WO2021109601A1 - 一种麻醉深度的测量方法、存储介质及电子设备 - Google Patents

一种麻醉深度的测量方法、存储介质及电子设备 Download PDF

Info

Publication number
WO2021109601A1
WO2021109601A1 PCT/CN2020/106066 CN2020106066W WO2021109601A1 WO 2021109601 A1 WO2021109601 A1 WO 2021109601A1 CN 2020106066 W CN2020106066 W CN 2020106066W WO 2021109601 A1 WO2021109601 A1 WO 2021109601A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
sub
eeg
anesthesia
feature
Prior art date
Application number
PCT/CN2020/106066
Other languages
English (en)
French (fr)
Inventor
韩如泉
温鹏
熊飞
任冠清
李兴
周赤宜
戴仁泉
王筱毅
李明
Original Assignee
深圳市德力凯医疗设备股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 深圳市德力凯医疗设备股份有限公司 filed Critical 深圳市德力凯医疗设备股份有限公司
Publication of WO2021109601A1 publication Critical patent/WO2021109601A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/48—Other medical applications
    • A61B5/4821—Determining level or depth of anaesthesia
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/369—Electroencephalography [EEG]
    • 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
    • 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
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • 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
    • A61B5/7246—Details of waveform analysis using correlation, e.g. template matching or determination of similarity
    • 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
    • A61B5/7271—Specific aspects of physiological measurement analysis
    • A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor

Definitions

  • the invention relates to the field of ultrasound technology, in particular to a method for measuring the depth of anesthesia, a storage medium and electronic equipment.
  • general anesthesia commonly known as general anesthesia
  • general anesthesia usually known as general anesthesia
  • general anesthesia has a very high risk.
  • too deep anesthesia can cause complications of anesthesia and even endanger the patient's life; too shallow anesthesia is likely to occur "intraoperative awareness ", produce pain, fear and mental sequelae; therefore, how to accurately estimate the depth of anesthesia, so that doctors can determine the amount of anesthesia according to the depth of anesthesia, to improve the safety of anesthesia.
  • the current methods of monitoring the depth of anesthesia based on EEG signals mainly include bispectral index, auditory evoked potential index, brain function state index, entropy index, complexity and wavelet analysis.
  • the BIS monitor launched by the American Aspect Company actually uses bispectral analysis to facilitate the use of an index of 0-100 to reflect the depth of anesthesia awareness.
  • the above-mentioned methods generally use the patient's real EEG signal, and the real EEG signal will have interference signals during the acquisition process, which will cause the real EEG signal to be abnormal.
  • doctors need to judge the patient's anesthesia status based on indirect indicators such as blood pressure, heart rate, respiratory rate, and degree of muscle relaxation. This depends on the doctor's professional level, and different doctors have different judgment results.
  • the present invention aims to provide a method for measuring the depth of anesthesia, a storage medium and an electronic device.
  • a method for measuring the depth of anesthesia which includes:
  • a simulated EEG signal corresponding to the EEG signal is generated according to the signal characteristic and the predicted signal characteristic, and an anesthesia depth value is determined according to the simulated EEG signal.
  • the process of generating the predicted EEG signal specifically includes:
  • the matrix brain network is driven to generate EEG signals according to preset connectivity parameters, where the preset connectivity parameters are used to control the state of each neuron group module in the matrix brain network.
  • the neuron group module includes an excitation neural cell network, a pyramidal cell network, and an inhibitory neural cell network; the excitation signal for exciting the neural cell network, an inhibitory signal for inhibiting the neural cell network, And the external excitation signal of the pyramidal cell network forms brain waves, and the brain waves are output to the external neuron group module, and the brain waves are respectively fed back to the stimulated nerve cell network and inhibited nerve cell of the neuron group module.
  • the obtaining an electroencephalogram signal and extracting the signal characteristics of the electroencephalogram signal specifically include:
  • the generating the analog EEG signal corresponding to the EEG signal according to the signal characteristic and the predicted signal characteristic specifically includes:
  • the analog signal characteristic corresponding to the brain electrical signal is determined according to the matching result, and the analog brain electrical signal is generated according to the analog signal characteristic.
  • the determining the analog signal characteristic corresponding to the brain electrical signal according to the matching result specifically includes:
  • sub-signal feature For sub-signal features that have the same matching result, use the sub-signal feature as the sub-analog signal feature corresponding to the sub-signal feature;
  • the analog signal characteristics corresponding to the brain electrical signal are generated according to the determined characteristics of all the sub-analog signals.
  • the matching result is a sub-signal feature that is different
  • the sub-signal feature is processed, and the processed sub-signal feature is used as the sub-analog signal feature corresponding to the sub-signal feature Specifically:
  • linear interpolation processing is performed on the sub-signal feature to obtain the sub-analog signal feature corresponding to the sub-signal feature.
  • the matching result is a sub-signal feature that is different
  • a moving window translation algorithm is used to determine the sub-analog signal feature corresponding to the sub-signal feature
  • the sub-prediction signal feature corresponding to the sub-signal feature is used as the sub-analog signal feature corresponding to the sub-signal feature;
  • the sub-signal feature and the sub-prediction signal feature corresponding to the sub-signal feature are weighted to obtain the sub-analog signal feature corresponding to the sub-signal feature.
  • a computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to realize the depth of anesthesia as described above The steps in the measurement method.
  • An electronic device comprising: a processor, a memory, and a communication bus; the memory stores a computer readable program that can be executed by the processor;
  • the communication bus realizes connection and communication between the processor and the memory
  • the present invention provides a method for measuring the depth of anesthesia, a storage medium, and electronic equipment.
  • the method includes acquiring an EEG signal, and extracting the signal characteristics of the EEG signal;
  • the predicted EEG signal corresponding to the EEG signal, and the predicted signal feature of the predicted EEG signal is extracted;
  • the analog EEG signal corresponding to the EEG signal is generated according to the signal feature and the predicted signal feature, and the analog EEG signal corresponding to the EEG signal is generated according to the Simulate EEG signals to determine the depth of anesthesia.
  • the present invention combines the predicted EEG signal generated by the network brain model with the actual measured EEG signal to obtain a simulated EEG signal, and then calculates the depth of anesthesia based on the simulated EEG signal, so that the predicted EEG signal can be compared to the actual Measuring EEG signals for verification can improve the accuracy of the measured anesthesia depth, improve the reliability of the anesthesia depth, and improve the dynamic real-time tracking ability and anti-interference ability of the anesthesia depth measurement.
  • Fig. 1 is a flow chart of the method for measuring the depth of anesthesia provided by the present invention.
  • Fig. 2 is a schematic diagram of the neuron group module in the method for measuring the depth of anesthesia provided by the present invention.
  • Fig. 3 is a schematic view of a human brain model in the method for measuring the depth of anesthesia provided by the present invention.
  • Fig. 4 is a schematic diagram of the human brain model from another angle in the method for measuring the depth of anesthesia provided by the present invention.
  • Fig. 5 is a schematic structural diagram of an electronic device provided by the present invention.
  • the present invention provides a method, storage medium and electronic equipment for measuring the depth of anesthesia.
  • the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not used to limit the present invention.
  • This implementation provides a method for measuring the depth of anesthesia, as shown in Figures 1 and 2, the method includes:
  • the EEG signal can be collected by an EEG acquisition device, can also be obtained by an electrode sheet, or can be collected by a sensor, and the EEG signal collected by the sensor can be an analog EEG signal, and Collecting the analog EEG signal can convert the analog EEG signal into a digital EEG signal, and use the digital EEG signal as the EEG signal, which can facilitate the storage of the EEG signal.
  • the EEG acquisition device can be various devices for acquiring EEG signals known to those skilled in the art.
  • the EEG signal may be the EEG signal of each segment of the patient during the whole process of anesthesia. For example, the signal length of each EEG signal is 10 seconds, and the EEG signal is controlled by the sliding window technology through the step length. Update of EEG signal data in seconds.
  • the signal feature may include at least one of a time domain feature, a frequency domain feature, and a non-frequency domain feature.
  • the signal feature includes a time domain feature or a frequency domain feature, that is, after the EEG signal is obtained, the time domain feature or the frequency domain feature of the EEG signal is extracted.
  • the acquiring the brain electrical signal and extracting the signal feature of the brain electrical signal specifically includes:
  • the denoising processing is to identify and process physiological interference signals and non-physiological interference signals. Denoising the EEG signal can improve the instruction of the EEG signal, thereby removing the interference signal to achieve a more accurate value of the depth of anesthesia.
  • the physiological interference signal may include an eye movement interference signal and an electromyographic interference signal
  • the non-physiological interference signal may include abnormal signal amplitude, abnormal signal slope, electrosurgical interference, and the like. Therefore, in order to remove physiological interference signals and non-physiological interference signals, the EEG signal may be digitally filtered, where the digital filtering may include low-pass filtering, high-pass filtering, and feature wave deletion.
  • the preset matrix brain network is established in advance, and the predicted brain electrical signal can be generated through the matrix brain network.
  • the predicted electroencephalogram signal may be used to determine anesthesia data corresponding to the electroencephalogram signal according to a preset anesthesia scheme, and dynamically obtain the predicted electroencephalogram signal corresponding to the electroencephalogram signal by changing model parameters in real time according to the anesthesia data.
  • the predicted EEG signal can be obtained offline in advance and stored in the standby anesthesia process according to the anesthesia data in the preset anesthesia plan by changing the model parameters.
  • the process of obtaining the predicted EEG signal is: after the anesthesia plan is determined, the model parameters corresponding to each moment are determined according to the anesthesia plan, and then the model parameters corresponding to each moment are used in chronological order.
  • Input to the matrix brain network the predicted EEG signal during the anesthesia process can be obtained, and after the predicted EEG signal is obtained, the predicted EEG signal corresponding to the anesthesia process can be stored.
  • the predicted EEG signal corresponding to the EEG signal may be selected from the predicted EEG signal generated by the preset matrix brain network according to the time period corresponding to the EEG signal, and the predicted EEG signal corresponding to the time period is selected.
  • the acquisition time of the predicted EEG signal can be reduced, and the real-time acquisition of the depth of anesthesia can be improved.
  • the anesthesia data may include anesthesia time, the speed of entering an anesthesia state, and the like.
  • the generation process of the predicted EEG information specifically includes:
  • M20 Drive the matrix brain network to generate EEG signals according to preset connectivity parameters, where the preset connectivity parameters are used to control the state of each neuron group module in the matrix brain network.
  • the neuron group module is preselected and established, each neuron group module represents a cerebral cortex area, the matrix brain network includes a plurality of neuron group modules, and each neuron group module corresponds to the cerebral cortex area
  • the cerebral cortex constitutes the cerebral cortex, that is, the cerebral cortex can be divided into a plurality of cerebral cortex regions, each cerebral cortex corresponds to a neuron group module, and the neuron group module is included in the plurality of neuron group modules forming the matrix brain network .
  • each of the multiple cerebral cortex areas is responsible for different functions, for example, the occipital lobe area is responsible for visual functions; the parietal lobe area is responsible for touch and space; the temporal lobe area is responsible for hearing and comprehensive perception; the frontal lobe motor cortex Responsible for coordinating the movement of the limbs; the prefrontal cortex is responsible for understanding, memory and judgment.
  • the neuron group module includes 76, that is, the left and right cerebral cortex are divided into 76 regions.
  • the matrix brain network has 76 elements. Currently, there are 76 elements. It is the best result we can achieve. Of course, more elements will make the model more accurate.
  • the EEG signal is a reflection of the activity of a large number of neurons, especially in the process of anesthesia, the EEG change of the EEG signal has a strong non-linear characteristic.
  • the EEG signal during anesthesia is formed by the oscillation coupling of a large number of neurons between the thalamus and the cerebral cortex, and with the change of the depth of anesthesia, the coupling strength will be significantly different, so that the EEG change of the EEG signal has Strong non-linear characteristics.
  • the function of the thalamus is very much like a perceptual "switch" or a monitor of overall brain activity, and the regular activity of the neural network of the thalamus and cerebral cortex. It is the key to anesthesia and perception.
  • the inhibitory process is transmitted from the base of the forebrain and the hypothalamus to the awakening core through GaBa+ ions, and at the same time the awakening core sends out to the deaf core.
  • a partial smooth process model is constructed based on the preset neuron group module, that is, the matrix brain network.
  • the matrix brain network is based on the preset neuron group module as the element, and according to each neuron group
  • the coupling strength between the modules establishes a matrix brain network, and the matrix brain network can combine the microscopic brain cell activity with the mesoscopic cortex activity and the macroscopic EEG signal number to make it pass all
  • the matrix brain network can simulate EEG signals.
  • the neuron population module is a JANSEN-RIT (JR) module
  • the JANSEN-RIT module includes an excitation neural cell network, a pyramidal cell network, and an inhibitor neural cell network;
  • the excitation signal for stimulating the neural cell network, the inhibitory signal for inhibiting the neural cell network, and the external excitation signal of the pyramidal cell network form brain waves, and output the brain electrical signals to an external neuron group module.
  • the excitation neural cell network, the pyramidal cell network, and the inhibitory neural cell network can all be expressed as neural circuits.
  • the excitation neural cell network includes a first pulse branch and a first feedback branch.
  • the voltage-pulse encoder is used to add The voltage signal of is converted into a pulse signal and then transmitted to the pyramidal cell network;
  • the suppressed neural cell network includes a second pulse branch and a second feedback branch, the voltage signal of the second pulse branch and the voltage of the second feedback branch
  • the added voltage signal is converted into a pulse signal by a voltage-pulse encoder and then transmitted to the pyramidal cell network;
  • the pyramidal cell network includes a local excitation branch, a local suppression branch and an external excitation branch.
  • the local excitation branch receives the pulse signal transmitted by the excitation neural network
  • the local suppression branch receives the pulse signal transmitted by the neural network
  • the external excitation branch receives the input pulse
  • the local excitation branch, the local suppression branch and the external The excitation branches merge to form a brain wave signal, which is output to other neuron cluster modules through a voltage-pulse encoder, and is fed back to the excitation neural cell network and the inhibition neural cell network of the JANSEN-RIT module respectively.
  • C 1T , C 13 , C 2T , C 23 , C 31 , C 32 and C 3T are all neural network groups Average number of synapses between; m 1T (t), m 1 (t), m 2T (t), m 2 (t), m 3 (t) and m 3T (t) are all pulse signals, v 1T (t ), v 13 (t), v 1 (t), v 2T (t), v 23 (t), v 2 (t), v 3T (t), v 31 (t), v 32 (t) and v 3 (t) is a voltage signal, where v 3 (t) is an EEG signal that can be detected.
  • the suggesting a matrix brain network with spatial characteristics using a pre-established neuron group module as an element specifically includes:
  • the brain data is obtained based on anatomy, and a structural brain network can be suggested based on the brain data.
  • the structural brain network includes 76 regions. After the structured brain network is established, for each area of the structured brain network, the coupling strength coefficient between the neural network nodes in the area is obtained, and the matrix of the area is determined according to all the coupling strength coefficients obtained, so that the structure type
  • the dynamic evolution of the brain network will produce a series of brain network matrices, and the dynamic brain network matrices at different time points form a matrix-type brain network.
  • the brain data includes the correlation between the cortex area and each cortex area, where the correlation strength of each cortex area can be represented by 0, 1, and -1, or it can be normalized from -1 to Any value of +1, and various brain activities can be represented by a set of dynamic time-domain equations describing large-scale brain neural networks. These equations can be selected according to the purpose of the research. For example, we can choose the following equations to compare conventional The EEG signal is simulated:
  • the parameters in the above dynamic time domain equation can be determined by known physiological parameter values, including but not limited to the conductivity of various brain tissues in the model.
  • the preset connection parameters are used to indicate the connection status of each JR module, where the connection parameters include unconnected, weak connected, normal connected, and strong connected, for example, 0, 1, 2, and 3 are used to indicate respectively .
  • the JR module is composed of three neural network loops as shown in Figure 2, and can generate brain waves according to preset parameters, where the preset parameters may be brain tissue conductivity, etc.
  • the conductivity of brain tissue can include scalp, skull, brain tissue fluid, white tissue, gray tissue, and blood vessels.
  • each neural network loop is represented by its key state variables, for example, the average membrane potential, the average activation rate and the mutual conversion between them (pulse-wave potential and wave potential-pulse), etc., so that you can
  • the connection status of each JR module is determined according to the oscillation frequency and the average PSP baseline of the cone, that is, according to the pre-selected connection parameters for controlling the state transition, the interaction between the various cerebral cortex will produce such as how stable , Synchronization and until the state of coordination, so as to generate the corresponding brain wave signal.
  • the driving the matrix brain network to generate EEG signals according to preset Unicom parameters specifically includes:
  • M22 Drive the matrix brain network according to the connection parameters, determine the potential information corresponding to the neural activity through the matrix brain network, and obtain an brain electrical signal according to the potential information.
  • the neural activity of the thalamus and cerebral cortex is known, which can be obtained through CT or magnetic resonance. After the neural activity of the thalamus and cerebral cortex is obtained, the neural activity of the thalamus and cerebral cortex is determined.
  • the connection parameters of the neuron group module, and the matrix brain network is driven by the connection parameters, so that the matrix brain network determines the potential information corresponding to the neural activity, and obtains it according to the potential information EEG signal.
  • the potential information corresponding to the neural activity is determined by the matrix brain network, and the brain electrical signal is obtained according to the potential information;
  • each electrode point unit of the brain scalp is determined according to the human brain model, and the brain electrical signal is generated according to all the determined electric potentials.
  • the brain image may be obtained through nuclear magnetic resonance, and the brain image includes a plurality of continuous nuclear magnetic resonance images, so that images of various parts of the brain can be obtained.
  • the brain image is identified to determine the brain components, where the brain components include dozens of human heads such as cranium, brain fluid, brain gray matter, and brain white matter. organization.
  • the brain components can be combined with known electrical parameters, such as scalp, skull, brain tissue fluid, white tissue, gray tissue, and blood vessels, to create a digital human head model, namely The human brain model is obtained, as shown in Figures 3 and 4.
  • the digital human brain model is abstracted into millions of small units.
  • Each unit represents a different part and organization of the human brain, and is assigned corresponding electrical quality parameters.
  • Cells can not only represent different human head tissues, but also different electrical parameters of the same tissue. After that, the positions of each small cell in the digital human brain model are determined, and then the cell positions corresponding to the electrode points on the scalp are found. To get EEG signals.
  • generating the analog EEG signal corresponding to the EEG signal according to the signal feature and the predicted signal feature refers to matching the signal feature with the predicted signal feature to divide the EEG signal , Are divided into matched signal set, linear interpolation signal set and completely mismatched signal.
  • the signal feature of the EEG signal in the matching signal set matches its corresponding prediction signal feature;
  • the signal feature of the EEG signal in the linear interpolation signal set can be matched to its corresponding prediction signal feature by means of linear interpolation
  • the signal feature of the EEG signal in the completely mismatched signal does not match its corresponding predicted signal feature, and cannot be matched to its corresponding predicted signal feature by means of linear interpolation.
  • the EEG signals in the matching signal set, linear interpolation signal set, and completely mismatched signals can be configured with different anesthesia estimation flags.
  • the EEG signals in the matching signal set can be configured with automatic flags and linear interpolation signals.
  • the concentrated EEG signal is configured with a semi-automatic identification
  • the EEG signal in the completely mismatched signal is configured with a semi-automatic identification after being corrected according to the predicted EEG signal. In this way, after matching the signal characteristics with the predicted signal characteristics, the depth of anesthesia can be determined according to the anesthesia estimation flag.
  • the generating the analog brain electrical signal corresponding to the brain electrical signal according to the signal characteristic and the predicted signal characteristic specifically includes:
  • the analog signal characteristic corresponding to the brain electrical signal is determined according to the matching result, and the analog brain electrical signal is generated according to the analog signal characteristic.
  • the matching result includes matching, matching after linear interpolation, and non-matching, which cannot be matched even through linear interpolation.
  • the EEG signal processing methods for different matching results are different.
  • the characteristic is that for EEG signals that are not matched and cannot be matched through linear interpolation, the EEG signals can be corrected by predicting the EEG signals, and according to the corrected EEG signals.
  • the brain blood signal generates a simulated brain blood signal, which can improve the accuracy of the simulated brain electrical signal, thereby improving the accuracy of the depth of anesthesia measurement.
  • the determining the analog signal characteristic corresponding to the brain electrical signal according to the matching result specifically includes:
  • sub-signal feature For sub-signal features that have the same matching result, use the sub-signal feature as the sub-analog signal feature corresponding to the sub-signal feature;
  • the analog signal characteristics corresponding to the brain electrical signal are generated according to the determined characteristics of all the sub-analog signals.
  • the matching result is the same, indicating that the signal feature is the same as the predicted signal feature. From this, it can be determined that the cerebral blood signal corresponding to the signal feature is an accurate EEG signal, so that the sub-signal feature can be used as the sub-signal feature Corresponding sub-analog signal characteristics. For sub-signal features whose matching results are not the same, the sub-signal features need to be processed to match the sub-signal features with their corresponding predicted signal features, but when the sub-signal features cannot be matched with their corresponding predicted signal features The matching is to correct the feature of the sub-signal according to the feature of the predicted signal.
  • the sub-signal features are processed, and the processed sub-signal feature is used as the sub-signal feature corresponding to the sub-signal feature.
  • the characteristics of the analog signal specifically include:
  • a moving window translation algorithm is used to determine the sub-analog signal feature corresponding to the sub-signal feature
  • the sub-prediction signal feature corresponding to the sub-signal feature is used as the sub-analog signal feature corresponding to the sub-signal feature;
  • the sub-signal feature and the sub-prediction signal feature corresponding to the sub-signal feature are weighted to obtain the sub-analog signal feature corresponding to the sub-signal feature.
  • the non-matching part will be integrated by linear interpolation to generate a semi-automatic identification.
  • the following three cases will be processed: 1) If the actual measurement If the EEG noise ratio is within the preset range, the moving window smoothing algorithm is applied to convert the signal characteristics of the measured EEG signal into sub-analog signal features. 2) If the EEG noise ratio of the measured EEG is outside the preset range, the EEG will be predicted The predicted signal characteristics of the signal are converted into sub-analog signal characteristics.
  • the preset range corresponding to the predicted signal-to-noise ratio may be determined according to the selected anesthesia feature, for example, between 10% and 50%.
  • matching is a process. Since anesthesia is a controlled and gradual continuous process, there can be no discontinuities in the middle, so when there is a mismatch, linear interpolation can be performed to match the signal characteristics to the predicted signal characteristics.
  • the depth of anesthesia can be determined according to the simulated EEG signal, where the depth of anesthesia can be normalized by a similar percentage to give a depth index from 0 to 100, for example, the The index can be shown in the chart below to reflect the depth of anesthesia of the patient.
  • Depth of Anesthesia Index Anesthesia Physiological/physical characteristics 100 wide awake All reactions are normal 80 Mild anesthesia Increased blood pressure and heart rate, responding to strong stimuli 60 General anesthesia Stable blood pressure, unconscious and unresponsive 40 Deep anesthesia High blood pressure, irregular heartbeat 20 Burst suppression Intermittent EEG activity
  • this embodiment provides a computer-readable storage medium that stores one or more programs, and the one or more programs can be processed by one or more The device executes to realize the steps in the method for measuring the depth of anesthesia as described in the above-mentioned embodiments.
  • the present invention also provides an electronic device, as shown in FIG. 5, which includes at least one processor (processor) 20; a display screen 21; and a memory (memory) 22, which may also include communication Interface (Communications Interface) 23 and bus 24.
  • processor processor
  • the display screen 21 is set to display a user guide interface preset in the initial setting mode.
  • the communication interface 23 can transmit information.
  • the processor 20 can call the logic instructions in the memory 22 to execute the method in the foregoing embodiment.
  • logic instructions in the memory 22 can be implemented in the form of software functional units and when sold or used as independent products, they can be stored in a computer readable storage medium.
  • the memory 22 can be configured to store software programs and computer-executable programs, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure.
  • the processor 20 executes functional applications and data processing by running software programs, instructions, or modules stored in the memory 22, that is, implements the methods in the foregoing embodiments.
  • the memory 22 may include a program storage area and a data storage area.
  • the program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created according to the use of the electronic device, and the like.
  • the memory 22 may include a high-speed random access memory, and may also include a non-volatile memory.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Pathology (AREA)
  • Physics & Mathematics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Veterinary Medicine (AREA)
  • Surgery (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Psychiatry (AREA)
  • Signal Processing (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physiology (AREA)
  • Anesthesiology (AREA)
  • Psychology (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

一种麻醉深度的测量方法、存储介质及电子设备，包括获取脑电信号，并提取脑电信号的信号特征（S10）；获取脑电信号对应的预测脑电信号，并提取预测脑电信号的预测信号特征（S20）；根据信号特征以及预测信号特性生成脑电信号对应的模拟脑电信号，并根据模拟脑电信号确定麻醉深度值（S30）。通过将网络脑模型生成的预测脑电信号与实际测量得到的脑电信号相结合得到模拟脑电信号，再根据该模拟脑电信号计算麻醉深度，这样可以通过预测脑电信号对实际测量脑电信号进行验证，可以提高测量得到的麻醉深度的准确性，提高麻醉深度的可靠性，并且提高麻醉深度测量的动态实时跟踪能力和抗干扰能力。

Description

一种麻醉深度的测量方法、存储介质及电子设备 技术领域
本发明涉及超声技术领域，特别涉及一种麻醉深度的测量方法、存储介质及电子设备。
背景技术
在进行重大手术前，需要对患者进行全身麻醉(俗称全麻)，而全麻具有极高风险，例如，麻醉过深会引起麻醉并发症，甚至危及患者生命；麻醉过浅易发生“术中知晓”，产生痛苦和恐惧及精神后遗症；因此如何精确地估算麻醉深度，使得医生可以根据麻醉深度确定麻醉用量，以提高麻醉安全性。
目前基于脑电信号展开的麻醉深度监测的方法主要包括双频指数、听觉诱发电位指数、脑功能状态指数,熵指数，复杂度和小波分析法等。例如，美国Aspect公司(now part of Covidien)主推出的BIS监护仪，其实采用双谱分析方便，并通过0-100的指数来反映麻醉意识深度。然而，上述方法中普遍采用的是患者的真实脑电信号，而真实脑电信号在采集过程中会存在干扰信号等，造成真实脑电信号异常。此时，医生需要根据血压、心率、呼吸频率、肌松程度等间接指标来判断病人的麻醉状态，这需要依赖医生的业务水平，并且不同医生的判断结果也会存在不同。
发明内容
鉴于现有技术的不足，本发明旨在提供一种麻醉深度的测量方法、存储介质及电子设备。
本发明所采用的技术方案如下：
一种麻醉深度的测量方法，其包括：
获取脑电信号，并提取所述脑电信号的信号特征；
获取所述脑电信号对应的预测脑电信号，并提取所述预测脑电信号的预测信号特征，其中，所述预测脑电信号为通过预设矩阵型脑网络生成；
根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号，并根据所述模拟脑电信号确定麻醉深度值。
所述麻醉深度的测量方法，其中，所述预测脑电信号的生成过程具体包括：
以预先建立的神经元群模块为元素建立矩阵型脑网络；
根据预设联通参数驱动所述矩阵型脑网络产生脑电信号，其中，所述预设联通参数用于控制所述矩阵型脑网络中各神经元群模块状态。
所述麻醉深度的测量方法，其中，所述神经元群模块包括激励神经细胞网络、锥体细胞网络以及抑制神经细胞网络；所述激励神经细胞网络的激励信号、抑制神经细胞网络的抑制信号、以及锥体细胞网络的外部激励信号形成脑电波，并将所述脑电波输出至外部的神经元群模块，并且所述脑电波分别反馈至该神经元群模块的激励神经细胞网络以及抑制神经细胞网络。
所述麻醉深度的测量方法，其中，所述获取脑电信号，并提取所述脑电信号的信号特征具体包括：
获取脑电信号，并对所述脑电信号进行去噪处理；
提取去噪处理后的脑电信号的信号特征。
所述麻醉深度的测量方法，其中，所述根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号具体包括：
将所述信号特征中各子信号特征分别与所述预测信号特征中对应的预测信号特征进行匹配；
根据所述匹配结果确定脑电信号对应的模拟信号特征，并根据所述模拟信号特征生成所述模拟脑电信号。
所述麻醉深度的测量方法，其中，所述根据所述匹配结果确定脑电信号对应的模拟信号特征具体包括：
对于匹配结果为相同的子信号特征，将该子信号特征作为该子信号特征对应的子模拟信号特征；
对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征；
根据确定所有子模拟信号特征生成所述脑电信号对应的模拟信号特征。
所述麻醉深度的测量方法，其中，所述对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征具体包括：
对于匹配结果为不相同的子信号特征，判断是否通过线性插值方式将该子信号特征匹配至其对应的子预测信号特征；
若是，则将该子信号特征进行线性插值处理以得到该子信号特征对应的子模拟信号特征。
所述麻醉深度的测量方法，其中，所述对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征包括：
若否，则获取所述脑电信号的信噪比；
当所述信噪比满足预设条件时，采用移动窗平移算法确定该子信号特征对应的子模拟信号特征；
当所述信噪比未满足预设条件时，将该子信号特征对应的子预测信号特征作为该子信号特征对应的子模拟信号特征；
当未检测到信噪比时，将该子信号特征与该子信号特征对应的子预测信号特征进行加权处理，以得到该子信号特征对应的子模拟信号特征。
一种计算机可读存储介质，所述计算机可读存储介质存储有一个或者多个程序，所述一个或者多个程序可被一个或者多个处理器执行，以实现如上任一所述的麻醉深度的测量方法中的步骤。
一种电子设备，其包括：处理器、存储器及通信总线；所述存储器上存储有可被所述处理器执行的计算机可读程序；
所述通信总线实现处理器和存储器之间的连接通信；
所述处理器执行所述计算机可读程序时实现如上任一所述的麻醉深度的测量方法中的步骤。
有益效果：与现有技术相比，本发明提供了一种麻醉深度的测量方法、存储介质及电子设备，所述方法包括获取脑电信号，并提取所述脑电信号的信号特征；获取所述脑电信号对应的预测脑电信号，并提取所述预测脑电信号的预测信号特征；根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号，并根据所述模拟脑电信号确定麻醉深度值。本发明通过将通过网络脑模型生成的预测脑电信号与实际测量得到的脑电信号相结合得到模拟脑电信号，再根据该模拟脑电信号计算麻醉深度，这样可以通过预测脑电信号对实际测量脑电信号进行验证，可以提高测量得到的麻醉深度的准确性，提高麻醉深度的可靠性，并且提高麻醉深度测量的动态实时跟踪能力和抗干扰能力。
附图说明
图1为本发明提供的麻醉深度的测量方法的流程图。
图2为本发明提供的麻醉深度的测量方法中神经元群模块的示意图。
图3为本发明提供的麻醉深度的测量方法中人脑模型的一个角度的示意图。
图4为本发明提供的麻醉深度的测量方法中人脑模型的另一个角度的示意图。
图5为本发明提供的电子设备的结构原理图。
具体实施方式
本发明提供一种麻醉深度的测量方法、存储介质及电子设备，为使本发明的目的、技术方案及效果更加清楚、明确，以下参照附图并举实施例对本发明进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本发明，并不用于限定本发明。
本技术领域技术人员可以理解，除非特意声明，这里使用的单数形式“一”、“一个”、“所述”和“该”也可包括复数形式。应该进一步理解的是，本发明的说明书中使用的措辞“包括”是指存在所述特征、整数、步骤、操作、元件和/或组件，但是并不排除存在或添加一个或多个其他特征、整数、步骤、操作、元件、组件和/或它们的组。应该理解，当我们称元件被“连接”或“耦接”到另一元件时，它可以直接连接或耦接到其他元件，或者也可以存在中间元件。此外，这里使用的“连接”或“耦接”可以包括无线连接或无线耦接。这里使用的措辞“和/或”包括一个或更多个相关联的列出项的全部或任一单元和全部组合。
本技术领域技术人员可以理解，除非另外定义，这里使用的所有术语(包括技术术语和科学术语)，具有与本发明所属领域中的普通技术人员的一般理解相同的意义。还应该理解的是，诸如通用字典中定义的那些术语，应该被理解为具有与现有技术的上下文中的意义一致的意义，并且除非像这里一样被特定定义，否则不会用理想化或过于正式的含义来解释。
下面结合附图，通过对实施例的描述，对发明内容作进一步说明。
本实施提供了一种麻醉深度的测量方法，如图1和图2所示，所述方法包括：
S10、获取脑电信号，并提取所述脑电信号的信号特征。
具体地，所述脑电信号可以通过EEG采集设备采集得到，也可以通过电极片获取得到，还可以通过传感器采集得到，并且通过传感器采集到所述脑电信号可以是模拟脑电信号，并且在采集到所述模拟脑电信号可以将模拟脑电信号转换为数字脑电信号，并将所述数字脑电信号作为所述脑电信号，这样可以便于脑电信号的存储。当然，值得说 明的，EEG采集设备可以为本领域技术人员所公知的各种采集脑电信号的设备。此外，所述脑电信号所述脑电信号可以是患者在麻醉全程中各段脑电信号，例如，每段脑电信号的信号长度为10秒，并通过滑动窗口技术，通过步长控制每秒脑电信号数据的更新。
所述信号特征可以包括时域特征、频域特征和非频域特征中的至少一种。在本实施例的一个实现方式中，所述信号特征包括时域特征或频域特征，即在获取到脑电信号后，提取该脑电信号的时域特征或频域特征。
进一步，在本实施例的一个实现方式中，所述获取脑电信号，并提取所述脑电信号的信号特征具体包括：
S11、获取脑电信号，并对所述脑电信号进行去噪处理；
S12、提取去噪处理后的脑电信号的信号特征。
具体地，所述去噪处理为对生理干扰信号和非生理干扰信号进行识别和处理。对脑电信号进行去噪处理，可以提高脑电信号的指令，由此，去除干扰信号能够更加准确的麻醉深度值。在本实施例的一个实现方式中，所述生理干扰信号可以包括眼动干扰信号和肌电干扰信号，所述非生理干扰信号可以包括信号幅度异常、信号斜率异常以及电刀干扰等。由此，为了去除生理干扰信号和非生理干扰信号，可以对所述脑电信号进行数字滤波，其中，所述数字滤波可以包括低通滤波，高通滤波以及特征波删除等方式。
S20、获取所述脑电信息对应的预测脑电信息，并提取所述预测脑电信号的预测信号特征，其中，所述预测脑电信息为通过预设矩阵型脑网络生成。
具体地，所述预设矩阵型脑网络为预先建立，并通过矩阵型脑网络可以生成预测脑电信号。所述预测脑电信号可以以根据预设麻醉方案确定脑电信号对应的麻醉数据，并根据所述麻醉数据实时改变模型参数动态地获取脑电信号对应的预测脑电信号。所述预测脑电信号可以根据预设麻醉方案中麻醉数据，通过改变模型参数提前离线地获取并存储备用麻醉过程中的预测脑电信号。在本实施例的一个可能实现方式中，所述预测脑电信号的获取过程为：在确定麻醉方案后，根据麻醉方案确定各时刻对应的模型参数，之后通过各时刻对应的模型参数按照时间顺序输入至矩阵型脑网，可以得到麻醉过程中的预测脑电信号，在得到预测脑电信号后，可以存储该麻醉过程对应的预测脑电信号。而所述脑电信号对应的预测脑电信号，可以是根据所述脑电信号对应的时间段在预设矩阵型脑网络生成的预测脑电信号中选取该时间段对应的预测脑电信号，以得到所述脑电信号对应的预测脑电信号，这样可以减少预测脑电信号的获取时间，提高麻醉深度获取的实时性。此外，所述麻醉数据可以包括麻醉时间以及进入麻醉状态的速度等。
进一步，在本实施例的一个实现方式中，所述预测脑电信息的生成过程具体包括：
M10、以预先建立的神经元群模块为元素建立矩阵型脑网络；
M20、根据预设联通参数驱动所述矩阵型脑网络产生脑电信号，其中，所述预设联通参数用于控制所述矩阵型脑网络中各神经元群模块状态。
具体地，所述神经元群模块为预选建立的，每个神经元群模块表示一脑皮层区域，所述矩阵脑网络包括多个神经元群模块，并且各神经元群模块对应的脑皮层区域构成脑皮层，即脑皮层可以划分为多个脑皮层区域，每个脑皮层对于一个神经元群模块，并且该神经元群模块包含于形成所述矩阵型脑网络的多个神经元群模块中。同时所述多个脑皮层区域中的每个脑皮层区域负责不同的功能，例如，枕叶区负责视觉功能；顶叶区负责触觉和空间；颞叶区负责听觉和综合感知；额叶运动皮层负责协调肢体运动；前额皮层区负责理解、记忆和判断等。在本实施例的一个可能实现方式中，所述神经元群模块包括76个，即将左右脑皮层共划分为76个区域，相应的，所述矩阵型脑网络具有76个元素，目前76个元素是我们能够达到的最好结果，当然更多元素会使模型更加准确。理论上越多的JR模型能模拟出来的脑电信号越接近真实脑电信号，但会带来系统效率低下。目前选择76这个值只是在保证模拟的脑电信号有效的前提下在系统运行性能与结果复杂度之间平衡的结果。
进一步，脑电信号EEG是大量神经元活动的反映，特别是在麻醉过程中，脑电信号EEG变化具有很强的非线性特征。其中，麻醉过程中脑电信号EEG是根据丘脑和脑皮层间的大量神经元会产生振荡耦合形成，并且且随着麻醉深度的变化，耦合强度会有明显不同，以使得脑电信号EEG变化具有很强的非线性特征，此外，在丘脑和脑皮层间的振荡耦合过程中，丘脑的功能很像是一个知觉“开关”或整体脑活动的显示器，并且丘脑和脑皮层神经网络的规则性活动是麻醉和知觉的关键。例如，在开始进入睡眠状态和维持睡眠的过程中，有一个正向的抑制过程，该抑制过程是由前脑底部和丘脑下部经由GaBa+离子传递到觉醒核心，同时觉醒核心同时向失觉核心发出以反馈过程，这说明当觉醒核心受到抑制时，该反馈正过程加强了失觉核心的神经活动，从而引发一个类似于单稳触发电路一样的有知觉和无知觉状态，并且人脑始终处于一个有知觉或无知觉的状态。由此，基于预设神经元群模块构建一个分部平滑的过程模型，即矩阵型脑网络，所述矩阵型脑网络是以预设建立的神经元群模块为元素，并按照各神经元群模块之间的耦合强度建立矩阵型脑网络，同时所述矩阵型脑网络可以把微观的脑细胞活动，同介观的脑皮层区活动和宏观的脑电图信号号结合起来，以使得通过所述矩阵型脑网络可以模 拟脑电信号。
进一步，在本实施例的一个实现方式中，所述神经元群模块为JANSEN-RIT(J-R)模块，所述JANSEN-RIT模块包括激励神经细胞网络、锥体细胞网络以及抑制神经细胞网络；所述激励神经细胞网络的激励信号、抑制神经细胞网络的抑制信号、以及锥体细胞网络的外部激励信号形成脑电波，并将所述脑电信号输出至外部的神经元群模块。如图2所示，所述激励神经细胞网络、锥体细胞网络以及抑制神经细胞网络均可以表示为神经环路。所述激励神经细胞网络包括第一脉冲支路以及第一反馈支路，第一脉冲支路的电压信号与第一反馈支路的电压信号相加后，通过电压-脉冲编码器将相加后的电压信号转换为脉冲信号后传输至锥体细胞网络；所述抑制神经细胞网络包括第二脉冲支路以及第二反馈支路，第二脉冲支路的电压信号与第二反馈支路的电压信号相加后，通过电压-脉冲编码器将相加后的电压信号转换为脉冲信号后传输至锥体细胞网络；锥体细胞网络包括局部激励支路、局部抑制支路以及外部激励支路，局部激励支路接收激励神经细胞网络传输的脉冲信号，局部抑制支路接收抑制神经网络传输的脉冲信号，外部激励支路接收输入脉冲，并且所述局部激励支路的、局部抑制支路以及外部激励支路汇合后形成脑电波信号，所述脑电波信号通过电压-脉冲编码器后输出至其他神经元群模块，通过分别反馈至该JANSEN-RIT模块的激励神经细胞网络以及抑制神经细胞网络。此外，如图2所示，在所述S i,i＝1,2,3为电压-脉冲编码器，用于将细胞内的电压信号转换成轴突丘上的脉冲信号；
Figure PCTCN2020106066-appb-000001
以及
Figure PCTCN2020106066-appb-000002
均为是脉冲-电压译码器，用于把收到的脉冲信号进行加权卷积后转换成突触后的电压，其中，其中，1T、2T以及3T均为外部输出信号的神经元群模块的编号；
Figure PCTCN2020106066-appb-000003
为加法器，用于把突触后的树突电压进行整合转换成细胞内的电压信号；C 1T，C 13，C 2T，C 23，C 31，C 32以及C 3T均为神经网络群之间平均突触数；m 1T(t)，m 1(t)，m 2T(t)，m 2(t)，m 3(t)以及m 3T(t)均为脉冲信号，v 1T(t)，v 13(t)，v 1(t)，v 2T(t)，v 23(t)，v 2(t)，v 3T(t)，v 31(t)，v 32(t)以及v 3(t)均为电压信号，其中，v 3(t)为可被检测到的脑电信号。
进一步，在本实施例的一个实现方式中，所述以预先建立的神经元群模块为元素建议具有空间特性的矩阵型脑网络具体包括：
S11、获取用户的脑部数据，并根据所述脑部数据生成结构型脑网络；
S12、获取该结构型脑网络中神经网络节点的耦合强度系数，并根据获取到的耦合强度系数形成矩阵序列，以得到矩阵型脑网络。
具体地，所述脑部数据为根据解剖学获取得到，根据所述脑部数据可以建议一个结构脑网络，例如，所述结构脑网络包括76个区域。在建立所述结构脑网络后，对于结构脑网络的每个区域，获取该区域中各神经网络节点之间的耦合强度系数，并根据获取到所有耦合强度系数确定该区域的矩阵，这样结构型脑网络的动态演变就会产生一系列脑网络矩阵，而在不同时间点的动态脑网络矩阵形成矩阵型脑网络。此外，所述脑部数据包括脑皮层区域和各脑皮层区域的关联性，其中，各脑皮层区域的关联强度可以用0，1和-1表示，也可以是归一化后从-1到+1的任何值，而各种脑活动则可以用一组描述大规模脑神经网络的动态时域方程表示，这些方程可以根据研究的目的进行选择，例如，我们可以选择如下的方程组对常规的脑电信号进行模拟：
Figure PCTCN2020106066-appb-000004
其中，
Figure PCTCN2020106066-appb-000005
代表节点i在时间t时基于局部动态函数f(x i(t))的平均电场，wij是连接节点i和j的各向同质电导矩阵，g是全局耦合函数，(t-Δt ij)代表时间延迟。
此外，上述动态时域方程中的参数可以由已知生理参数值，包括不限于模型中各种脑组织的导电率等来确定。
进一步，所述预设联通参数用于表示各J-R模块的连通状态，其中，所述联通参数包括未联通、弱联通、正常联通以及强联通，例如，分别采用0，1，2，和3表示。此外，所述J-R模块是由如图2所示的三个神经网络环路组成，并且能够根据预设参数而产生脑电波，其中，所述预设参数可以为脑组织导电系数等，所述脑组织导电系数可以包括头皮，头盖骨，脑组织液，白色组织，灰色组织以及血管等。
进一步，每一个神经网路环路都是用其关键状态变量来表示，例如，平均膜电位，平均激活率和它们之间的相互转换(脉冲-波电位及波电位-脉冲)等，这样可以根据震荡频率和锥体的平均PSP基准线确定各J-R模块的连通状态，即根据预先选定用于控制状态转移的联通参数，可以使得各个脑皮区之间的相互作用将会产生诸如多稳、同步以及直至协同等状态，从而产生相应的脑电波信号。
进一步，在本实施例的一个实现方式中，所述根据预设联通参数驱动所述矩阵型脑网络产生脑电信号具体包括：
M21、获取丘脑以及脑皮层的神经活动；
M22、根据所述连通参数驱动所述矩阵型脑网络，通过所述矩阵型脑网络确定所述神经活动对应的电位信息，并根据所述电位信息得到脑电信号。
具体地，所述丘脑以及脑皮层的神经活动为已知的，可以通过CT或者磁共振等途径获取到，在获取到丘脑和脑皮层的神经活动后，根据丘脑和脑皮层的神经活动确定各神经元群模块的连通参数，并通过所述连通参数驱动所述矩阵型脑网络，以使得所述通过所述矩阵型脑网络确定所述神经活动对应的电位信息，并根据所述电位信息得到脑电信号。
进一步，在本实施例的一个实现方式中，所述通过所述矩阵型脑网络确定所述神经活动对应的电位信息，并根据所述电位信息得到脑电信号；
对丘脑以及脑皮层对应的脑部图像进行识别以区分脑部组分，并将各脑部组分与其对应的电质参数相结合，以得到人脑模型；
根据所述人脑模型确定脑部头皮的各电极点单元的电位，并根据确定到的所有电位生成所述脑电信号。
具体地，所述脑部图像可以是通过核磁共振获取到，并且所述脑部图像包括多张连续核磁共振图像，以便于可以获取到脑部各部位的图像。在获取到脑部图像后，对所述脑部图像进行识别以确定脑部组分，其中，所述脑部组分包括盖骨，脑液，脑灰色物质，脑白色物质等几十种人头组织。在识别到脑部组分后，可以将脑部组分与已知电质参数，例如，头皮，头盖骨，脑组织液，白色组织，灰色组织以及血管等，相结合而建立以数字人头模型，即得到人脑模型，如图3和4所示。此外，在建立人脑模型后，将该数字化的人脑模型抽象成几百万个小的单元，每一个单元代表人脑的不同部位和组织，并被赋予相应的电质参数，这样每一个单元不仅可以代表不同的人头组织，也可以代表同样组织的不同电质参数，之后确定数字化的人脑模型中各个小单元的点位，然后从中找出头皮上对应电极点的单元的点位，以得到脑电信号。
S30、根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号，并根据所述模拟脑电信号确定麻醉深度值。
具体地，所述根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号指的是将所述信号特征与预测信号特征进行匹配，以将所述脑电信号进行划分，分别划分为匹配信号集、线性插值信号集以及完全不匹配信号。其中，所述匹配信号集中的脑电信号的信号特征与其对应的预测信号特征相匹配；所述线性插值信号集中的脑电信号的信号特征可以通过线性插值的方式匹配至其对应的预测信号特征；所述完全不匹配信号中的脑电信号的信号特征与其对应的预测信号特征不匹配，并且无法通过线性插值的方式匹配至其对应的预测信号特征。此外，为了快速计算麻醉深度，匹配信号集、 线性插值信号集以及完全不匹配信号中的脑电信号可以配置不同的麻醉估算标识，例如，匹配信号集中脑电信号配置有自动标识，线性插值信号集中脑电信号配置有半自动标识，完全不匹配信号中的脑电信号在根据预测脑电信号进行修正后，配置有半自动标识。这样在对信号特征与预测信号特征匹配后，可以根据麻醉估算标识确定麻醉深度。
进一步，在本实施例的一个实现方式中，所述根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号具体包括：
将所述信号特征中各子信号特征分别与所述预测信号特征中对应的预测信号特征进行匹配；
根据所述匹配结果确定脑电信号对应的模拟信号特征，并根据所述模拟信号特征生成所述模拟脑电信号。
具体地，所述匹配结果包括匹配、通过线性插值后匹配以及不匹配且通过线性插值也无法匹配。对于不同的匹配结果的脑电信号的处理方式不同，特征是对于不匹配且通过线性插值也无法匹配的脑电信号，可以通过预测脑电信号对所述脑电信号进行修正，并根据修正后的脑血信号生成模拟脑血信号，这样可以提高模拟脑电信号的准确性，从而提高麻醉深度测量的准确性，
进一步，在本实施例的一个实现方式中，所述根据所述匹配结果确定脑电信号对应的模拟信号特征具体包括：
对于匹配结果为相同的子信号特征，将该子信号特征作为该子信号特征对应的子模拟信号特征；
对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征；
根据确定所有子模拟信号特征生成所述脑电信号对应的模拟信号特征。
具体地，所述匹配结果为相同，说明信号特征与预测信号特征一样，由此，可以确定该信号特征对应的脑血信号为准确脑电信号，从而可以将该子信号特征作为该子信号特征对应的子模拟信号特征。而对于匹配结果为不相同的子信号特征，则需要对该子信号特征进行处理，以使得所述子信号特征与其对应的预测信号特征匹配，而当无法使得子信号特征与其对应的预测信号特征匹配是，根据所述预测信号特征对所述该子信号特征进行修正。相应的，在本实施例的一个实现方式中，所述对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征具体包括：
对于匹配结果为不相同的子信号特征，判断是否通过线性插值方式将该子信号特征匹配至其对应的子预测信号特征；
若是，则将该子信号特征进行线性插值处理以得到该子信号特征对应的子模拟信号特征；
若否，则获取所述脑电信号的信噪比；
当所述信噪比满足预设条件时，采用移动窗平移算法确定该子信号特征对应的子模拟信号特征；
当所述信噪比未满足预设条件时，将该子信号特征对应的子预测信号特征作为该子信号特征对应的子模拟信号特征；
当未检测到信噪比时，将该子信号特征与该子信号特征对应的子预测信号特征进行加权处理，以得到该子信号特征对应的子模拟信号特征。
具体地，所述对非吻合部分将用线性插值的方法进行数据整合后产生一个半自动的标识，对于不匹配的但又无法进行插值整合的部分，分以下三种情况进行处理：1)如果实测脑电信噪比在预设范围内，应用移动窗平滑算法将实测脑电信号的信号特征转化为子模拟信号特征，2)如果实测脑电信噪比在预设范围外，将预测脑电信号的预测信号特征转化为子模拟信号特征，3)如果得不到实测脑电信噪比，则用信号特征A和预测信号特征B两者的加权值(αA+(1-α)B)，以期得到子模拟信号特征。其中，所述预测信噪比对应的预设范围可以随选取的麻醉特征而定，例如，在10％到50％之间。在此，在匹配是一个过程，由于麻醉是一个受控制的渐进的连续过程，中间不可以有间断，从而在不匹配时，可以进行线性插值以使得信号特征匹配至预测信号特征。
进一步，在确定模拟脑电信号后，可以根据模拟脑电信号确定麻醉深度，其中，所述麻醉深度可以经过一个类似百分比的归一化处理给出一个0到100的深度指数，例如，所述指数可以如下图表所示，反映病人的麻醉深度。
麻醉深度指数 麻醉状态 生理/物理特征
100 清醒 各项反应正常
80 轻度麻醉 血压心律升高，对强刺激有反应
60 全麻醉 血压稳定，无知觉无反应
40 深度麻醉 血压高不稳，心律不齐
20 爆发抑制 间歇性脑电活动
0 无脑电活动 无脑电活动
基于上述麻醉深度的测量方法，本实施例提供了一种计算机可读存储介质，所述计算机可读存储介质存储有一个或者多个程序，所述一个或者多个程序可被一个或者多个处理器执行，以实现如上述实施例所述的麻醉深度的测量方法中的步骤。
基于上述麻醉深度的测量方法，本发明还提供了一种电子设备，如图5所示，其包括至少一个处理器(processor)20；显示屏21；以及存储器(memory)22，还可以包括通信接口(Communications Interface)23和总线24。其中，处理器20、显示屏21、存储器22和通信接口23可以通过总线24完成相互间的通信。显示屏21设置为显示初始设置模式中预设的用户引导界面。通信接口23可以传输信息。处理器20可以调用存储器22中的逻辑指令，以执行上述实施例中的方法。
此外，上述的存储器22中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。
存储器22作为一种计算机可读存储介质，可设置为存储软件程序、计算机可执行程序，如本公开实施例中的方法对应的程序指令或模块。处理器20通过运行存储在存储器22中的软件程序、指令或模块，从而执行功能应用以及数据处理，即实现上述实施例中的方法。
存储器22可包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据电子设备的使用所创建的数据等。此外，存储器22可以包括高速随机存取存储器，还可以包括非易失性存储器。例如，U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等多种可以存储程序代码的介质，也可以是暂态存储介质。
此外，上述存储介质以及移动终端中的多条指令处理器加载并执行的具体过程在上述方法中已经详细说明，在这里就不再一一陈述。
最后应说明的是：以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims (10)

  1. 一种麻醉深度的测量方法，其特征在于，其包括：
    获取脑电信号，并提取所述脑电信号的信号特征；
    获取所述脑电信号对应的预测脑电信号，并提取所述预测脑电信号的预测信号特征，其中，所述预测脑电信号为通过预设矩阵型脑网络生成；
    根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号，并根据所述模拟脑电信号确定麻醉深度值。
  2. 根据权利要求1所述麻醉深度的测量方法，其特征在于，所述预测脑电信号的生成过程具体包括：
    以预先建立的神经元群模块为元素建立矩阵型脑网络；
    根据预设联通参数驱动所述矩阵型脑网络产生脑电信号，其中，所述预设联通参数用于控制所述矩阵型脑网络中各神经元群模块状态。
  3. 根据权利要求2所述麻醉深度的测量方法，其特征在于，所述神经元群模块包括激励神经细胞网络、锥体细胞网络以及抑制神经细胞网络；所述激励神经细胞网络的激励信号、抑制神经细胞网络的抑制信号、以及锥体细胞网络的外部激励信号形成脑电波，并将所述脑电波输出至外部的神经元群模块，并且所述脑电波分别反馈至该神经元群模块的激励神经细胞网络以及抑制神经细胞网络。
  4. 根据权利要求1所述麻醉深度的测量方法，其特征在于，所述获取脑电信号，并提取所述脑电信号的信号特征具体包括：
    获取脑电信号，并对所述脑电信号进行去噪处理；
    提取去噪处理后的脑电信号的信号特征。
  5. 根据权利要求1所述麻醉深度的测量方法，其特征在于，所述根据所述信号特征以及预测信号特性生成所述脑电信号对应的模拟脑电信号具体包括：
    将所述信号特征中各子信号特征分别与所述预测信号特征中对应的预测信号特征进行匹配；
    根据所述匹配结果确定脑电信号对应的模拟信号特征，并根据所述模拟信号特征生成所述模拟脑电信号。
  6. 根据权利要求5所述麻醉深度的测量方法，其特征在于，所述根据所述匹配结果确定脑电信号对应的模拟信号特征具体包括：
    对于匹配结果为相同的子信号特征，将该子信号特征作为该子信号特征对应的子模拟信号特征；
    对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征；
    根据确定所有子模拟信号特征生成所述脑电信号对应的模拟信号特征。
  7. 根据权利要求6所述麻醉深度的测量方法，其特征在于，所述对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征具体包括：
    对于匹配结果为不相同的子信号特征，判断是否通过线性插值方式将该子信号特征匹配至其对应的子预测信号特征；
    若是，则将该子信号特征进行线性插值处理以得到该子信号特征对应的子模拟信号特征。
  8. 根据权利要求7所述麻醉深度的测量方法，其特征在于，所述对于匹配结果为不相同的子信号特征，将该子信号特征进行处理，并将处理后的子信号特征作为该子信号特征对应的子模拟信号特征包括：
    若否，则获取所述脑电信号的信噪比；
    当所述信噪比满足预设条件时，采用移动窗平移算法确定该子信号特征对应的子模拟信号特征；
    当所述信噪比未满足预设条件时，将该子信号特征对应的子预测信号特征作为该子信号特征对应的子模拟信号特征；
    当未检测到信噪比时，将该子信号特征与该子信号特征对应的子预测信号特征进行加权处理，以得到该子信号特征对应的子模拟信号特征。
  9. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质存储有一个或者多个程序，所述一个或者多个程序可被一个或者多个处理器执行，以实现如权利要求1～8任意一项所述的麻醉深度的测量方法中的步骤。
  10. 一种电子设备，其特征在于，包括：处理器、存储器及通信总线；所述存储器上存储有可被所述处理器执行的计算机可读程序；
    所述通信总线实现处理器和存储器之间的连接通信；
    所述处理器执行所述计算机可读程序时实现如权利要求1-8任意一项所述的麻醉深度的测量方法中的步骤。
PCT/CN2020/106066 2019-12-06 2020-07-31 一种麻醉深度的测量方法、存储介质及电子设备 WO2021109601A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911245011.5 2019-12-06
CN201911245011.5A CN110840411B (zh) 2019-12-06 2019-12-06 一种麻醉深度的测量装置、存储介质及电子设备

Publications (1)

Publication Number Publication Date
WO2021109601A1 true WO2021109601A1 (zh) 2021-06-10

Family

ID=69608120

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/106066 WO2021109601A1 (zh) 2019-12-06 2020-07-31 一种麻醉深度的测量方法、存储介质及电子设备

Country Status (2)

Country Link
CN (1) CN110840411B (zh)
WO (1) WO2021109601A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117281534A (zh) * 2023-11-22 2023-12-26 广东省人民医院 一种多指标的麻醉状态监测方法及系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110840411B (zh) * 2019-12-06 2022-03-11 深圳市德力凯医疗设备股份有限公司 一种麻醉深度的测量装置、存储介质及电子设备
CN113133744A (zh) * 2021-04-30 2021-07-20 鹤壁市人民医院 一种多功能麻醉科用麻醉深度监测装置
CN114521904B (zh) * 2022-01-25 2023-09-26 中山大学 一种基于耦合神经元群的脑电活动模拟方法及系统
CN116491960B (zh) * 2023-06-28 2023-09-19 南昌大学第一附属医院 脑瞬态监测设备、电子设备及存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040082876A1 (en) * 2000-10-16 2004-04-29 Viertio-Oja Hanna E. Method and apparatus for determining the cerebral state of a patient with fast response
KR20120131036A (ko) * 2011-05-24 2012-12-04 한국과학기술원 뇌활성도 및 마취심도 변화에 따른 ｅｅｇ신호 모델 및 시뮬레이터
CN105769146A (zh) * 2016-03-24 2016-07-20 美合实业（苏州）有限公司 一种多监测指标的麻醉深度监护仪
CN107595247A (zh) * 2017-08-29 2018-01-19 深圳市德力凯医疗设备股份有限公司 一种基于脑电信号的麻醉深度的监测方法及系统
US20190069841A1 (en) * 2017-09-05 2019-03-07 Korea University Research And Business Foundation Method and apparatus of monitoring anaesthesia and consciousness depth through brain network analysis
CN109645989A (zh) * 2018-12-10 2019-04-19 燕山大学 一种麻醉深度估计方法及系统
WO2019127558A1 (zh) * 2017-12-29 2019-07-04 深圳迈瑞生物医疗电子股份有限公司 基于脑电的麻醉深度监测方法和装置
CN110840411A (zh) * 2019-12-06 2020-02-28 深圳市德力凯医疗设备股份有限公司 一种麻醉深度的测量方法、存储介质及电子设备

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101259015B (zh) * 2007-03-06 2010-05-26 李小俚 一种脑电信号分析监测方法及其装置
JP4836140B2 (ja) * 2007-03-23 2011-12-14 独立行政法人産業技術総合研究所 脳活動解析方法および装置
CN101488162B (zh) * 2008-01-17 2012-03-21 复旦大学 一种用于脑电信号自动评估的脑电信号特征提取方法
US8588899B2 (en) * 2010-03-24 2013-11-19 Steven John Schiff Model based control of Parkinson's disease
JP2012146116A (ja) * 2011-01-12 2012-08-02 Kyushu Institute Of Technology スピーチ内容を識別する装置及び方法
CN102509282B (zh) * 2011-09-26 2014-06-25 东南大学 一种融合结构连接的各脑区间的效能连接分析方法
CN102499675B (zh) * 2011-10-27 2013-09-18 杭州电子科技大学 一种皮层脑电信号的反馈系统随机共振增强方法
CN103006211B (zh) * 2013-01-17 2015-01-07 西安电子科技大学 一种基于脑电网络分析的地形图描绘装置
AU2017349924A1 (en) * 2016-10-25 2019-05-23 Bgn Technologies Ltd. Apparatus and methods for predicting therapy outcome
KR20180059985A (ko) * 2016-11-28 2018-06-07 참엔지니어링(주) 히든 마르코프 모델을 이용한 마취 심도 측정 방법 및 장치
KR101911506B1 (ko) * 2016-12-27 2018-10-25 울산과학기술원 뇌신호 기반 3차원 상지운동 외부 보조기기 제어를 위한 가상 뇌파 생성장치와 시뮬레이션 장치 및 방법
US11241186B2 (en) * 2017-01-03 2022-02-08 Myndlift Ltd. Systems and methods for processing EEG signals of a neurofeedback protocol
CN109065128A (zh) * 2018-09-28 2018-12-21 郑州大学 一种加权图正则化稀疏脑网络构建方法
CN109701160A (zh) * 2019-01-23 2019-05-03 中国人民解放军总医院 影像引导下可见光定位导航装置及方法
CN110393525B (zh) * 2019-06-18 2020-12-15 浙江大学 一种基于深度循环自编码器的大脑活动检测方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040082876A1 (en) * 2000-10-16 2004-04-29 Viertio-Oja Hanna E. Method and apparatus for determining the cerebral state of a patient with fast response
KR20120131036A (ko) * 2011-05-24 2012-12-04 한국과학기술원 뇌활성도 및 마취심도 변화에 따른 ｅｅｇ신호 모델 및 시뮬레이터
CN105769146A (zh) * 2016-03-24 2016-07-20 美合实业（苏州）有限公司 一种多监测指标的麻醉深度监护仪
CN107595247A (zh) * 2017-08-29 2018-01-19 深圳市德力凯医疗设备股份有限公司 一种基于脑电信号的麻醉深度的监测方法及系统
US20190069841A1 (en) * 2017-09-05 2019-03-07 Korea University Research And Business Foundation Method and apparatus of monitoring anaesthesia and consciousness depth through brain network analysis
WO2019127558A1 (zh) * 2017-12-29 2019-07-04 深圳迈瑞生物医疗电子股份有限公司 基于脑电的麻醉深度监测方法和装置
CN109645989A (zh) * 2018-12-10 2019-04-19 燕山大学 一种麻醉深度估计方法及系统
CN110840411A (zh) * 2019-12-06 2020-02-28 深圳市德力凯医疗设备股份有限公司 一种麻醉深度的测量方法、存储介质及电子设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"Chinese Doctoral Dissertations Full-Text Database, Medical and Health Sciences", vol. 3, 30 May 2017, ISSN: 2317-563X, article GENG, SHUJUAN: "Nonlinear Dynamic Analysis of Neural Mass Model Based on Bifurcation Theory", XP055819698, DOI: 10.17565/gesta.v3i2.15098 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117281534A (zh) * 2023-11-22 2023-12-26 广东省人民医院 一种多指标的麻醉状态监测方法及系统
CN117281534B (zh) * 2023-11-22 2024-03-22 广东省人民医院 一种多指标的麻醉状态监测方法及系统

Also Published As

Publication number Publication date
CN110840411B (zh) 2022-03-11
CN110840411A (zh) 2020-02-28

Similar Documents

Publication Publication Date Title
WO2021109601A1 (zh) 一种麻醉深度的测量方法、存储介质及电子设备
WO2021109600A1 (zh) 一种生成脑电信号的方法、存储介质及电子设备
JP6322194B2 (ja) ニューロフィードバックシステム
CN204931634U (zh) 基于生理信息的抑郁症评估系统
CN106333652B (zh) 一种睡眠状态分析方法
CN110353704B (zh) 基于穿戴式心电监测的情绪评估方法与装置
US20150065813A1 (en) System for recording and processing signal for diagnosing auditory system and method for recording and processing signal for diagnosing auditory system
CN105125186B (zh) 一种确定干预治疗方式的方法及系统
CN109157232A (zh) 心率变异性反馈训练辅助方法、装置、设备及存储介质
Kokonozi et al. A study of heart rate and brain system complexity and their interaction in sleep-deprived subjects
CN103816007B (zh) 一种基于脑电频域特征指标化算法的耳鸣治疗设备及方法
Alshaikhli et al. A study on the effects of EEG and ECG signals while listening to Qur'an recitation
PL198745B1 (pl) Sposób monitorowania sygnałów wywoływanych słuchowo potencjałów wskazujących poziom przytomności pacjenta
Zeng et al. Classifying driving fatigue by using EEG signals
AU2001213824A1 (en) Monitoring auditory evoked potentials
CN114648040A (zh) 生命体征多生理信号提取、融合分析方法
CN110931123B (zh) 一种矩阵型脑网络及其构建方法
CN116313029B (zh) 一种数字针灸动态控制优化的方法、系统和装置
Ruan et al. Feature extraction of SSVEP-based brain-computer interface with ICA and HHT method
CN106491118A (zh) 基于安卓系统的心电图机的实时心率计算系统及其方法
CN111248900A (zh) 一种基于单通道心脑信息耦合分析方法及系统
CN115192907A (zh) 一种实时生物反馈经皮迷走神经电子针灸装置
CN204745287U (zh) 一种基于心率变异性的催眠治疗抑郁症效果评价系统
Paulraj et al. Fractal feature based detection of muscular and ocular artifacts in EEG signals
CN109124605B (zh) 一种减少icu内误报警的方法、装置及设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20895395

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205N DATED 12.10.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20895395

Country of ref document: EP

Kind code of ref document: A1