CN111554398A - Remote vital sign evaluation method and system based on 5G - Google Patents
Remote vital sign evaluation method and system based on 5G Download PDFInfo
- Publication number
- CN111554398A CN111554398A CN202010390520.3A CN202010390520A CN111554398A CN 111554398 A CN111554398 A CN 111554398A CN 202010390520 A CN202010390520 A CN 202010390520A CN 111554398 A CN111554398 A CN 111554398A
- Authority
- CN
- China
- Prior art keywords
- blood pressure
- algorithm
- frequency
- layer
- network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000002360 preparation method Methods 0.000 claims abstract description 11
- 238000005516 engineering process Methods 0.000 claims abstract description 5
- 230000036772 blood pressure Effects 0.000 claims description 51
- 238000000354 decomposition reaction Methods 0.000 claims description 42
- 210000002569 neuron Anatomy 0.000 claims description 39
- 238000013528 artificial neural network Methods 0.000 claims description 22
- 230000008859 change Effects 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 15
- 238000009530 blood pressure measurement Methods 0.000 claims description 12
- 230000036760 body temperature Effects 0.000 claims description 12
- 230000000740 bleeding effect Effects 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 239000000284 extract Substances 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 9
- 230000036407 pain Effects 0.000 claims description 7
- 230000036387 respiratory rate Effects 0.000 claims description 7
- 238000012546 transfer Methods 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 210000001747 pupil Anatomy 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 210000004087 cornea Anatomy 0.000 claims description 5
- 230000009977 dual effect Effects 0.000 claims description 5
- 206010062519 Poor quality sleep Diseases 0.000 claims description 3
- 239000003086 colorant Substances 0.000 claims description 3
- 210000003128 head Anatomy 0.000 claims description 3
- 210000002364 input neuron Anatomy 0.000 claims description 3
- 230000002618 waking effect Effects 0.000 claims description 3
- 230000003321 amplification Effects 0.000 claims 4
- 238000003199 nucleic acid amplification method Methods 0.000 claims 4
- 238000011478 gradient descent method Methods 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000029142 excretion Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000474 nursing effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- 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
- G16H40/60—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 operation of medical equipment or devices
- G16H40/67—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 operation of medical equipment or devices for remote operation
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0002—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
- A61B5/0015—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system
- A61B5/0022—Monitoring a patient using a global network, e.g. telephone networks, internet
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0059—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence
- A61B5/0077—Devices for viewing the surface of the body, e.g. camera, magnifying lens
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow
- A61B5/021—Measuring pressure in heart or blood vessels
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow
- A61B5/024—Measuring pulse rate or heart rate
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7225—Details of analogue processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- 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
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Pathology (AREA)
- Heart & Thoracic Surgery (AREA)
- Veterinary Medicine (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Physiology (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Cardiology (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Vascular Medicine (AREA)
- Power Engineering (AREA)
- Computer Networks & Wireless Communication (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses a method and a system for remotely evaluating vital signs based on 5G, belonging to the field of remote medical care, aiming at solving the technical problem of how to enable medical staff to acquire patient information in time, making rescue preparation in advance according to the specific situation of a patient, improving the success rate of rescue and saving medical resources, and adopting the technical scheme that: the method utilizes a 5G network to send videos, pictures and text information acquired on site to a medical server of a hospital emergency center, and the medical server analyzes the videos, the pictures and the text information by combining an AI algorithm technology and then carries out grade evaluation on vital signs of patients, so that medical staff of the emergency center can make rescue preparation in advance according to specific conditions of the patients. The system comprises head-mounted equipment and a medical server, wherein the head-mounted equipment comprises a high-definition camera, a voice recognition and recording module and a thermal infrared imager; the high-definition camera, the voice recognition and recording module and the thermal infrared imager are respectively connected with the medical server through a 5G network.
Description
Technical Field
The invention relates to the field of remote medical care, in particular to a method and a system for remotely evaluating vital signs based on 5G.
Background
The medical emergency rescue is an important link of national economic development, and plays an important role in realizing the operation management and control modernization of the medical emergency rescue, ensuring that the wounded is treated in time, preventing the wounded from getting worse and saving the life of the wounded.
At present, medical resources are in a state of shortage all the time, especially in emergency centers of various hospitals. The key to ensure the success rate of rescue is to prepare all rescue work when the patient arrives at the emergency center. The main reason why the rescue doctor who waits in the rescue room can not understand the patient condition deeply enough can lead to insufficient rescue preparation and serious waste of medical resources is lack of real-time and reliable information of the patient condition. Therefore, how to enable medical staff to acquire patient information in time and to make rescue preparation in advance according to specific conditions of patients is a technical problem to be solved urgently at present.
Patent document No. CN109473166A discloses an intelligent remote medical care system and method based on multi-network fusion, which is provided with a processor, wherein the processor receives signals of a vital sign detection module, a video monitoring module and a calling module, and is connected with a network access module; the network access module is connected with the remote controller through a network. The invention is provided with vital sign detection, can remotely know the vital signs of a patient at any time, and can treat the patient at any time after the vital signs of the patient are abnormal; the invention is provided with the excretion processing module which can process the excretion of the patient at any time; the invention is provided with a remote controller which can remotely control the nursing process. This technical scheme delay of information probably appears, influences patient information transfer's real-time nature, and then influences medical personnel and in time obtain patient's information, probably reduces the success rate of rescuing.
Disclosure of Invention
The technical task of the invention is to provide a method and a system for remotely evaluating vital signs based on 5G, which solve the problems that how to enable medical staff to acquire patient information in time, and how to make rescue preparation in advance according to the specific conditions of patients, improve the rescue success rate and save medical resources.
The technical task of the invention is realized in the following way, a method for remotely evaluating vital signs based on 5G is characterized in that a 5G network is utilized to send videos, pictures and text information acquired on site to a medical server of a hospital emergency center, and the medical server analyzes the videos, the pictures and the text information by combining an AI algorithm technology and then carries out grade evaluation on the vital signs of patients, so that medical staff in the emergency center can make rescue preparations in advance according to the specific conditions of the patients; the method comprises the following specific steps:
s1, wearing head-mounted equipment by the on-site emergency doctor, wherein the head-mounted equipment comprises a high-definition camera, a voice recognition and recording module and a thermal infrared imager;
s2, shooting videos and pictures by the high-definition camera, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
s3, the medical server acquires heart rate signals from the videos and the pictures by combining an AI algorithm and completes blood pressure measurement;
s4, shooting and observing the head and the bleeding position of the patient by using a high-definition camera, acquiring pupil and cornea information and bleeding conditions, sending the information and the bleeding conditions to a medical server, and calculating information change and giving an evaluation result by the medical server;
s5, acquiring the body temperature of the patient in a long distance by using a thermal infrared imager, and directly sending the body temperature to a medical server of an emergency center through a 5G network;
s6, the on-site emergency doctor inputs the information of the pulse, the respiratory rate, the pain and the wakefulness of the patient by voice through the voice recognition and recording module, converts the voice into character information by the voice recognition and recording module, and then sends the character information to the medical server of the emergency center through the 5G network.
Preferably, the step S3 of acquiring the heart rate signal specifically includes:
s301-1, extracting face color change information from videos and pictures by the medical server in combination with an AI algorithm;
s301-2, amplifying the change of the human face color signal through an Eulerian Video magnetic algorithm to obtain an RGB three-channel signal;
s301-3, carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
s301-4, converting the chrominance signal into a frequency domain by Fourier transform (Fourier transform) to obtain the heart rate.
Preferably, the blood pressure measurement in step S3 is completed as follows:
s302-1, the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
s302-2, carrying out skin color segmentation and extraction through an Adaptive Boosting algorithm;
s302-3, processing the image through an Eulerian video verification algorithm (Eulerian video verification) and acquiring a human body double-path pulse signal from the image;
s302-4, completing detection of blood pressure based on a model of double blood pressure prediction of the BP neural network.
Preferably, the detection of the blood pressure by the model of dual blood pressure prediction based on the BP neural network in step S302-4 is specifically as follows:
(1) and human blood pressure sequence single-scale decomposition: using db4 wavelet as the basis function of blood pressure sequence wavelet analysis, performing low-frequency and high-frequency wavelet decomposition on the original blood pressure sequence X (t), decomposing by Mallat algorithm of wavelet transformation, filtering the original signal by a low-pass filter and a high-pass filter, and then sampling to obtain a decomposition coefficient, wherein the formula is as follows:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow and high frequency information at resolution; h represents a low-pass filter; g represents a high-pass filter; j represents the decomposition scale, J is 1, 2, …, J;
(2) reconstruction of low and high frequency components: respectively performing single-branch reconstruction on a low-frequency function and a high-frequency function of a blood pressure sequence based on wavelet basis functions selected by decomposition, and reconstructing low-frequency and high-frequency decomposition coefficients to an original scale through a wavelet algorithm to obtain a low-frequency component L (t) and a high-frequency component H (t) of blood pressure;
(3) and predicting the neural network: respectively predicting L (t) and H (t) by using a BP neural network model, and processing layer by layer from an input layer through a hidden layer until components of each layer are output by an output layer
And
(4) and predicted sequence synthesis: will be provided with
And
the sequence prediction results are superposed to obtain the predicted value of X (t)
Namely, a prediction result corresponding to the original blood pressure sequence is obtained, and the formula is as follows:
preferably, the neural network prediction in the step (3) is specifically as follows:
firstly, a prediction model with a single hidden layer BP neural network structure is arranged, the number of input layer neurons of the prediction model is 6, the number of input layer neurons is 1, and the number of hidden layer neurons is determined according to the following formula:
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
determining the optimal number of neurons in a hidden layer of the low-frequency component prediction model to be 12, the optimal number of neurons in a hidden layer of the high-frequency component prediction model to be 7, wherein the neurons in the hidden layer adopt sigmoid transform functions, the neurons in an output layer adopt linear transfer functions, a network training function adopts train lm, the functions train a forward network by utilizing a levenberg-Marquard algorithm, and the training speed of the algorithm is faster than that of a gradient descent method;
③, the input of the network is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs each layer of component
And
a remote vital sign evaluation system based on 5G comprises a head-mounted device and a medical server, wherein the head-mounted device comprises a high-definition camera, a voice recognition input module and a thermal infrared imager; the high-definition camera, the voice recognition and input module and the thermal infrared imager are respectively connected with the medical server through a 5G network;
the high-definition camera is used for shooting videos and pictures, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
the voice recognition and recording module is used for converting voice information of pulse, respiratory rate, pain and waking degree of a patient of an on-site emergency doctor into text information and sending the text information to a medical server of an emergency center through a 5G network;
the thermal infrared imager is used for obtaining the body temperature of a patient in a long distance and directly sending the body temperature to a medical server of the emergency center through a 5G network;
the medical server is used for counting all vital signs of the patient, acquiring a heart rate signal by combining an AI algorithm, completing blood pressure measurement, evaluating the degree of danger and preparing rescue in advance.
Preferably, the acquiring of the heart rate signal is specifically as follows:
the method comprises the following steps that (I) a medical server extracts face color change information from videos and pictures by combining an AI algorithm;
(II) amplifying the change of the human face color signal through an Eulerian Video magnetic algorithm to obtain an RGB three-channel signal;
(III) carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
and (IV) converting the chrominance signal into a frequency domain by Fourier transform (Fourier transform) to obtain the heart rate.
More preferably, the blood pressure measurement is specifically performed as follows:
the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
(ii) carrying out segmentation and extraction on skin colors by an Adaptive Boosting algorithm;
(iii) processing the images by an Eulerian Video magnetic algorithm and obtaining a human body two-way pulse signal therefrom;
(iv) completing the detection of the blood pressure based on a dual blood pressure prediction model of the BP neural network; the method comprises the following specific steps:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow and high frequency information at resolution; h represents a low-pass filter; g represents a high-pass filter; j is a function ofDenotes the decomposition scale, J ═ 1, 2, …, J;
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
through a trial and error method, through multiple comparisons, the optimal number of neurons in a hidden layer of a low-frequency component prediction model is determined to be 12, the optimal number of neurons in a hidden layer of a high-frequency component prediction model is determined to be 7, the neurons in the hidden layer adopt a sigmoid transformation function, the neurons in an output layer adopt a linear transfer function, a network training function adopts a train lm, the function trains a forward network by utilizing a levenberg-Marquard algorithm, and the training speed of the algorithm is faster than that of a gradient descent method;
the network input is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs components of each layer
And
the method and the system for remotely evaluating the vital signs based on the 5G have the following advantages:
based on the characteristics of low delay and high bandwidth of a 5G network, various vital signs related to a wounded person or a patient in a life critical state are sent to an emergency center in a video, picture and text information mode by using head-mounted equipment at the first time, and the vital signs of the patient are collected from the video by combining with an artificial intelligence algorithm, so that the emergency center can make rescue preparation in advance according to the specific condition of the patient; compared with the prior art, the invention can enable the rescue doctor in the emergency center to clearly know the condition of the patient, is convenient for medical personnel to make rescue preparation in advance according to the specific condition of the patient, effectively improves the working efficiency of the medical personnel in the emergency center of the hospital and greatly saves public medical resources;
by means of the 5G technology, the invention ensures that the network is more sensitive, the information transmission is quicker, the real-time performance is more excellent, the transmission efficiency of the real-time video is very high, the rescue efficiency of medical personnel is greatly improved, and a large amount of medical resources are saved;
the invention not only judges the current condition of the patient based on four vital signs, but also scores the danger level of the patient according to other vital signs such as heart rate, pupil change and the like, and helps the waiting medical personnel to prepare for rescue in advance;
according to the invention, through videos, pictures and information observed by a doctor on site, medical personnel in the emergency center can know the condition of a patient in real time, and can carry out rescue preparation as required, so that the working efficiency is improved.
And (V) the head-mounted equipment comprises a voice recognition and entry module, and because the video cannot detect all physical signs, such as pulse, respiratory rate, pain, sobriety degree and the like, the voice recognition and entry module is convenient for a doctor to convert voice into text information and send the text information to an emergency center in a voice input mode in a very short time, so that the timeliness is ensured.
Drawings
The invention is further described below with reference to the accompanying drawings.
FIG. 1 is a block flow diagram of a 5G-based method for remotely assessing vital signs;
fig. 2 is a block diagram of a 5G-based remote vital signs assessment system.
Detailed Description
A5G-based remote vital sign assessment method and system of the present invention is described in detail below with reference to the drawings and the detailed description.
Example 1:
as shown in the attached figure 1, the method for remotely evaluating the vital signs based on 5G sends video, pictures and text information acquired on site to a medical server of a hospital emergency center by using a 5G network, and the medical server analyzes the video, the pictures and the text information by combining an AI algorithm technology and then carries out grade evaluation on the vital signs of patients, so that medical staff in the emergency center can make rescue preparations in advance according to the specific conditions of the patients; the method comprises the following specific steps:
s1, wearing head-mounted equipment by the on-site emergency doctor, wherein the head-mounted equipment comprises a high-definition camera, a voice recognition and recording module and a thermal infrared imager;
s2, shooting videos and pictures by the high-definition camera, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
s3, the medical server acquires heart rate signals from the videos and the pictures by combining an AI algorithm and completes blood pressure measurement; the acquisition of the heart rate signal is specifically as follows:
s301-1, extracting face color change information from videos and pictures by the medical server in combination with an AI algorithm;
s301-2, amplifying the change of the human face color signal through an Eulerian Video magnetic algorithm to obtain an RGB three-channel signal;
s301-3, carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
s301-4, converting the chrominance signal into a frequency domain by Fourier transform (Fourier transform) to obtain the heart rate.
The blood pressure measurement is accomplished as follows:
s302-1, the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
s302-2, carrying out skin color segmentation and extraction through an Adaptive Boosting algorithm;
s302-3, processing the image through an Eulerian video verification algorithm (Eulerian video verification) and acquiring a human body double-path pulse signal from the image;
s302-4, completing detection of blood pressure based on a model of double blood pressure prediction of the BP neural network; the method comprises the following specific steps:
(1) and human blood pressure sequence single-scale decomposition: using db4 wavelet as the basis function of blood pressure sequence wavelet analysis, performing low-frequency and high-frequency wavelet decomposition on the original blood pressure sequence X (t), decomposing by Mallat algorithm of wavelet transformation, filtering the original signal by a low-pass filter and a high-pass filter, and then sampling to obtain a decomposition coefficient, wherein the formula is as follows:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow at resolutionFrequency and high frequency information; h represents a low-pass filter; g represents a high-pass filter; j represents the decomposition scale, J is 1, 2, …, J;
(2) reconstruction of low and high frequency components: respectively performing single-branch reconstruction on a low-frequency function and a high-frequency function of a blood pressure sequence based on wavelet basis functions selected by decomposition, and reconstructing low-frequency and high-frequency decomposition coefficients to an original scale through a wavelet algorithm to obtain a low-frequency component L (t) and a high-frequency component H (t) of blood pressure;
(3) and predicting the neural network: respectively predicting L (t) and H (t) by using a BP neural network model; the method comprises the following specific steps:
firstly, a prediction model with a single hidden layer BP neural network structure is arranged, the number of input layer neurons of the prediction model is 6, the number of input layer neurons is 1, and the number of hidden layer neurons is determined according to the following formula:
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
determining the optimal number of neurons in a hidden layer of the low-frequency component prediction model to be 12, the optimal number of neurons in a hidden layer of the high-frequency component prediction model to be 7, wherein the neurons in the hidden layer adopt sigmoid transform functions, the neurons in an output layer adopt linear transfer functions, a network training function adopts train lm, the functions train a forward network by utilizing a levenberg-Marquard algorithm, and the training speed of the algorithm is faster than that of a gradient descent method;
③, the input of the network is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs each layer of component
And
(4) and predicted sequence synthesis: will be provided with
And
the sequence prediction results are superposed to obtain the predicted value of X (t)
Namely, a prediction result corresponding to the original blood pressure sequence is obtained, and the formula is as follows:
s4, shooting and observing the head and the bleeding position of the patient by using a high-definition camera, acquiring pupil and cornea information and bleeding conditions, sending the information and the bleeding conditions to a medical server, and calculating information change and giving an evaluation result by the medical server; on-spot doctor utilizes high definition digtal camera can closely shoot the doctor and suffer from the fact like the wound face, and the picture can be sent the doctor that waits for first aid center, makes things convenient for them to deepen the understanding to the condition of patient, simultaneously according to the patient condition of bleeding, calculates a grade for the patient, can be included overall danger grade calculation later.
The frequency of the pupil and cornea changes of the patient in the video can be simply calculated and analyzed aiming at the detection of the pupil and cornea changes.
S5, acquiring the body temperature of the patient in a long distance by using a thermal infrared imager, and directly sending the body temperature to a medical server of an emergency center through a 5G network;
s6, the on-site emergency doctor inputs the information of the pulse, the respiratory rate, the pain and the wakefulness of the patient by voice through the voice recognition and recording module, converts the voice into character information by the voice recognition and recording module, and then sends the character information to the medical server of the emergency center through the 5G network.
Example 2:
as shown in fig. 2, the remote vital sign evaluation system based on 5G of the present invention includes a head-mounted device and a medical server, wherein the head-mounted device includes a high definition camera, a voice recognition and entry module, and a thermal infrared imager; the high-definition camera, the voice recognition and input module and the thermal infrared imager are respectively connected with the medical server through a 5G network;
the high-definition camera is used for shooting videos and pictures, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
the voice recognition and recording module is used for converting voice information of pulse, respiratory rate, pain and waking degree of a patient of an on-site emergency doctor into text information and sending the text information to a medical server of an emergency center through a 5G network;
the thermal infrared imager is used for obtaining the body temperature of a patient in a long distance and directly sending the body temperature to a medical server of the emergency center through a 5G network;
the medical server is used for counting all vital signs of the patient, acquiring a heart rate signal by combining an AI algorithm, completing blood pressure measurement, evaluating the degree of danger and preparing rescue in advance; the acquisition of the heart rate signal is specifically as follows:
the method comprises the following steps that (I) a medical server extracts face color change information from videos and pictures by combining an AI algorithm;
(II) amplifying the change of the human face color signal through an Eulerian Video magnetic algorithm to obtain an RGB three-channel signal;
(III) carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
and (IV) converting the chrominance signal into a frequency domain by Fourier transform (Fourier transform) to obtain the heart rate.
The blood pressure measurement is accomplished as follows:
the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
(ii) carrying out segmentation and extraction on skin colors by an Adaptive Boosting algorithm;
(iii) processing the images by an Eulerian Video magnetic algorithm and obtaining a human body two-way pulse signal therefrom;
(iv) completing the detection of the blood pressure based on a dual blood pressure prediction model of the BP neural network; the method comprises the following specific steps:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow and high frequency information at resolution; h represents a low-pass filter; g represents a high-pass filter; j represents the decomposition scale, J is 1, 2, …, J;
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
through a trial and error method, through multiple comparisons, the optimal number of neurons in a hidden layer of a low-frequency component prediction model is determined to be 12, the optimal number of neurons in a hidden layer of a high-frequency component prediction model is determined to be 7, the neurons in the hidden layer adopt a sigmoid transformation function, the neurons in an output layer adopt a linear transfer function, a network training function adopts a train lm, the function trains a forward network by utilizing a levenberg-Marquard algorithm, and the training speed of the algorithm is faster than that of a gradient descent method;
the network input is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs components of each layer
And
finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (8)
1. A remote vital sign assessment method based on 5G is characterized in that a 5G network is utilized to send videos, pictures and text information acquired on site to a medical server of a hospital emergency center, the medical server analyzes the videos, the pictures and the text information by combining an AI algorithm technology, and then carries out grade assessment on vital signs of patients, so that medical staff of the emergency center can make rescue preparations in advance according to specific conditions of the patients; the method comprises the following specific steps:
s1, wearing head-mounted equipment by the on-site emergency doctor, wherein the head-mounted equipment comprises a high-definition camera, a voice recognition and recording module and a thermal infrared imager;
s2, shooting videos and pictures by the high-definition camera, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
s3, the medical server acquires heart rate signals from the videos and the pictures by combining an AI algorithm and completes blood pressure measurement;
s4, shooting and observing the head and the bleeding position of the patient by using a high-definition camera, acquiring pupil and cornea information and bleeding conditions, sending the information and the bleeding conditions to a medical server, and calculating information change and giving an evaluation result by the medical server;
s5, acquiring the body temperature of the patient in a long distance by using a thermal infrared imager, and directly sending the body temperature to a medical server of an emergency center through a 5G network;
s6, the on-site emergency doctor inputs the information of the pulse, the respiratory rate, the pain and the wakefulness of the patient by voice through the voice recognition and recording module, converts the voice into character information by the voice recognition and recording module, and then sends the character information to the medical server of the emergency center through the 5G network.
2. The 5G-based remote vital sign assessment method according to claim 1, wherein the step S3 of acquiring the heart rate signal is as follows:
s301-1, extracting face color change information from videos and pictures by the medical server in combination with an AI algorithm;
s301-2, amplifying the change of the face color signal through an Euler amplification algorithm to obtain RGB three-channel signals;
s301-3, carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
s301-4, converting the chrominance signal into a frequency domain by adopting Fourier transform to obtain the heart rate.
3. The 5G-based remote vital sign assessment method according to claim 1 or 2, wherein the blood pressure measurement in step S3 is accomplished as follows:
s302-1, the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
s302-2, carrying out skin color segmentation and extraction through an Adaptive Boosting algorithm;
s302-3, processing the image through an Euler amplification algorithm and obtaining a human body double-path pulse signal from the image;
s302-4, completing detection of blood pressure based on a model of double blood pressure prediction of the BP neural network.
4. The 5G-based remote vital sign assessment method according to claim 3, wherein the BP neural network-based dual blood pressure prediction model in step S302-4 performs the following steps:
(1) and human blood pressure sequence single-scale decomposition: using db4 wavelet as the basis function of blood pressure sequence wavelet analysis, performing low-frequency and high-frequency wavelet decomposition on the original blood pressure sequence X (t), decomposing by Mallat algorithm of wavelet transformation, filtering the original signal by a low-pass filter and a high-pass filter, and then sampling to obtain a decomposition coefficient, wherein the formula is as follows:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow and high frequency information at resolution; h represents a low-pass filter; g represents a high-pass filter; j represents the decomposition scale, J is 1, 2, …, J;
(2) reconstruction of low and high frequency components: respectively performing single-branch reconstruction on a low-frequency function and a high-frequency function of a blood pressure sequence based on wavelet basis functions selected by decomposition, and reconstructing low-frequency and high-frequency decomposition coefficients to an original scale through a wavelet algorithm to obtain a low-frequency component L (t) and a high-frequency component H (t) of blood pressure;
(3) and predicting the neural network: respectively predicting L (t) and H (t) by using a BP neural network model, and processing layer by layer from an input layer through a hidden layer until components of each layer are output by an output layer
And
(4) and predicted sequence synthesis: will be provided with
And
the sequence prediction results are superposed to obtain the predicted value of X (t)
Namely, a prediction result corresponding to the original blood pressure sequence is obtained, and the formula is as follows:
5. the 5G-based remote vital sign assessment method according to claim 4, wherein the neural network prediction in step (3) is specifically as follows:
firstly, a prediction model with a single hidden layer BP neural network structure is arranged, the number of input layer neurons of the prediction model is 6, the number of input layer neurons is 1, and the number of hidden layer neurons is determined according to the following formula:
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
determining the optimal number of neurons in a hidden layer of the low-frequency component prediction model to be 12, the optimal number of neurons in a hidden layer of the high-frequency component prediction model to be 7, wherein the neurons in the hidden layer adopt sigmoid transform functions, the neurons in an output layer adopt linear transfer functions, a network training function adopts train lm, and the functions train a forward network by utilizing a levenberg-Marquard algorithm through a trial and error method and multiple comparisons;
③, the input of the network is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs each layer of component
And
6. a remote vital sign evaluation system based on 5G is characterized by comprising head-mounted equipment and a medical server, wherein the head-mounted equipment comprises a high-definition camera, a voice recognition input module and a thermal infrared imager; the high-definition camera, the voice recognition and input module and the thermal infrared imager are respectively connected with the medical server through a 5G network;
the high-definition camera is used for shooting videos and pictures, and transmitting the videos and the pictures to a medical server in emergency in real time through a 5G network;
the voice recognition and recording module is used for converting voice information of pulse, respiratory rate, pain and waking degree of a patient of an on-site emergency doctor into text information and sending the text information to a medical server of an emergency center through a 5G network;
the thermal infrared imager is used for obtaining the body temperature of a patient in a long distance and directly sending the body temperature to a medical server of the emergency center through a 5G network;
the medical server is used for counting all vital signs of the patient, acquiring a heart rate signal by combining an AI algorithm, completing blood pressure measurement, evaluating the degree of danger and preparing rescue in advance.
7. The 5G-based remote vital signs assessment system according to claim 6, wherein said acquiring a heart rate signal is specifically as follows:
the method comprises the following steps that (I) a medical server extracts face color change information from videos and pictures by combining an AI algorithm;
(II) amplifying the change of the human face color signal through an Euler amplification algorithm to obtain an RGB three-channel signal;
(III) carrying out normalization processing on the RGB three-channel signals to obtain chrominance signals;
and (IV) converting the chrominance signal into a frequency domain by adopting Fourier transform to obtain the heart rate.
8. 5G-based remote assessment vital signs system according to claim 6 or 7, wherein said performing of blood pressure measurements is specifically as follows:
the medical server extracts skin color images of different human body parts from videos and pictures by combining an AI algorithm;
(ii) carrying out segmentation and extraction on skin colors by an Adaptive Boosting algorithm;
(iii) processing the image through an Euler amplification algorithm and acquiring a human body two-way pulse signal from the image;
(iv) completing the detection of the blood pressure based on a dual blood pressure prediction model of the BP neural network; the method comprises the following specific steps:
Cj＝HCj-1；
Dj＝GCj-1；
wherein, CjAnd DjRespectively, represent the original signals at 2-jLow and high frequency information at resolution; h represents a low-pass filter; g represents a high-pass filter; j represents the decomposition scale, J is 1, 2, …, J;
wherein m represents the number of input layer nodes; n represents the number of output nodes; a belongs to [0,10 ];
determining the optimal number of neurons in a hidden layer of a low-frequency component prediction model to be 12, the optimal number of neurons in a hidden layer of a high-frequency component prediction model to be 7, wherein the neurons in the hidden layer adopt a sigmoid transformation function, the neurons in an output layer adopt a linear transfer function, a network training function adopts a train lm, and the function trains a forward network by utilizing a levenberg-Marquard algorithm through a trial and error method and multiple comparisons;
the network input is L (t) and H (t) after wavelet decomposition and reconstruction, and the input layer is processed layer by layer through the hidden layer until the output layer outputs components of each layer
And
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010390520.3A CN111554398A (en) | 2020-05-11 | 2020-05-11 | Remote vital sign evaluation method and system based on 5G |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010390520.3A CN111554398A (en) | 2020-05-11 | 2020-05-11 | Remote vital sign evaluation method and system based on 5G |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111554398A true CN111554398A (en) | 2020-08-18 |
Family
ID=72008073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010390520.3A Pending CN111554398A (en) | 2020-05-11 | 2020-05-11 | Remote vital sign evaluation method and system based on 5G |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111554398A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114171195A (en) * | 2021-10-28 | 2022-03-11 | 北京大学人民医院 | Emergency patient pre-hospital index calculation method, electronic device and storage medium |
| CN114390204A (en) * | 2022-01-27 | 2022-04-22 | 复旦大学 | A smart medical treatment method and system |
| CN115363582A (en) * | 2022-08-20 | 2022-11-22 | 中国科学院心理研究所 | Face clue-based life perception measurement method and system |
| CN119811646A (en) * | 2025-03-11 | 2025-04-11 | 宁波市威尔信息科技有限公司 | An emergency first aid data connection and transmission system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080310600A1 (en) * | 2007-06-13 | 2008-12-18 | Clawson Jeffrey J | Diagnostic and intervention tools for emergency medical dispatch |
| CN106236049A (en) * | 2016-10-12 | 2016-12-21 | 南京工程学院 | Blood pressure measuring method based on video image |
| CN109009033A (en) * | 2018-07-10 | 2018-12-18 | 华南理工大学 | A kind of blood pressure prediction technique based on wavelet analysis and echo state network |
| CN109949934A (en) * | 2018-11-15 | 2019-06-28 | 陕西医链区块链集团有限公司 | Calculation method for evaluating body health state by utilizing AI algorithm |
| CN110236511A (en) * | 2019-05-30 | 2019-09-17 | 云南东巴文健康管理有限公司 | A kind of noninvasive method for measuring heart rate based on video |
| CN110706826A (en) * | 2019-10-29 | 2020-01-17 | 江苏大学 | Non-contact real-time multi-person heart rate and blood pressure measuring method based on video images |
| CN110719333A (en) * | 2019-10-18 | 2020-01-21 | 江苏盖睿健康科技有限公司 | Intelligent 5G emergency ambulance cloud emergency system and method |
-
2020
- 2020-05-11 CN CN202010390520.3A patent/CN111554398A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080310600A1 (en) * | 2007-06-13 | 2008-12-18 | Clawson Jeffrey J | Diagnostic and intervention tools for emergency medical dispatch |
| CN106236049A (en) * | 2016-10-12 | 2016-12-21 | 南京工程学院 | Blood pressure measuring method based on video image |
| CN109009033A (en) * | 2018-07-10 | 2018-12-18 | 华南理工大学 | A kind of blood pressure prediction technique based on wavelet analysis and echo state network |
| CN109949934A (en) * | 2018-11-15 | 2019-06-28 | 陕西医链区块链集团有限公司 | Calculation method for evaluating body health state by utilizing AI algorithm |
| CN110236511A (en) * | 2019-05-30 | 2019-09-17 | 云南东巴文健康管理有限公司 | A kind of noninvasive method for measuring heart rate based on video |
| CN110719333A (en) * | 2019-10-18 | 2020-01-21 | 江苏盖睿健康科技有限公司 | Intelligent 5G emergency ambulance cloud emergency system and method |
| CN110706826A (en) * | 2019-10-29 | 2020-01-17 | 江苏大学 | Non-contact real-time multi-person heart rate and blood pressure measuring method based on video images |
Non-Patent Citations (4)
| Title |
|---|
| 方杰、张征: "基于小波分析与BP神经网络的人体血压预测", 《计算机系统应用》 * |
| 李强等: "构建基于5G通讯技术的大型社会活动医疗急救保障系统", 《中华急诊医学杂志》 * |
| 许可: "基于肤色图像的非接触式血压测量方法研究", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 * |
| 陈玉国: "《急诊医学新进展》", 31 January 2016, 中华医学电子音像出版社 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114171195A (en) * | 2021-10-28 | 2022-03-11 | 北京大学人民医院 | Emergency patient pre-hospital index calculation method, electronic device and storage medium |
| CN114390204A (en) * | 2022-01-27 | 2022-04-22 | 复旦大学 | A smart medical treatment method and system |
| CN114390204B (en) * | 2022-01-27 | 2022-10-04 | 复旦大学 | A smart medical treatment method and system |
| CN115363582A (en) * | 2022-08-20 | 2022-11-22 | 中国科学院心理研究所 | Face clue-based life perception measurement method and system |
| CN119811646A (en) * | 2025-03-11 | 2025-04-11 | 宁波市威尔信息科技有限公司 | An emergency first aid data connection and transmission system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111554398A (en) | Remote vital sign evaluation method and system based on 5G | |
| CN108272448B (en) | A non-contact day and night monitoring method for infant physiological parameters | |
| WO2019052063A1 (en) | Medical image classification processing system and method based on artificial intelligence | |
| CN112754431A (en) | Respiration and heartbeat monitoring system based on millimeter wave radar and lightweight neural network | |
| Zhao et al. | An IoT-based wearable system using accelerometers and machine learning for fetal movement monitoring | |
| CN114038578B (en) | A 5G-based medical supervision system | |
| CN114677826A (en) | A violent early warning system for mental patients based on individual behavior and physiological characteristics | |
| CN114628020A (en) | Model construction, detection method, device and application of remote plethysmographic signal detection | |
| Ahmed et al. | Internet of health things driven deep learning-based system for non-invasive patient discomfort detection using time frame rules and pairwise keypoints distance feature | |
| CN110689935A (en) | IoT emergency service platform and early warning method | |
| CN113706824A (en) | Old man nurses system at home based on thing networking control | |
| CN114938952B (en) | Hemoglobin concentration prediction method based on a priori causal model | |
| CN116343302A (en) | A Micro-expression Classification and Recognition System Based on Machine Vision | |
| CN115721322A (en) | A system and method for monitoring abnormal sleep of the elderly based on EEG signals | |
| CN119745362A (en) | An artificial intelligence-based system for assessing the effects of anesthetic overdose on brain function | |
| CN113768523A (en) | Method and system for prewarning stool based on countermeasure generation network | |
| CN115154828B (en) | Brain function remodeling method, system and equipment based on brain-computer interface technology | |
| CN118352055A (en) | Remote nursing method and system for child internal medicine patient | |
| CN115272977A (en) | A Heart Attack Detection System Based on Convolutional Neural Networks and Transfer Learning | |
| CN117292420A (en) | Depression detection and recognition system based on facial submicron features | |
| CN106709260A (en) | Understanding doctor type doctor-patient online interactive method | |
| KR102676918B1 (en) | Apparatus for estimating biosignals from face image | |
| García-Torres et al. | AI-Based Thermal Video Analysis in Privacy-Preserving Healthcare: A Case Study on Detecting Time of Birth | |
| CN110432897A (en) | Fetal ECG extraction system and method based on residual error convolution encoding and decoding neural network | |
| CN114580913B (en) | An intelligent management platform for public transportation services |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200818 |