. | Analysis of Heart Disease Prediction using Various Machine Learning Techniques: A Review Study | |
| Al-Sagheer et al. | Data Mining and RBF Neural Networks to Analyze Data from COVID-19 Patients and Predict New Cases Based on Symptoms |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |