CN110473616B - Voice signal processing method, device and system - Google Patents
Voice signal processing method, device and system Download PDFInfo
- Publication number
- CN110473616B CN110473616B CN201910757443.8A CN201910757443A CN110473616B CN 110473616 B CN110473616 B CN 110473616B CN 201910757443 A CN201910757443 A CN 201910757443A CN 110473616 B CN110473616 B CN 110473616B
- Authority
- CN
- China
- Prior art keywords
- patient
- preset
- information
- department
- vital sign
- 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.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 11
- 208000024891 symptom Diseases 0.000 claims abstract description 81
- 239000013598 vector Substances 0.000 claims abstract description 43
- 238000012549 training Methods 0.000 claims abstract description 39
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 22
- 230000001815 facial effect Effects 0.000 claims description 22
- 230000036760 body temperature Effects 0.000 claims description 19
- 210000002345 respiratory system Anatomy 0.000 claims description 13
- 238000010276 construction Methods 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 description 21
- 238000003062 neural network model Methods 0.000 description 11
- 230000008569 process Effects 0.000 description 11
- 206010015150 Erythema Diseases 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 231100000321 erythema Toxicity 0.000 description 4
- 210000002741 palatine tonsil Anatomy 0.000 description 4
- 206010061218 Inflammation Diseases 0.000 description 2
- 230000004054 inflammatory process Effects 0.000 description 2
- 230000001755 vocal effect Effects 0.000 description 2
- 241000212749 Zesius chrysomallus Species 0.000 description 1
- 230000002567 autonomic effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000010801 machine learning 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
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 206010044008 tonsillitis Diseases 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 238000004383 yellowing Methods 0.000 description 1
Images
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
The application provides a voice signal processing method, a voice signal processing device and a voice signal processing system, wherein the method comprises the following steps: under the condition that a first preset instruction is received, first voice information is collected, the first voice information comprises symptom information of a patient, and the first voice information is identified to obtain the symptom information of the patient. Reading the medical history information of the patient from the social security card of the patient, at least constructing a characteristic vector of the symptom information and the medical history information of the patient, inputting the characteristic vector into a preset model, taking a registered department required by the patient as a first target department, training the preset model by taking the symptom information and the medical history information of the historical patient and the marked registered department as training samples, acquiring the number of the patient in the first target department for treatment, and outputting the number. The required registration department of patient can be accurately determined to this application.
Description
Technical Field
The present application relates to the field of electronic information, and in particular, to a method, an apparatus, and a system for processing a voice signal.
Background
In the hospital registration process, the patient is not clearly registered in the department, which is needed in some cases, for example, as the division of hospital departments is more and more detailed as the times develop, or the symptoms of the patient are not obvious, so that the patient is not clearly registered in the department. Currently, registration can be performed on an autonomous registration machine.
However, the registration accuracy of the autonomous registration machine is low, that is, the probability that the department, which is hung by the autonomous registration machine, is adopted by the user as the department, which the user actually should register, is low.
Disclosure of Invention
The application provides a voice signal processing method, a voice signal processing device and a voice signal processing system, and aims to solve the problem of low registration accuracy of an autonomous registration machine.
In order to achieve the above object, the present application provides the following technical solutions:
the application provides a voice signal processing method, which comprises the following steps:
under the condition of receiving a first preset instruction, acquiring first voice information; the first voice information includes symptom information of the patient;
recognizing the first voice information to obtain symptom information of the patient;
reading medical history information of the patient from the social security card of the patient;
constructing a feature vector from at least symptom information and medical history information of the patient;
inputting the feature vector into a preset model to obtain a registration department required by the patient as a first target department; the preset model is obtained by training at least the symptom information and the medical history information of the historical patient and the marked registration department as training samples;
acquiring the number of the patient in the first target department for treatment;
and outputting the number.
Optionally, after the acquiring the first voice information, the method further includes:
recognizing a voiceprint from the first speech information;
identifying a current value of a first preset vital sign of the patient from the voiceprint if the voiceprint is of the patient; the first preset vital sign is a preset vital sign representing a respiratory system; and obtaining a current value of a second preset vital sign of the patient; the second preset vital sign is a sign representing body temperature and a preset facial feature;
constructing a feature vector from at least symptom information and medical history information of the patient, including:
and constructing a feature vector according to the symptom information, the medical history information, the current value of the first preset vital sign of the patient and the current value of the second preset vital sign of the patient.
Optionally, the output layer of the preset model includes a preset regression function; the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the registered department of which the probability is greater than a preset probability threshold and the registered number of people does not reach the preset threshold is taken as the first target department by the output layer.
Optionally, the recognizing the first voice information to obtain the symptom information of the patient includes:
performing gain and/or noise reduction on the first voice information to obtain processed first voice information;
and identifying the processed first voice information to obtain the symptom information of the patient.
Optionally, the method further includes:
collecting second voice information under the condition of receiving a second preset instruction; the second voice message comprises information of a department for which the patient needs to register;
identifying the second voice information to obtain a department needing registration of the patient as a second target department;
acquiring the number of the patient in the second target department;
and outputting the number.
The present application also provides a speech signal processing apparatus, including:
the acquisition module is used for acquiring first voice information under the condition of receiving a first preset instruction; the first voice information includes symptom information of the patient;
the recognition module is used for recognizing the first voice information to obtain the symptom information of the patient;
the reading module is used for reading the medical history information of the patient from the social security card of the patient;
the construction module is used for constructing a characteristic vector at least from the symptom information and the medical history information of the patient;
the input module is used for inputting the feature vector into a preset model to obtain a registration department required by the patient as a first target department; the preset model is obtained by training at least the symptom information and the medical history information of the historical patient and the marked registration department as training samples;
the acquiring module is used for acquiring the number of the patient in the first target department;
and the output module is used for outputting the serial number.
Optionally, the acquiring module is further configured to identify a voiceprint from the first voice information after the acquiring of the first voice information;
the identification module is further configured to identify a current value of a first preset vital sign of the patient from the voiceprint if the voiceprint is of the patient; the first preset vital sign is a preset vital sign representing a respiratory system; and obtaining a current value of a second preset vital sign of the patient; the second preset vital sign is a sign representing body temperature and a preset facial feature;
the construction module is used for constructing a feature vector at least from the symptom information and the medical history information of the patient, and comprises:
the construction module is specifically configured to construct a feature vector from the symptom information, the medical history information, the current value of the first preset vital sign of the patient, and the current value of the second preset vital sign of the patient.
Optionally, the output layer of the preset model includes a preset regression function; the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the registered department of which the probability is greater than a preset probability threshold and the registered number of people does not reach the preset threshold is taken as the first target department by the output layer.
Optionally, the recognizing module is configured to recognize the first voice information to obtain symptom information of the patient, and includes:
the recognition module is specifically configured to perform gain and/or noise reduction on the first voice information to obtain processed first voice information; and identifying the processed first voice information to obtain the symptom information of the patient.
Optionally, the method further includes: the processing module is used for acquiring second voice information under the condition of receiving a second preset instruction; the second voice message comprises information of a department for which the patient needs to register;
identifying the second voice information to obtain a department needing to be registered by the patient as a second target department;
acquiring the number of the patient in the second target department for treatment;
and outputting the number.
The present application also provides a speech signal processing system, comprising: a processor;
the processor is used for acquiring first voice information under the condition of receiving a first preset instruction; the first voice information includes symptom information of the patient;
recognizing the first voice information to obtain symptom information of the patient;
reading medical history information of the patient from the social security card of the patient;
constructing a feature vector from at least symptom information and medical history information of the patient;
reading and inputting a preset model from the social security card of the patient, and obtaining a registered department required by the patient as a first target department; the preset model is obtained by training at least the symptom information and the medical history information of the historical patient and the marked registration department as training samples;
acquiring the number of the patient in the first target department for treatment;
and outputting the number.
Optionally, the system further includes: a remote temperature sensor and a face sensor; the remote temperature sensor and the face sensor are respectively connected with the processor;
the remote temperature sensor is used for measuring body temperature;
the face sensor is used for extracting facial features;
the processor is further configured to identify a voiceprint from the first voice message after receiving the first voice message, and identify a current value of a first preset vital sign of the patient from the voiceprint if the voiceprint is the voiceprint of the patient; the first preset vital sign is a preset vital sign representing a respiratory system; and, extracting a current value of a second preset vital sign of the patient from the remote temperature sensor and the facial sensor; the second preset vital sign is a sign representing body temperature and a preset facial feature;
the processor is used for constructing a feature vector at least from symptom information and medical history information of the patient, and comprises the following steps:
the processor is specifically configured to construct a feature vector from the symptom information, the medical history information, the current value of the first preset vital sign of the patient, and the current value of the second preset vital sign of the patient.
Optionally, the output layer of the preset model includes a preset regression function; the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the registered department of which the probability is greater than a preset probability threshold and the registered number of people does not reach the preset threshold is taken as the first target department by the output layer.
Optionally, the processor is further configured to collect second voice information when a second preset instruction is received; the second voice message comprises information of a department for which the patient needs to register; identifying the second voice information to obtain a department needing registration of the patient as a second target department; registering the patient in the second target department to obtain the number of the patient visiting the second target department; and outputting the number.
In the voice signal processing method, the voice signal processing device and the voice signal processing system, under the condition that a first preset instruction is received, first voice information is collected, wherein the first voice information comprises symptom information of a patient, the symptom information of the patient is identified from the first voice information, medical history information of the patient is read from a social security card of the patient, at least characteristic vectors are constructed by the symptom information and the medical history information of the patient, the constructed characteristic vectors are input into a preset model, a registration department required by the patient is obtained and is a first target department, a number of the patient in a treatment of the first target department is obtained, and the number is output. The preset model is obtained by training at least the symptom information and the medical history information of the historical patients and the marked registration departments as training samples.
The method and the device have the advantages that the symptom information of the patient is obtained through voice recognition, and the registered department, namely the first target department, required by the patient is determined by using the preset model in combination with the medical history information of the patient. And, the number of the patient's visit in the first target department is acquired and output. Therefore, the registration department required by the patient can be determined by the method and the registration department required by the patient determined by the model has certain accuracy. And the patient's number of seeing a doctor in the required department of registering is obtained, and the patient registers in required department of registering promptly, consequently, has reduced the user and has not definited the wrong department of registering that leads to of actually required department of registering, and then, can improve the accuracy that adopts the autonomic machine of registering of this application scheme to register.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic diagram of a training process of a preset model disclosed in an embodiment of the present application;
fig. 2 is a flowchart of a speech signal processing method disclosed in an embodiment of the present application;
fig. 3 is a schematic structural diagram of a speech signal processing apparatus according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a speech signal processing system according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The inventor finds in research that the reason that the registration accuracy of the existing autonomous registration machine is low is that: the existing autonomous registration machine cannot accurately predict the department (registration department according with the symptoms of the patient) where the patient should be actually registered. Therefore, in the embodiment of the application, the voice signal processing device is integrated in the autonomous registration machine, so that the integrated autonomous registration machine determines the symptom information of the patient according to the collected voice, the medical history information of the patient is acquired from the social security card of the patient, the symptom information and the medical history information of the patient are input into the preset model, and the registration department required by the patient is obtained, wherein the preset model is obtained by training the symptom information and the medical history information of the historical patient and the marked registration department for the training sample, so that the registration department required by the patient determined by the preset model is ensured to have certain accuracy, and the registration accuracy is improved.
Fig. 1 is a schematic diagram of a training process of a preset model according to an embodiment of the present application, including the following steps:
and S101, obtaining a training sample.
In this embodiment, the training samples at least include: symptom information of a historic patient, medical history information of a historic patient, and a registered department of a historic patient.
In order to make the accuracy of the registration department required by the patient output by the trained model higher, in this embodiment, the training sample may further include: the registration time of the historical patient, the registration place of the historical patient, the value of a first preset vital sign of the historical patient, the value of a second preset vital sign of the historical patient, and the registration condition of each registration department preset during registration of the historical patient.
Wherein, the registration time can be the season of the user when registering, for example, summer. The registered location may be a region where the user is located, e.g., XX city XX county, XX province, etc.
The first preset vital sign is a preset sign indicative of the respiratory system, for example: whether there is a cold and inflammation of the tonsils. The value of the first preset vital sign is a value of a preset vital sign of the respiratory system, for example, whether the result is a cold (e.g., cold), whether the tonsil is inflamed (e.g., inflammation of tonsil), and the like.
The second preset vital sign is a sign representing body temperature and a preset facial feature, wherein the preset facial feature can be skin color, whether red spots exist in the skin, and the like. The second predetermined vital sign is a result of the second predetermined vital sign, such as a result of body temperature (e.g., body temperature of 37 degrees celsius), a result of skin color (e.g., yellowing), and a result of whether erythema is present on the skin (e.g., no erythema on the skin).
The preset registration condition of each registration department can include: the number of registered personnel of each registered department is preset, and whether the number of registered personnel reaches a preset threshold value or not is judged. The preset registration department can be a registration department owned by a hospital, specifically, the specific content of the preset registration department can be set according to actual requirements, and the specific content of the preset registration department is not limited in the embodiment.
The preset threshold refers to an upper limit value of the number of registered people that can be accepted in the registration department, specifically, a specific value of the preset threshold may be set according to an actual situation, and the specific value of the preset threshold is not limited in this embodiment.
S102, inputting the training sample into a preset model for training to obtain the trained preset model.
In this embodiment, the preset model may be a deep neural network model, and certainly, may also be other models, and the specific content of the preset model is not limited in this embodiment.
Under the condition that the preset model is the deep neural network model, the deep neural network model comprises an output layer, the output layer comprises a preset regression function, after the characteristic vector used for predicting the registration departments needed by the patient is input into the deep neural network model, the regression function outputs the probability that the patient needs to register in each preset department, and the output layer of the deep neural network model outputs the registration departments with the probability being greater than the preset probability threshold value and the registered number not reaching the preset threshold value.
For example, the training samples include symptom information of the historical patient, medical history information of the historical patient, registration time of the historical patient, registration location of the historical patient, value of a first preset vital sign of the historical patient, value of a second preset vital sign of the historical patient, registration condition of each registration department preset during registration of the historical patient, and registration department of the historical patient.
The process of training the deep neural network model by using the training samples can comprise the following steps:
in this embodiment, firstly, the symptom information of the historical patient, the medical history information of the historical patient, the registration time of the historical patient, the registration location of the historical patient, the value of the first preset vital sign of the historical patient, the value of the second preset vital sign of the historical patient, and the registration condition of each registration department when the historical patient registers in the training sample are constructed as the feature vector.
Secondly, inputting the constructed feature vector into a deep neural network, and outputting the probability that each preset registration department is the registration department required by the historical patient by the regression function of the output layer of the deep neural network, namely determining the probability that each preset registration department in the registration departments is the registration department required by the historical patient.
Thirdly, the output layer of the deep neural network outputs registered departments with the probabilities larger than a preset probability threshold value and the number of registered people not reaching the preset threshold value in each registered department output by the regression function.
Fourthly, after the registered departments required by the historical patients are output by the output layer of the deep neural network, loss function values between the registered departments output by the output layer and the registered departments marked in the training samples of the historical patients are calculated through a preset loss function. And adjusting the value of each parameter matrix in the deep neural network model according to the loss function value.
Fifthly, by adopting the four training methods from the first step to the fourth step, the deep neural network model is circularly trained through a plurality of training samples, and the trained deep neural network model is obtained.
The beneficial effects of this embodiment include:
in the prior art, an output layer of a neural network model predicts one registered department by using a classification function, and under the condition that the number of registered people in the predicted registered department reaches a preset threshold value, it indicates that no department which can be registered by a patient exists currently.
In this embodiment, the output layer of the deep neural network model includes a preset regression function, each preset registered department output by the regression function serves as the probability of the department required to be registered by the patient, and the output layer outputs registered departments with the probabilities greater than a preset probability threshold and with the registered number of people not reaching the preset threshold, so that the number of registered departments with the probabilities greater than the preset threshold may be multiple.
Fig. 2 is a speech signal processing method according to an embodiment of the present application, including the following steps:
s201, judging whether a first preset instruction is received or not, if so, executing S202, and if not, executing S213.
In this embodiment, the first preset instruction represents a registration department where the user does not clearly specify the patient needs, wherein the user may be the patient himself or a person who helps the patient to register. In this embodiment, two button indicia may be displayed on the autonomous registration machine, one button indicating a clearly registered department and the other button indicating a not clearly registered department. And the user selects the button to be clicked according to the actual situation. In this step, whether a first preset instruction is received or not is judged by clicking an instruction triggered by a certain button by a user. It should be noted that, this step is only to provide one implementation manner, and in practice, the step may also be implemented by other manners, and this embodiment does not limit the specific implementation manner.
S202, collecting first voice information.
In this embodiment, the first speech information includes symptom information of the patient.
And S203, recognizing the first voice information to obtain the symptom information of the patient.
Specifically, in order to ensure that the symptom information of the patient can be identified from the first speech information collected in the noisy environment, in this embodiment, the collected first speech information is processed to obtain the processed first speech information, and the symptom information of the patient is identified from the processed first speech information.
Wherein, the processing of the collected first voice information comprises: and carrying out directional voice gain on the first voice information, and/or carrying out directional noise reduction gain. The directional noise reduction gain may be specifically a DOA directional noise reduction gain. Specifically, the implementation processes of the directional speech gain and the directional noise reduction gain are all the prior art, and are not described herein again.
And S204, acquiring medical history information of the patient.
In this embodiment, in the process of using the device (for example, a self-help registration machine) integrating the voice signal processing method provided by the present invention, the user may insert the social security card of the patient first. In this step, the patient's medical history information may be read from the social security card.
And S205, identifying the voiceprint from the first voice information.
Specifically, a MFCC voiceprint recognition algorithm may be used to recognize a voiceprint from the first speech information, and specifically, the process of recognizing a voiceprint is the prior art and is not described herein again.
S206, judging whether the identified voiceprint is the voiceprint of the patient, if so, executing S207, and if not, executing S210.
In this embodiment, when the social security card is first inserted into the device for integrating the voice signal processing method for applying for a request for registration, the voice of the holder of the social security card is collected, the voiceprint is identified from the collected voice, and the social security card and the identified voiceprint are stored, that is, the identified voiceprint is used as the voiceprint of the holder of the social security card.
In this step, in the case where the voiceprint recognized from the first speech information is the voiceprint of the patient, S207 and S208 are executed, otherwise, S210 is executed.
And S207, identifying the current value of the first preset vital sign of the patient from the voiceprint.
In this step, the first preset vital sign is a preset sign representative of the respiratory system, for example: whether there is a cold and inflammation of the tonsils. The current value of the first preset vital sign is a value of a preset vital sign of the respiratory system, for example, a result of whether the tonsil is inflamed or not, and the like. When a person feels common or diseases related to the respiratory system, the vocal print characteristics of the voice of the person change, the related vocal print training set is labeled, and an intelligent recognition model can be obtained through machine learning or deep learning. The following related training is similar, and after a reasonable data set is obtained, an effective intelligent recognition model can be formed.
And S208, acquiring the current value of the second preset vital sign of the patient.
In this step, the second preset vital sign may be a sign indicating body temperature and a preset facial feature, where the preset facial feature may be skin color, whether erythema exists on the skin, and the like. The current value of the second preset vital sign represents a current result value of the second preset vital sign, for example, a result of body temperature, a result of skin color, and a result of whether erythema is present on the skin.
This step may obtain a current value of a second preset vital sign from a preset device, where the preset device may be a temperature sensor and a facial sensor. Wherein the temperature sensor may be a remote temperature sensor. Specifically, the body temperature is measured by a body temperature sensor, and the facial features are measured by a facial sensor. In this step, a current value of the body temperature of the patient may be obtained from the temperature sensor, a current value of the facial feature of the patient may be obtained from the facial sensor, and a current value of the second preset vital sign of the patient may be obtained.
S209, constructing a feature vector by using the symptom information, the medical history information, the current value of the first preset vital sign and the current value of the second preset vital sign of the patient.
Specifically, the process of constructing the feature vector is the prior art, and is not described herein again.
In order to improve the accuracy of the preset model for determining the registered departments required by the patient, in this embodiment, a feature vector may be constructed by using the symptom information, the medical history information, the current value of the first preset vital sign, the current value of the second preset vital sign, the time, the location, and the result value of whether the registered number of the patient in each department reaches the preset threshold value.
It should be noted that the meanings respectively expressed by the terms for constructing the feature vector may refer to the meanings given in the embodiment corresponding to fig. 1, and are not described herein again.
After the present step is executed, S211 is executed again.
S210, constructing a feature vector by using the symptom information and the medical history information of the patient.
Specifically, the process of constructing the feature vector is the prior art, and is not described herein again.
In order to improve the accuracy of the preset model for determining the registered departments required by the patient, in this embodiment, the feature vector may be further constructed by using the result values of whether the symptom information, the medical history information, the time and the location of the patient and the number of registered people in each preset department reach the preset threshold value.
After the present step is executed, S211 is executed again.
S211, inputting the feature vector into a preset model, and taking the registration department required by the patient as a first target department.
In this step, the preset model is the model obtained by training in the embodiment corresponding to fig. 1.
S212, acquiring the number of the patient in the first target department.
In this step, a message may be sent to the message queue of the system of the first target department, and the sequence number of the currently sent message in the message queue is obtained, where the sequence number is the number of the patient visiting the first target department.
After the present step is executed, S216 is executed again.
And S213, collecting second voice information.
In this embodiment, the step is executed by default when the second preset instruction is received. Specifically, the second voice message includes information of a department for which the patient needs to be registered. Specifically, the process of acquiring the second speech information is the prior art, and is not described herein again.
S214, the second voice information is recognized, and the department with the registration required by the patient is the second target department.
Specifically, the step processes the second voice message to obtain the processed second voice message, and identifies the department of the patient who needs to register from the processed second voice message. Specifically, the processing procedure of the second speech information is the same as the processing procedure of the first speech information, and is not described herein again.
S215, acquiring the number of the patient in the second target department.
In this step, the message may be sent to the message queue of the system of the registered department required by the patient, and the sequence number of the currently sent message in the message queue is obtained, where the sequence number is the number of the patient for the visit in the required registered department.
And S216, outputting the number.
The number is displayed, and a number sheet can also be printed.
The embodiment of the application has the following beneficial effects:
has the beneficial effects of,
The embodiment of the application provides a method for obtaining symptom information of a patient in a voice recognition mode, and determining a registration department, namely a first target department, required by the patient by using a preset model in combination with medical history information of the patient. And, the patient is registered at the first target department. Therefore, the method and the device can determine the registration department required by the patient and register the patient in the required registration department, so that the user can complete registration only by queuing once, the condition that the required registration department and registration are determined by queuing twice in the prior art is avoided, the queuing time is further reduced, and the time consumed by registration of the user is reduced.
Has the beneficial effects of,
In the embodiment of the present application, the collected speech is processed, specifically, the processing includes directional speech gain and/or directional noise reduction gain, to obtain processed speech, so that the required information is identified from the processed speech. So that in a noisy environment like a hospital, the probability of accurately recognizing the desired information from the user's voice is increased.
Fig. 3 is a processing apparatus of a speech signal according to an embodiment of the present application, including: an acquisition module 301, an identification module 302, a reading module 303, a construction module 304, an input module 305, an acquisition module 306 and an output module 307.
The acquisition module 301 is configured to acquire first voice information under the condition that a first preset instruction is received, where the first voice information includes symptom information of a patient. The recognition module 302 is configured to recognize the first voice message to obtain symptom information of the patient. The reading module 303 is used for reading medical history information of a patient from a social security card of the patient. The construction module 304 is configured to construct feature vectors from at least symptom information and medical history information of the patient. The input module 305 is configured to input the feature vector into a preset model, where the obtained registration department required by the patient is a first target department, and the preset model is obtained by training a training sample with at least symptom information and medical history information of the historical patient and the labeled registration department. The obtaining module 306 is used for obtaining the number of the patient's visit in the first target department. The output module 307 is used for outputting the number.
Optionally, the collecting module 301 is further configured to identify a voiceprint from the first voice information after collecting the first voice information. The identification module 302 is further configured to identify a current value of a first preset vital sign of the patient from the voiceprint when the voiceprint is the voiceprint of the patient, the first preset vital sign being a preset vital sign representing a respiratory system, and obtain a current value of a second preset vital sign of the patient, the second preset vital sign being a vital sign representing a body temperature and a preset facial feature.
The construction module 304 constructs feature vectors from at least symptom information and medical history information of the patient, including:
the construction module 304 is specifically configured to construct a feature vector from the symptom information, the medical history information, the current value of the first preset vital sign of the patient, and the current value of the second preset vital sign of the patient.
Optionally, the output layer of the preset model includes a preset regression function, the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the output layer takes the registered department with the probability greater than the preset probability threshold and the registered number of people not reaching the preset threshold as the first target department.
Optionally, the recognition module 302 is configured to recognize the first voice information to obtain the symptom information of the patient, and includes:
the recognition module 302 is specifically configured to perform gain and/or noise reduction on the first voice information to obtain processed first voice information, and recognize the processed first voice information to obtain symptom information of the patient.
Optionally, the apparatus further comprises: the processing module 308 is configured to collect second voice information under the condition that the second preset instruction is received, where the second voice information includes information of a department that needs to be registered by the patient, identify the second voice information, obtain that the department that needs to be registered by the patient is a second target department, obtain a serial number of the patient seeing a doctor in the second target department, and output the serial number.
Fig. 4 is a speech signal processing system according to an embodiment of the present application, including: a processor.
The processor is used for acquiring first voice information under the condition that a first preset instruction is received, wherein the first voice information comprises symptom information of a patient, identifying the first voice information to obtain the symptom information of the patient, reading medical history information of the patient from a social security card of the patient, at least constructing a feature vector by the symptom information and the medical history information of the patient, reading and inputting a preset model from the social security card of the patient, obtaining a registration department required by the patient as a first target department, training the preset model by taking at least the symptom information and the medical history information of the historical patient and the marked registration department as training samples, acquiring a number of the patient in the first target department for treatment, and outputting the number.
Optionally, the system further comprises: remote temperature sensor and face sensor, remote temperature sensor links to each other with the treater respectively with face sensor.
And the remote temperature sensor is used for measuring the body temperature. And the face sensor is used for extracting facial features. The processor is further used for identifying a voiceprint from the first voice information after receiving the first voice information, identifying a current value of a first preset vital sign of the patient from the voiceprint under the condition that the voiceprint is the voiceprint of the patient, wherein the first preset vital sign is a preset vital sign representing a respiratory system, and extracting a current value of a second preset vital sign of the patient from the remote temperature sensor and the face sensor, wherein the second preset vital sign is a vital sign representing body temperature and preset facial features.
A processor for constructing a feature vector from at least symptom information and medical history information of a patient, comprising:
the processor is specifically configured to construct a feature vector from the symptom information, the medical history information, the current value of the first preset vital sign of the patient, and the current value of the second preset vital sign of the patient.
Optionally, the output layer of the preset model includes a preset regression function, the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the output layer takes the registered department with the probability greater than a preset probability threshold and the number of registered people not reaching the preset threshold as the first target department.
Optionally, the processor is further configured to acquire second voice information under the condition that a second preset instruction is received, where the second voice information includes information of a department to which the patient needs to be registered, identify the second voice information, obtain that the department to which the patient needs to be registered is a second target department, register the patient in the second target department, obtain a serial number of the patient in the second target department, and output the serial number.
The functions described in the method of the embodiment of the present application, if implemented in the form of software functional units and sold or used as independent products, may be stored in a storage medium readable by a computing device. Based on such understanding, part of the technical solutions or portions of the embodiments contributing to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computing device (which may be a personal computer, a server, a mobile computing device, a network device, or the like) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A speech signal processing method, comprising:
under the condition of receiving a first preset instruction, acquiring first voice information; the first voice information includes symptom information of the patient;
recognizing the first voice information to obtain symptom information of the patient; recognizing a voiceprint from the first speech information; identifying a current value of a first preset vital sign of the patient from the voiceprint if the voiceprint is of the patient; the first preset vital sign is a preset vital sign representing a respiratory system; and obtaining a current value of a second preset vital sign of the patient; the second preset vital sign is a sign representing body temperature and a preset facial feature;
reading medical history information of the patient from the social security card of the patient;
constructing a feature vector from at least symptom information and medical history information of the patient; constructing a feature vector from at least symptom information and medical history information of the patient, including: constructing a feature vector according to the symptom information, the medical history information, the current value of the first preset vital sign of the patient and the current value of the second preset vital sign of the patient;
inputting the feature vector into a preset model to obtain a registration department required by the patient as a first target department; the preset model is obtained by training at least with symptom information and medical history information of historical patients and marked registration departments as training samples;
acquiring the number of the patient in the first target department;
and outputting the number.
2. The method of claim 1, wherein the output layer of the predetermined model comprises a predetermined regression function; the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the registered department of which the probability is greater than a preset probability threshold and the registered number of people does not reach the preset threshold is taken as the first target department by the output layer.
3. The method of claim 1, wherein the recognizing the first speech information to obtain symptom information of the patient comprises:
performing gain and/or noise reduction on the first voice information to obtain processed first voice information;
and identifying the processed first voice information to obtain the symptom information of the patient.
4. The method of claim 1, further comprising:
under the condition of receiving a second preset instruction, acquiring second voice information; the second voice message comprises information of a department for which the patient needs to register;
identifying the second voice information to obtain a department needing to be registered by the patient as a second target department;
acquiring the number of the patient in the second target department;
and outputting the number.
5. A speech signal processing apparatus, comprising:
the acquisition module is used for acquiring first voice information under the condition of receiving a first preset instruction; the first voice information includes symptom information of the patient; the acquisition module is also used for identifying the voiceprint from the first voice information after acquiring the first voice information;
the recognition module is used for recognizing the first voice information to obtain the symptom information of the patient; the identification module is further configured to identify a current value of a first preset vital sign of the patient from the voiceprint when the voiceprint is the voiceprint of the patient, the first preset vital sign being a preset sign representing a respiratory system, and acquire a current value of a second preset vital sign of the patient, the second preset vital sign being a sign representing a body temperature and a preset facial feature;
the reading module is used for reading the medical history information of the patient from the social security card of the patient;
the construction module is used for constructing a characteristic vector at least from the symptom information and the medical history information of the patient; the construction module is specifically used for constructing a feature vector from the symptom information, the medical history information, the current value of a first preset vital sign of the patient and the current value of a second preset vital sign of the patient;
the input module is used for inputting the feature vector into a preset model to obtain a registration department required by the patient as a first target department; the preset model is obtained by training at least with symptom information and medical history information of historical patients and marked registration departments as training samples;
the acquiring module is used for acquiring the number of the patient in the first target department;
and the output module is used for outputting the serial number.
6. A speech signal processing system, comprising: a processor; a remote temperature sensor and a face sensor; the remote temperature sensor and the face sensor are respectively connected with the processor;
the processor is used for acquiring first voice information under the condition of receiving a first preset instruction; the first voice information includes symptom information of the patient;
recognizing the first voice information to obtain symptom information of the patient;
reading medical history information of the patient from the social security card of the patient;
constructing a feature vector from at least symptom information and medical history information of the patient;
reading and inputting a preset model from the social security card of the patient, and obtaining a registered department required by the patient as a first target department; the preset model is obtained by training at least the symptom information and the medical history information of the historical patient and the marked registration department as training samples;
acquiring the number of the patient in the first target department;
outputting the number;
the remote temperature sensor is used for measuring body temperature;
the face sensor is used for extracting facial features;
the processor is further configured to identify a voiceprint from the first voice message after receiving the first voice message, and identify a current value of a first preset vital sign of the patient from the voiceprint if the voiceprint is the voiceprint of the patient; the first preset vital sign is a preset vital sign representing a respiratory system; and, extracting a current value of a second preset vital sign of the patient from the remote temperature sensor and the facial sensor; the second preset vital sign is a sign representing body temperature and a preset facial feature;
the processor is used for constructing a feature vector at least from symptom information and medical history information of the patient, and comprises the following steps:
the processor is specifically configured to construct a feature vector from the symptom information, the medical history information, the current value of the first preset vital sign of the patient, and the current value of the second preset vital sign of the patient.
7. The system of claim 6, wherein the output layer of the predetermined model comprises a predetermined regression function; the regression function outputs the probability that each preset registered department is the registered department required by the patient, and the registered department of which the probability is greater than a preset probability threshold and the registered number of people does not reach the preset threshold is taken as the first target department by the output layer.
8. The system of claim 6,
the processor is further used for acquiring second voice information under the condition that a second preset instruction is received; the second voice information comprises information of a department for which the patient needs to register; identifying the second voice information to obtain a department needing registration of the patient as a second target department; registering the patient in the second target department to obtain the number of the patient in the second target department; and outputting 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 CN110473616A (en) | 2019-11-19 |
CN110473616B true 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) |
Families Citing this family (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 (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
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 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105913846B (en) * | 2016-05-25 | 2019-12-06 | 北京云知声信息技术有限公司 | voice registration realization method, device and system |
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 |
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 |
-
2019
- 2019-08-16 CN CN201910757443.8A patent/CN110473616B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
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 |
---|
基于O2O的医疗服务模式创新研究;孙俊菲;《中国优秀硕士学位论文全文数据库 (医药卫生科技辑)》;20171115(第11期);第E053-42页 * |
Also Published As
Publication number | Publication date |
---|---|
CN110473616A (en) | 2019-11-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111899878B (en) | Old person health detection system, method, computer device and readable storage medium | |
CN110473615B (en) | Information processing method and device | |
US20170147777A1 (en) | Method and apparatus for predicting health data value through generation of health data pattern | |
US9861308B2 (en) | Method and system for monitoring stress conditions | |
CN109493979A (en) | A kind of disease forecasting method and apparatus based on intelligent decision | |
CN111785371A (en) | Medical service method, medical service device, computer equipment and storage medium | |
CN115132356A (en) | Internet medical triage method and device, electronic equipment and storage medium | |
CN115797868A (en) | Behavior early warning method, system, device and medium for monitoring object | |
CN113164163A (en) | Device, method and program for inferring depression state | |
JP2018195164A (en) | Analysis device, analysis program, and analysis method | |
JP2013148996A (en) | Seriousness determination device, and seriousness determination method | |
CN110473616B (en) | Voice signal processing method, device and system | |
CN110752027A (en) | Electronic medical record data pushing method and device, computer equipment and storage medium | |
CN116942149B (en) | Lumbar vertebra monitoring method, device, equipment and storage medium based on millimeter wave radar | |
CN110648754A (en) | Department recommendation method, device and equipment | |
CN112329633A (en) | Emotion identification method, device, medium and electronic equipment based on tensor decomposition | |
CN109119131B (en) | Physical examination method and system based on medical examination expert intelligence library platform | |
KR20150102880A (en) | System and method for depression monitoring based on user information | |
CN114388123A (en) | Intelligent auxiliary diagnosis method, device, equipment and storage medium | |
CN115146179A (en) | Information recommendation method and device based on cross-domain medical data and computer equipment | |
KR102179850B1 (en) | System and method for predicting health using analysis device for intraoral microbes (bacteria, virus, viroid, and/or fungi) | |
CN113095153A (en) | Mobile terminal human situation recognition method based on depth residual error network | |
CN112487980A (en) | Micro-expression-based treatment method, device, system and computer-readable storage medium | |
CN113302701A (en) | Data processing system and method for determining risk of transfer of individual to emergency department | |
CN114822830B (en) | Inquiry interaction method and related device, electronic equipment and storage medium |
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 |