A method and system for auxiliary identification of anesthesia depth

Abstract

The present invention relates to the field of electroencephalogram analysis technology, and specifically to an auxiliary identification method and system for anesthesia depth. The present invention first analyzes the change speed of the bispectral index to obtain the patient's anesthesia sensitivity coefficient; then divides the monitoring curve into a real-time monitoring segment and a historical monitoring segment according to the patient's anesthesia sensitivity coefficient; according to the difference in the fluctuation stability state of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, the early warning coefficient of the bispectral index at the current moment is obtained, and then the patient's anesthesia depth state is abnormally warned. The present invention first obtains the anesthesia sensitivity coefficient that reflects the individual characteristics of the patient, and further determines the analysis time domain of the local anesthesia depth state change at the current moment according to the anesthesia sensitivity coefficient, thereby improving the accuracy of real-time monitoring and identification of the patient's anesthesia depth state.

Description

Technical Field

[0001] The present invention relates to the technical field of electroencephalogram analysis, and in particular to an auxiliary identification method and system for anesthesia depth. Background Art

[0002] Proper anesthesia is crucial to the safety and success of the patient's surgery. If the anesthesia is too deep, the patient may experience serious complications such as respiratory depression and circulatory depression. If the anesthesia is insufficient, the patient may feel pain or have conscious reactions, increasing the risk of surgery. The Bispectral Index (BIS) reflects the patient's cortical and subcortical EEG activity and consciousness state. By real-time monitoring of the patient's BIS, the changing trend of the depth of anesthesia and the changes in the patient's consciousness state can be evaluated, thereby assisting in the precise control of the anesthesia process to ensure the safety of the patient's surgery.

[0003] However, long-term monitoring of BIS will over-smooth the patient's real-time BIS change trend, affecting the accuracy of the current anesthesia depth assessment; local analysis of the BIS change trend at the current moment can reduce the impact of the long historical time series. However, different patients may have different responses to anesthetic drugs, and the time series length selected for local analysis is affected by individual differences in patients, thus affecting the accuracy of real-time monitoring and identification of the anesthesia depth of different patients. Summary of the invention

[0004] In order to solve the technical problem of low accuracy in real-time monitoring and identification of anesthesia depth in the prior art, the present invention aims to provide an auxiliary identification method and system for anesthesia depth. The technical solution adopted is as follows:

[0005] Acquire a monitoring curve of the patient's bispectral index in real time, and acquire anesthesia induction segments in the monitoring curve;

[0006] According to the change rate of the bispectral index in the anesthesia induction segment, the anesthesia sensitivity coefficient of the patient is obtained; according to the anesthesia sensitivity coefficient of the patient, the monitoring curve is divided into a real-time monitoring segment and a historical monitoring segment;

[0007] According to the difference in the fluctuation stability of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, the early warning coefficient of the bispectral index at the current moment is obtained;

[0008] An abnormal warning is issued for the patient's anesthesia depth state according to the warning coefficient at the current moment.

[0009] Furthermore, the monitoring curve is divided into a real-time monitoring segment and a historical monitoring segment according to the anesthesia sensitivity coefficient of the patient, including:

[0010] The anesthesia sensitivity coefficient is negatively correlated and normalized to obtain an interception coefficient;

[0011] The interception coefficient is multiplied by the total length of the non-anesthesia induction segment in the monitoring curve and then rounded up to obtain the real-time monitoring segment length; taking the current time as the starting point, in the non-anesthesia induction segment of the monitoring curve, a monitoring curve segment of the real-time monitoring segment length is obtained in the reverse direction of the time sequence as the real-time monitoring segment;

[0012] The monitoring curve other than the real-time monitoring segment in the monitoring curve is used as a historical monitoring segment.

[0013] Furthermore, the method for obtaining the warning coefficient includes:

[0014] In the real-time monitoring segment and the historical monitoring segment, according to the fluctuation stability of the bispectral index in the corresponding monitoring segment and the deviation from the preset EEG index range, respectively obtain the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment;

[0015] Obtaining a trend item in the monitoring curve based on an STL decomposition algorithm, intercepting a real-time monitoring trend segment in the trend item, and obtaining a trend stability index of the bispectral index in the real-time monitoring trend segment;

[0016] The early warning coefficient is obtained according to the trend stability index and the difference between the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment.

[0017] Furthermore, the calculation formula of the warning coefficient includes:

[0018] ;in, is the warning coefficient at the current moment; is the overall stability coefficient of the historical monitoring section; is the real-time stability coefficient of the real-time monitoring section; Brackets for Iverson; To monitor the trend stability index of the bispectral index in the trend segment in real time; is the standard normalization function.

[0019] Furthermore, the calculation formula of the real-time stability coefficient includes:

[0020] ;in, is the real-time stability coefficient of the real-time monitoring section; is the standard deviation of all bispectral indices in the real-time monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the real-time monitoring segment; For real-time monitoring The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range;

[0021] The calculation formula of the overall stability coefficient includes:

[0022] ;in, is the overall stability coefficient of the historical monitoring section; is the standard deviation of all bispectral indices in the historical monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the historical monitoring segment; This is the first The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range.

[0023] Furthermore, the calculation formula of the trend stability index is:

[0024] ;in, To monitor the trend stability index of the bispectral index in the trend segment in real time; To monitor the maximum value of all bispectral indices in the trend segment in real time; It is the minimum value of all bispectral indices in the real-time monitoring trend segment; It is the standard deviation of all bispectral indices in the real-time monitoring trend segment.

[0025] Furthermore, the method for providing an abnormal warning for the patient's anesthesia depth state includes:

[0026] During the real-time monitoring of the patient's bi-frequency EEG, the warning coefficient is obtained at each current moment. When the warning coefficient exceeds the preset threshold, the corresponding current moment is marked as an abnormal moment; when there are a continuous preset number of abnormal moments, an abnormal warning is issued.

[0027] Furthermore, the method for obtaining the anesthesia induction segment includes:

[0028] In the monitoring curve, the moment corresponding to the bispectral index reaching the preset expected value for the first time is taken as the dividing point, and the monitoring curve corresponding to the time period from the initial moment to the moment corresponding to the dividing point in the monitoring curve is taken as the anesthesia induction segment, and the anesthesia induction stage includes the dividing point.

[0029] Furthermore, the method for obtaining the anesthesia sensitivity coefficient includes:

[0030] In the anesthesia induction segment, the difference between the first and last bispectral indexes is divided by the corresponding time interval to obtain the patient's anesthesia sensitivity coefficient.

[0031] The present invention also proposes an auxiliary identification system for anesthesia depth, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of an auxiliary identification method for anesthesia depth when executing the computer program.

[0032] The present invention has the following beneficial effects:

[0033] The present invention first obtains the monitoring curve of the patient's bispectral index in real time, and further obtains the anesthesia induction segment in the monitoring curve, so as to obtain the patient's anesthesia sensitivity coefficient according to the change speed of the bispectral index, and the anesthesia sensitivity coefficient reflects the patient's metabolic ability of anesthetics, and determines the frequency of monitoring the depth of anesthesia of the patient according to the individual characteristics of the patient in the subsequent monitoring process; then at the current moment, the monitoring curve is divided into a real-time monitoring segment and a historical monitoring segment according to the patient's anesthesia sensitivity coefficient, and the real-time monitoring time domain is adjusted according to individual signs, so as to comprehensively analyze the change of anesthesia depth by integrating the real-time monitoring segment and the historical monitoring segment; according to the difference in the fluctuation stability state of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, the early warning coefficient of the bispectral index at the current moment is obtained; according to the early warning coefficient at the current moment, the patient's anesthesia depth state is abnormally warned. The present invention first obtains the anesthesia sensitivity coefficient reflecting the individual characteristics of the patient, and further determines the local anesthesia depth state change analysis time domain at the current moment according to the anesthesia sensitivity coefficient, thereby improving the accuracy of real-time monitoring and identification of the patient's anesthesia depth state. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1A flowchart of a method for auxiliary identification of anesthesia depth provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the auxiliary identification method and system of anesthesia depth proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0037] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0038] The specific scheme of the auxiliary identification method and system for anesthesia depth provided by the present invention is described in detail below with reference to the accompanying drawings.

[0039] See also Figure 1 , which shows a flow chart of an auxiliary identification method for anesthesia depth provided by an embodiment of the present invention, specifically comprising:

[0040] Step S1, obtaining a monitoring curve of the patient's bispectral index in real time, and obtaining anesthesia induction segments in the monitoring curve.

[0041] In one embodiment of the present invention, the bispectral index (BIS) value of the patient is collected in real time from the beginning of anesthesia preparation to the end of the operation through a bispectral index monitor. During the collection process, the sampling frequency of the monitor is set to 128Hz, that is, 128 BIS values ​​are obtained per second, and the collected BIS values ​​are used to construct a monitoring curve according to the collection time sequence. The BIS value range is 0-100. Generally, the higher the BIS value, the clearer the consciousness; BIS values ​​of 0-20 are explosive inhibition, and the patient may be in danger of life; 20-60 are deep sedation, and the patient is unlikely to wake up and has no response to external stimuli; 60-80 are moderate sedation, and there is a response to loud commands or shaking; 80-100 are awake and can respond normally.

[0042] It should be noted that during the collection process, the temperature in the operating room should be kept at , humidity maintained at , to ensure that the electrode sheets adhere firmly and to avoid myoelectric interference caused by chills due to low temperature; use 75% alcohol to clean and degrease the patient's skin, and stick the electrodes of the monitor to the designated patch positions during the drying process; the collection of bispectral index is an existing technology well known to those skilled in the art; in other embodiments of the present invention, the implementer may also set other sampling frequencies according to the specific implementation situation.

[0043] The anesthesia process mainly includes three stages: anesthesia induction, maintenance of anesthesia and anesthesia recovery. Anesthesia induction is the process of the patient's change from a conscious state to an unconscious state. Maintenance of anesthesia means that after the patient enters an appropriate depth of anesthesia, the anesthesiologist performs appropriate anesthetic injections based on the patient's consciousness state and vital signs to maintain a stable state of anesthesia. At the same time, the doctor can perform surgery on the patient during this stage. After the operation is completed, the patient will gradually regain consciousness.

[0044] Due to the different physical conditions of different patients, the time of losing consciousness is also different, that is, each patient has a different sensitivity to anesthetics; the process of losing consciousness and entering an anesthesia state usually occurs during the anesthesia induction stage. To facilitate the subsequent analysis of each patient's sensitivity to anesthetics, the embodiment of the present invention further obtains anesthesia induction segments in the monitoring curve.

[0045] Preferably, in one embodiment of the present invention, the method for obtaining anesthesia induction segmentation includes: in the monitoring curve, the corresponding moment of the bispectral index that first reaches the preset expected value is used as the demarcation point, and the monitoring curve corresponding to the time period from the initial moment to the moment corresponding to the demarcation point in the monitoring curve is used as the anesthesia induction segment, and the anesthesia induction stage includes the demarcation point. Among them, the preset expected value is set to 60. During the anesthesia process, when the patient's bispectral index drops to 60 for the first time, it is considered that the patient's anesthesia induction stage ends and enters the maintenance anesthesia stage. The corresponding moment of the bispectral index that first reaches the preset expected value is the demarcation point between the two stages; therefore, the historical monitoring curve segment of the monitoring curve from the start of monitoring to the moment corresponding to the demarcation point is the anesthesia induction segment, and the anesthesia induction segment includes the demarcation point.

[0046] In other embodiments of the present invention, the implementer may use other values ​​of the upper limit of the BIS value interval corresponding to deep anesthesia as preset expected values ​​according to actual conditions, or may combine the rate of change of the instantaneous slope of the bispectral index at each moment to evaluate whether the patient's bispectral index has entered deep anesthesia and maintained in a stable anesthetic state, thereby judging whether the patient has entered the maintenance anesthesia stage, so as to obtain the dividing point and obtain the anesthesia induction segmentation, which are all existing technologies well known to those skilled in the art and will not be elaborated here.

[0047] It should be noted that during the real-time monitoring process, since the monitoring time is too short and the patient is still in the anesthesia induction stage, the monitoring curve may only collect the bispectral index of the anesthesia induction stage, and it is impossible to determine the anesthesia sensitivity index and perform abnormal monitoring. Therefore, when the patient's anesthesia induction stage ends and enters the continuous anesthesia stage, the bispectral index at the current moment will be monitored for abnormalities.

[0048] Step S2, obtaining the patient's anesthesia sensitivity coefficient according to the change rate of the bispectral index in the anesthesia induction segment; and dividing the monitoring curve into a real-time monitoring segment and a historical monitoring segment according to the patient's anesthesia sensitivity coefficient.

[0049] During the anesthesia induction stage, the faster the patient enters the anesthesia maintenance stage, i.e., the deep anesthesia state, the stronger the patient's ability to metabolize anesthetics. In the subsequent anesthesia maintenance stage, more frequent attention should be paid to the depth of anesthesia to ensure the safety of the operation. Based on this, the embodiment of the present invention first obtains the patient's anesthesia sensitivity coefficient according to the change rate of the bispectral index in the anesthesia induction segment; and then determines the real-time monitoring and analysis sequence according to the patient's anesthesia sensitivity coefficient, so as to accurately and real-time pay attention to the patient's anesthesia depth state according to the individual characteristics of the patient.

[0050] Preferably, in one embodiment of the present invention, considering that the speed at which the patient enters deep anesthesia from wakefulness during the anesthesia induction stage reflects his or her sensitivity to anesthetics, the method for obtaining the anesthesia sensitivity coefficient includes: in the anesthesia induction segment, dividing the difference between the first and last bispectral indexes by the corresponding time interval to obtain the patient's anesthesia sensitivity coefficient.

[0051] The calculation formula of anesthesia sensitivity coefficient is:

[0052] ;

[0053] in, is the patient's anesthesia sensitivity coefficient; The first bispectral index in the anesthesia induction segment corresponds to the acquisition time; is the value of the first bispectral index in the anesthesia induction segment; The caudal bispectral index in the anesthesia induction segment corresponds to the acquisition time; Values ​​of the caudal bispectral index during the anesthesia induction segment.

[0054] It should be noted that since the BIS value corresponding to the state of awakeness to deep anesthesia decreases, the sign logic of the numerator and the denominator is opposite; the anesthesia sensitivity coefficient is calculated by the speed formula. The faster the speed of change, the greater the anesthesia sensitivity coefficient.

[0055] Since the monitoring curve of the bispectral index during anesthesia is a time series curve, the data is constantly updated. As time changes, the long historical time series will smooth out the changes in the patient's bispectral index in the local time domain corresponding to the current moment, and increase the amount of calculation for change analysis; and if only the adjacent fixed segments of the current moment are selected to analyze the changes in the bispectral index, the individual characteristics of the patient, that is, the sensitive reaction to anesthetics, will be ignored, which may make it impossible to fully reflect the changes in the depth of anesthesia, affecting the anesthesiologist's timely intervention. Therefore, the embodiment of the present invention divides the monitoring curve into a real-time monitoring segment and a historical monitoring segment according to the patient's anesthesia sensitivity coefficient, constructs a local neighborhood for real-time bispectral index change analysis through different anesthesia sensitivity coefficients of different patients, and adjusts the real-time monitoring range according to individual signs, so as to comprehensively analyze the changes in anesthesia depth by combining the real-time monitoring segment and the historical monitoring segment.

[0056] Preferably, in one embodiment of the present invention, the monitoring curve is divided into real-time monitoring segments and historical monitoring segments according to the patient's anesthesia sensitivity coefficient, including: negatively mapping the anesthesia sensitivity coefficient and normalizing it to obtain a cutoff coefficient; multiplying the cutoff coefficient by the total length of the non-anesthesia induced segment in the monitoring curve and rounding it up to obtain the length of the real-time monitoring segment; taking the current moment as the starting point, in the non-anesthesia induced segment of the monitoring curve, obtaining a monitoring curve segment of the real-time monitoring segment length in the reverse direction of the time sequence as a real-time monitoring segment; and taking the monitoring curve other than the real-time monitoring segment in the monitoring curve as a historical monitoring segment.

[0057] In the embodiment of the present invention, considering that it usually takes several minutes to more than ten minutes for a patient to enter a deep anesthesia state from awake, the corresponding anesthetic sensitivity coefficient is usually a decimal, so the negative correlation mapping is implemented and normalized by subtracting the anesthetic sensitivity coefficient from 1. The implementer may also adopt other negative correlation mapping methods; the higher the anesthetic sensitivity coefficient, the faster the anesthetic metabolism, and the faster the fluctuation of the bispectral index of the patient is likely to be, and the smaller the local time domain is, the more consistent with the individual characteristics of the patient's individual bispectral index analysis and the higher the accuracy.

[0058] In another embodiment of the present invention, the implementer may also obtain the length of the real-time monitoring segment by multiplying the negative correlation mapping of the anesthesia sensitivity coefficient by the preset unit length and rounding it up to obtain the real-time monitoring segment in the monitoring curve at the current moment, and the monitoring curve segments in the monitoring curve other than the real-time monitoring segment are historical monitoring segments; in the embodiment of the present invention, the preset unit length is 20s. In other embodiments of the present invention, the implementer may also set the preset unit length according to the specific implementation situation; if the anesthesia sensitivity coefficient is 0.3, the inverse negative correlation mapping is taken and multiplied by the preset unit length of 60s, then a 60s monitoring curve segment is intercepted from the monitoring curve with the current moment as the starting point in the reverse direction of the timing as the real-time monitoring segment, and the remaining monitoring curve segments are historical monitoring segments.

[0059] Step S3, according to the difference in the fluctuation stability of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, obtain the warning coefficient of the bispectral index at the current moment.

[0060] When the patient enters the maintenance anesthesia stage, in order to ensure smooth entry into the surgical state and ensure intraoperative safety, the anesthesiologist should focus on the patient's anesthesia depth state changes, that is, the fluctuation stability of the bispectral index. Considering that when the time to enter the maintenance anesthesia stage is too short, the fluctuation trend of the bispectral index in the historical monitoring segment will be more intense, and the bispectral index in the real-time monitoring segment may also have certain fluctuations due to the just entering the maintenance anesthesia stage; after a certain period of monitoring time, the fluctuation performance of the historical monitoring segment in the historical long time series may be relatively gentle, reflecting the overall change trend of the anesthetic state changes during the anesthesia process, while the real-time monitoring segment is more focused on the fluctuation trend of the anesthetic state changes in the current adjacent time domain. If the fluctuation trend in the real-time monitoring segment is more intense, the difference between the two changes is greater, and the real-time changes of the bispectral index need to be paid more attention. Based on this, the embodiment of the present invention obtains the warning coefficient of the bispectral index at the current moment according to the difference in the fluctuation stability of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment.

[0061] Preferably, in one embodiment of the present invention, considering that the fluctuation of data in the monitoring curve may contain different components such as seasonality and trend, and may also contain noise interference, the changing trend of the depth of anesthesia state may not be accurately obtained based on the raw data changes of the monitoring curve, and the time series decomposition algorithm (Seasonal and Trenddecomposition using Loess, STL) can decompose each component, so as to accurately obtain the trend fluctuation of the patient's bispectral index, that is, the depth of anesthesia state change; the greater the fluctuation difference of the anesthetic state change in the historical monitoring segment, that is, the real-time monitoring segment, and the more the fluctuation of the bispectral index in the real-time monitoring segment tends to deviate or be drastic, the more attention should be paid to the early warning; based on this, the method for obtaining the early warning coefficient includes:

[0062] In the real-time monitoring segment and the historical monitoring segment, the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment are obtained respectively according to the fluctuation stability of the bispectral index in the corresponding monitoring segment and the deviation from the preset EEG index range; the trend item in the monitoring curve is obtained based on the STL decomposition algorithm, the real-time monitoring trend segment in the trend item is intercepted, and the trend stability index of the bispectral index in the real-time monitoring trend segment is obtained; according to the trend stability index and the difference between the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment, the early warning coefficient is obtained.

[0063] In the embodiment of the present invention, the real-time stability coefficient and the overall stability coefficient both reflect whether the change of the bispectral index in the corresponding monitoring segment within the preset EEG index range is stable. The more stable, the larger the corresponding coefficient is, and the lower the necessity of early warning may be. The preset EEG index range is a limiting interval for maintaining the patient's depth of anesthesia. Exceeding the interval may lead to surgical risks. The trend item in the monitoring curve is obtained by the STL algorithm, and then the trend item segment in the corresponding real-time monitoring segment is intercepted from the trend item as the real-time monitoring trend segment, and then the trend stability index is obtained according to the changes in the real-time monitoring trend segment. The trend stability index reflects the stability of the fluctuation trend of the bispectral index in the real-time monitoring trend segment. The larger the value, the more stable the current trend is, and the greater the possibility of subsequent changes according to this trend is. Whether an early warning should be given is further evaluated in combination with the stability coefficient.

[0064] It should be noted that the preset EEG index range is 60 to 80, and the implementer may also set it to other ranges such as 40 to 60 according to the specific implementation situation; the STL algorithm is an existing technology well known to those skilled in the art and will not be described in detail here.

[0065] The calculation formula of the early warning coefficient includes:

[0066] ;

[0067] in, is the warning coefficient at the current moment; is the overall stability coefficient of the historical monitoring section; is the real-time stability coefficient of the real-time monitoring section; Brackets for Iverson; To monitor the trend stability index of the bispectral index in the trend segment in real time; is the standard normalization function.

[0068] In the calculation formula of the early warning coefficient, when entering the initial stage of maintaining anesthesia, the historical monitoring segment mainly includes the bispectral index of the anesthesia induction segment, which shows a downward trend, and the corresponding overall stability coefficient is low, while the real-time monitoring segment mainly includes the unstable fluctuation state when entering the maintenance anesthesia stage, which is a normal situation. However, its real-time stability coefficient is higher than the overall stability coefficient of the historical monitoring segment. Then the condition in Iverson brackets does not hold, and the early warning coefficient is 0. When entering the maintenance anesthesia stage and lasting for a period of time, the historical monitoring segment shows a relatively stable overall fluctuation state as the time series lengthens, and the corresponding overall stability coefficient becomes larger. If the real-time stability coefficient in the real-time monitoring segment is larger at this time, and when the real-time stability coefficient is greater than the overall stability coefficient, then the corresponding The condition in the Ferson bracket is still 0, and the need for early warning is lower. When the real-time stability coefficient is less than or equal to the overall stability coefficient, and the difference between the two is greater, it means that the fluctuation of the bispectral index in the real-time monitoring segment is more drastic relative to the overall fluctuation, and the corresponding early warning coefficient is larger; if the real-time stability coefficient in the real-time monitoring segment is smaller at this time, it means that the fluctuation of the bispectral index in the real-time monitoring segment is more drastic, then the need for early warning is greater, and the corresponding early warning coefficient is also greater; at the same time, combined with the trend stability index in the real-time monitoring trend segment, the higher the trend stability index, the higher the possibility of maintaining the current drastic fluctuation trend, and the higher the corresponding early warning coefficient; on the contrary, it means that the subsequent drastic fluctuation may slow down, and the early warning coefficient is reduced.

[0069] Preferably, in one embodiment of the present invention, considering that in the historical monitoring segment or the real-time monitoring segment, the standard deviation of the bispectral index in each monitoring segment reflects its change stability, and at the same time, the greater the difference between the bispectral index at each moment and the median of the preset EEG index range, the greater the possibility of its fluctuation and the more unstable the change; based on this, the calculation formula of the real-time stability coefficient and the calculation formula of the overall stability coefficient can be constructed;

[0070] The calculation formula of the real-time stability coefficient includes:

[0071] ;

[0072] in, is the real-time stability coefficient of the real-time monitoring section; is the standard deviation of all bispectral indices in the real-time monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the real-time monitoring segment; For real-time monitoring The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range;

[0073] The calculation formula of the overall stability factor includes:

[0074] ;

[0075] in, is the overall stability coefficient of the historical monitoring section; is the standard deviation of all bispectral indices in the historical monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the historical monitoring segment; This is the first The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range.

[0076] In the calculation formula of the real-time stability coefficient and the calculation formula of the overall stability coefficient, the corresponding logic is adjusted by performing the inverse operation of the standard deviation, so that the larger the standard deviation, the lower the corresponding change stability coefficient; the differences between the median values ​​of the relative range of the bispectral index at all times are averaged to obtain the average fluctuation stability level of the bispectral index in the corresponding monitoring segment, and then negatively correlated to the exponential function to adjust the logic, the larger the average difference, the lower the corresponding change stability coefficient; finally, the two are multiplied to comprehensively evaluate the change stability; in other embodiments of the present invention, other negative correlation mapping methods can also be used to adjust the logic.

[0077] Preferably, in one embodiment of the present invention, the calculation formula of the trend stability index is:

[0078] ;

[0079] in, To monitor the trend stability index of the bispectral index in the trend segment in real time; To monitor the maximum value of all bispectral indices in the trend segment in real time; It is the minimum value of all bispectral indices in the real-time monitoring trend segment; It is the standard deviation of all bispectral indices in the real-time monitoring trend segment.

[0080] In the calculation formula of trend stability index, if The smaller it is, the more stable the fluctuation of the real-time monitoring trend segment is, and the more likely it is to maintain the current trend; The smaller it is, the more stable the trend fluctuation in the real-time monitoring trend segment is, which makes the denominator larger; It reflects the stability of the fluctuation trend of the data in the real-time monitoring trend segment and whether the current trend change will be maintained in the future. The smaller the ratio is, the smaller the possibility of deviation from the trend is. Through negative correlation mapping and normalization adjustment logic, the trend stability index becomes larger.

[0081] Step S4, issuing an abnormal warning for the patient's anesthesia depth state according to the warning coefficient at the current moment.

[0082] Since the patient's anesthesia depth status is monitored in real time and the current moment is constantly updated, the patient's anesthesia depth status can be evaluated by obtaining the warning coefficient at each current moment to provide abnormal warning.

[0083] Preferably, in one embodiment of the present invention, the method for providing an abnormal early warning of the depth of anesthesia of a patient comprises:

[0084] During the real-time monitoring of the patient's bi-frequency EEG, the warning coefficient is obtained at each current moment. When the warning coefficient exceeds the preset threshold, the corresponding current moment is marked as an abnormal moment; when there are a preset number of consecutive abnormal moments, an abnormal warning is issued.

[0085] In an embodiment of the present invention, the preset threshold is set to 0.8, and the preset number is set to 10. In other embodiments of the present invention, the implementer may also choose other values; when 10 consecutive current moments are marked as abnormal moments, it means that the patient's anesthesia depth state is not stable enough. At this time, an abnormal warning is required to assist the anesthesiologist in making adjustments based on changes in the patient's real-time bispectral index to ensure the safety of the operation.

[0086] In summary, the embodiment of the present invention first obtains the monitoring curve of the patient's bispectral index in real time, and further obtains the anesthesia induction segment in the monitoring curve, so as to obtain the patient's anesthesia sensitivity coefficient according to the change speed of the bispectral index; then divides the monitoring curve into a real-time monitoring segment and a historical monitoring segment according to the patient's anesthesia sensitivity coefficient; according to the difference in the fluctuation stability state of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, obtain the early warning coefficient of the bispectral index at the current moment; and perform abnormal early warning on the patient's anesthesia depth state according to the early warning coefficient at the current moment. The present invention first obtains the anesthesia sensitivity coefficient that reflects the individual characteristics of the patient, and further determines the local anesthesia depth state change analysis time domain at the current moment according to the anesthesia sensitivity coefficient, thereby improving the accuracy of real-time monitoring and identification of the patient's anesthesia depth state.

[0087] The present invention also proposes an auxiliary identification system for anesthesia depth, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, steps of an auxiliary identification method for anesthesia depth are implemented.

[0088] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0089] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for auxiliary identification of anesthesia depth, characterized in that: The method comprises: Acquire a monitoring curve of the patient's bispectral index in real time, and acquire anesthesia induction segments in the monitoring curve; According to the change rate of the bispectral index in the anesthesia induction segment, the anesthesia sensitivity coefficient of the patient is obtained; according to the anesthesia sensitivity coefficient of the patient, the monitoring curve is divided into a real-time monitoring segment and a historical monitoring segment; According to the difference in the fluctuation stability of the bispectral index in the real-time monitoring segment and the historical monitoring segment, combined with the change trend of the bispectral index in the real-time monitoring segment, the early warning coefficient of the bispectral index at the current moment is obtained; Providing an abnormal warning of the patient's anesthesia depth state according to the warning coefficient at the current moment; The method for obtaining the early warning coefficient includes: In the real-time monitoring segment and the historical monitoring segment, according to the fluctuation stability of the bispectral index in the corresponding monitoring segment and the deviation from the preset EEG index range, respectively obtain the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment; Obtaining a trend item in the monitoring curve based on an STL decomposition algorithm, intercepting a real-time monitoring trend segment in the trend item, and obtaining a trend stability index of the bispectral index in the real-time monitoring trend segment; Obtaining an early warning coefficient according to the trend stability index and the difference between the real-time stability coefficient of the real-time monitoring segment and the overall stability coefficient in the historical monitoring segment; The calculation formula of the real-time stability coefficient includes: ;in, is the real-time stability coefficient of the real-time monitoring section; is the standard deviation of all bispectral indices in the real-time monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the real-time monitoring segment; For real-time monitoring The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range; The calculation formula of the overall stability coefficient includes: ;in, is the overall stability coefficient of the historical monitoring section; is the standard deviation of all bispectral indices in the historical monitoring segment; The natural constant The exponential function with base ; is the number of all bispectral indices in the historical monitoring segment; This is the first The amplitude of the bispectral index; The upper limit of the preset EEG index range; It is the lower limit of the preset EEG index range.

2. The auxiliary identification method of anesthesia depth according to claim 1, characterized in that: The monitoring curve is divided into a real-time monitoring segment and a historical monitoring segment according to the anesthesia sensitivity coefficient of the patient, including: The anesthesia sensitivity coefficient is negatively correlated and normalized to obtain an interception coefficient; The interception coefficient is multiplied by the total length of the non-anesthesia induction segment in the monitoring curve and then rounded up to obtain the real-time monitoring segment length; taking the current time as the starting point, in the non-anesthesia induction segment of the monitoring curve, a monitoring curve segment of the real-time monitoring segment length is obtained in the reverse direction of the time sequence as the real-time monitoring segment; The monitoring curve other than the real-time monitoring segment in the monitoring curve is used as a historical monitoring segment.

3. The auxiliary identification method of anesthesia depth according to claim 1, characterized in that: The calculation formula of the early warning coefficient includes: ;in, is the warning coefficient at the current moment; is the overall stability coefficient of the historical monitoring section; is the real-time stability coefficient of the real-time monitoring section; Brackets for Iverson; To monitor the trend stability index of the bispectral index in the trend segment in real time; is the standard normalization function.

4. The auxiliary identification method of anesthesia depth according to claim 1, characterized in that: The calculation formula of the trend stability index is: ;in, To monitor the trend stability index of the bispectral index in the trend segment in real time; To monitor the maximum value of all bispectral indices in the trend segment in real time; It is the minimum value of all bispectral indices in the real-time monitoring trend segment; It is the standard deviation of all bispectral indices in the real-time monitoring trend segment.

5. The auxiliary identification method of anesthesia depth according to claim 1, characterized in that: Methods for providing abnormal warning of the patient's depth of anesthesia include: During the real-time monitoring of the patient's bi-frequency EEG, the warning coefficient is obtained at each current moment. When the warning coefficient exceeds the preset threshold, the corresponding current moment is marked as an abnormal moment; when there are a continuous preset number of abnormal moments, an abnormal warning is issued.

6. The method for auxiliary identification of anesthesia depth according to claim 1, characterized in that: The method for obtaining the anesthesia induction segment comprises: In the monitoring curve, the moment corresponding to the bispectral index reaching the preset expected value for the first time is taken as the dividing point, and the monitoring curve corresponding to the time period from the initial moment to the moment corresponding to the dividing point in the monitoring curve is taken as the anesthesia induction segment, and the anesthesia induction stage includes the dividing point.

7. The auxiliary identification method of anesthesia depth according to claim 1, characterized in that: The method for obtaining the anesthesia sensitivity coefficient includes: In the anesthesia induction segment, the difference between the first and last bispectral indexes is divided by the corresponding time interval to obtain the patient's anesthesia sensitivity coefficient.

8. An auxiliary identification system for anesthesia depth, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of an auxiliary identification method for anesthesia depth as described in any one of claims 1 to 7 are implemented.

Images