Anesthetic state assessment method and system based on EEG data analysis

Abstract

The present application relates to an anesthetic state assessment method and system based on EEG data analysis, which belongs to the technical field of EEG data analysis. The method includes: determining an anesthesia assessment strategy based on the patient's attribute information, the attribute information includes age, medical history and surgical category, and the anesthesia assessment strategy is one or more of human-computer dialogue assessment, periodic instruction assessment, handheld object falling assessment and pressing device assessment; in the process of anesthetizing the patient, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score; in the process of anesthetizing the patient, the patient's EEG signal is collected; a mapping relationship between the consciousness score and the EEG signal is established to obtain a consciousness assessment model, and the consciousness assessment model is used to assess whether the patient has entered an anesthetized state. The present application has the effect of improving the accuracy of assessing the anesthetic state.

Description

Technical Field

[0001] The present application relates to the technical field of electroencephalogram (EEG) data analysis, and in particular to an anesthetic state assessment method and system based on EEG data analysis. Background Art

[0002] At present, in clinical research on brain science, how to apply EEG data to the assessment of anesthetic status and thus assist in anesthesia work is an especially important research topic.

[0003] The traditional way of applying EEG data to the evaluation of anesthesia status is: during the anesthesia process, the patient's EEG signals are collected, and the spectral characteristics, coherence and other indicators of the EEG signals are integrated to obtain a comprehensive value as the bispectral index, and then the patient's anesthesia status is evaluated by the bispectral index. The bispectral index is also called the BIS value. The BIS value is a monitoring indicator used to evaluate the depth of anesthesia, and its value range is 0 to 100. The larger the BIS value, the more awake the patient is; the smaller the BIS value, the more severe the inhibition of the cerebral cortex and the deeper the patient's anesthesia.

[0004] However, due to individual differences in age, resistance, etc., different patients may have different BIS values ​​obtained based on EEG signals even if the same anesthetic is used. The same patient may also have different degrees of resistance to different anesthetics, so the same BIS value is difficult to reflect the same anesthesia depth. Therefore, the current method of evaluating anesthesia status through EEG data has the problem of inaccurate evaluation. Summary of the invention

[0005] In order to solve the problem of low accuracy in evaluating anesthesia status based on EEG data, the present application provides an anesthesia status evaluation method and system based on EEG data analysis.

[0006] In a first aspect of the present application, a method for evaluating anesthesia state based on EEG data analysis is provided. The method comprises:

[0007] Determine an anesthesia assessment strategy according to the patient's attribute information, wherein the attribute information includes age, medical history, and surgery type, and the anesthesia assessment strategy is one or more of a human-machine dialogue assessment, a periodic instruction assessment, a handheld object falling assessment, and a pressing device assessment;

[0008] During the process of anesthetizing the patient, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, where the consciousness score is used to reflect the anesthesia depth of the patient;

[0009] During the process of anesthetizing the patient, collecting the patient's electroencephalogram signal;

[0010] A mapping relationship between the consciousness score and the electroencephalogram signal is established to obtain a consciousness assessment model, and the consciousness assessment model is used to assess whether the patient enters an anesthetized state.

[0011] By adopting the above technical solution, first, an anesthesia assessment strategy corresponding to the patient is selected, and during the anesthesia process of the patient, the selected anesthesia assessment strategy is used to evaluate the patient's anesthesia depth, thereby obtaining a consciousness score. At the same time, during the anesthesia process of the patient, the patient's EEG signal is also monitored in real time, and then a mapping relationship between the consciousness score and the EEG signal is established, and iterative training is performed to obtain a consciousness assessment model, so that after the EEG signal is obtained, the consciousness score corresponding to the EEG signal and the anesthesia depth corresponding to the consciousness score can be matched by the consciousness assessment model. It can be seen from this that the present application no longer uses EEG signals alone as the basis for evaluating whether the patient has entered an anesthesia state, but rather judges the patient's anesthesia depth (i.e., obtains a consciousness score) based on whether the patient can respond during anesthesia, and evaluates the patient's anesthesia depth together with the EEG signal, thereby ensuring the accuracy of evaluating whether the patient has entered an anesthesia state.

[0012] In a possible implementation, when the human-computer dialogue assessment is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including:

[0013] Determine the timing of the assessment;

[0014] When the patient is anesthetized to the evaluation time point, outputting a dialogue instruction;

[0015] collecting actual audio emitted by the patient;

[0016] A consciousness score is obtained according to the similarity between the actual audio and the reference audio, wherein the reference audio is the audio corresponding to the dialogue instruction.

[0017] In a possible implementation, when the periodic instruction evaluation is determined as the anesthesia assessment strategy, the step of using the anesthesia assessment strategy to assess the patient to obtain a consciousness score includes:

[0018] Determine the timing of the assessment;

[0019] When the patient is anesthetized to the evaluation time point, outputting an instruction;

[0020] collecting actual response behaviors performed by the patient;

[0021] An awareness score is obtained according to the similarity between the actual response behavior and the reference response behavior, wherein the reference response behavior is a behavior corresponding to the instruction.

[0022] In a possible implementation, when the handheld object falling assessment is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including:

[0023] Determine the timing of the assessment;

[0024] When the patient is anesthetized to the evaluation time point, driving the handheld object to fall;

[0025] collecting the actual time the patient holds the handheld object;

[0026] An awareness score is obtained according to the difference between the actual time and the reference time.

[0027] In a possible implementation, when the compression device evaluation determines that the anesthesia evaluation strategy is used, the anesthesia evaluation strategy is used to evaluate the patient to obtain a consciousness score, including:

[0028] Determine the timing of the assessment;

[0029] When the patient is anesthetized to the evaluation time point, outputting a compression instruction;

[0030] Collecting the patient's actual pressing time on the pressing device and a pressing force variation trend during the actual pressing time;

[0031] The consciousness score is obtained according to the changing trend of the pressing force during the actual pressing time.

[0032] In a possible implementation, obtaining the consciousness score according to the pressing force change trend during the actual pressing time includes:

[0033] When it is determined that the pressing force of the patient during the actual pressing time is in a decreasing trend, a preset initial score is obtained;

[0034] Calculate the change in the decreasing pressure intensity to get the additional score;

[0035] The sum of the initial score and the additional score is calculated and determined as the awareness score.

[0036] By adopting the above-mentioned technical scheme, the present application provides four evaluation methods for evaluating the depth of anesthesia of patients during anesthesia. Therefore, in actual use, the evaluation method corresponding to the patient can be matched according to the patient's attribute information, thereby ensuring the accuracy of the evaluation results and further ensuring the accuracy of the calculated consciousness score.

[0037] In a possible implementation, the step of establishing a mapping relationship between the consciousness score and the EEG signal to obtain a consciousness assessment model includes:

[0038] The consciousness score calculated and the EEG signal collected at the same evaluation time point are used as a pair of associated data, and the consciousness score and the EEG signal in the pair of associated data have a mapping relationship;

[0039] The associated data is input into a training model for iterative training to obtain a consciousness assessment model.

[0040] By adopting the above technical solution, the associated data is used for model training to obtain a consciousness assessment model, so that after the EEG signal is obtained, the consciousness assessment model can be matched to obtain the consciousness score corresponding to the EEG signal and the anesthesia depth corresponding to the consciousness score. Therefore, this application no longer uses the EEG signal as the sole basis for evaluating whether the patient has entered an anesthetized state, but judges the patient's anesthesia depth based on whether the patient can respond during the anesthesia process (that is, obtains the consciousness score), and adds the strength value of the EEG signal to evaluate the patient's anesthesia depth, thereby ensuring the accuracy of evaluating whether the patient has entered an anesthetized state.

[0041] In a possible implementation, determining an anesthesia assessment strategy according to the patient's attribute information includes:

[0042] Retrieving historical anesthesia status assessment data of the patient, wherein the historical anesthesia status assessment data includes a historical anesthesia assessment strategy and a historical age, historical medical history, and historical surgery category corresponding to the historical anesthesia assessment strategy;

[0043] Calculating a total similarity value between the historical anesthesia state assessment data and the attribute information;

[0044] Determine whether the total similarity value is greater than the similarity threshold;

[0045] If yes, determining the historical anesthesia assessment strategy as the anesthesia assessment strategy;

[0046] If not, the anesthesia assessment strategy is obtained according to the attribute information matching.

[0047] In a possible implementation manner: the calculating the total similarity value of the historical anesthesia state assessment data and the attribute information includes:

[0048] Obtaining a first similarity value according to the historical age and the age;

[0049] obtaining a second similarity value based on the historical medical history and the medical history;

[0050] Obtaining a third similarity value according to the historical surgery category and the surgery category;

[0051] A total similarity value is obtained according to the first similarity value, the second similarity value and the third similarity value.

[0052] By adopting the above technical solution, we first determine whether to select the historical anesthesia assessment strategy as the current anesthesia assessment strategy based on the patient's historical anesthesia status assessment data. On the one hand, this can reduce the time for selecting the anesthesia assessment strategy, and on the other hand, it can reduce the assessment error caused by the patient's discomfort caused by changing the anesthesia assessment strategy, thereby ensuring the accuracy of the calculated consciousness score.

[0053] In a second aspect of the present application, a system for evaluating anesthesia status based on EEG data analysis is provided. The system comprises:

[0054] A data determination module, configured to determine an anesthesia assessment strategy according to the patient's attribute information, wherein the attribute information includes age, medical history, and surgery type, and the anesthesia assessment strategy is one or more of a human-machine dialogue assessment, a periodic instruction assessment, a handheld object falling assessment, and a pressing device assessment;

[0055] a data calculation module, used for evaluating the patient by using the anesthesia evaluation strategy to obtain a consciousness score during anesthesia of the patient, wherein the consciousness score is used to reflect the anesthesia depth of the patient;

[0056] A data acquisition module, used for collecting the patient's electroencephalogram signal during anesthesia of the patient;

[0057] A data generation module is used to establish a mapping relationship between the consciousness score and the electroencephalogram signal to obtain a consciousness assessment model, and the consciousness assessment model is used to assess whether the patient enters an anesthetized state.

[0058] In summary, the present application includes the following beneficial technical effects:

[0059] First, select an anesthesia assessment strategy corresponding to the patient, and in the process of anesthetizing the patient, use the selected anesthesia assessment strategy to evaluate the patient's anesthesia depth, so as to obtain a consciousness score. At the same time, the patient's EEG signal is also monitored in real time during the anesthesia process, and then the consciousness score calculated at the same evaluation time point and the intensity value corresponding to the collected EEG signal are used as a pair of associated data. The associated data is used for model training to obtain a consciousness assessment model, so that after the EEG signal is obtained, the consciousness score corresponding to the EEG signal and the depth of anesthesia corresponding to the consciousness score can be matched by the consciousness assessment model. It can be seen from this that the present application no longer uses EEG signals as the sole basis for evaluating whether a patient has entered an anesthetized state, but judges the patient's anesthesia depth (i.e., obtains a consciousness score) based on whether the patient can respond during anesthesia, and evaluates the patient's anesthesia depth together with the intensity value of the EEG signal, thereby ensuring the accuracy of evaluating whether the patient has entered an anesthetized state. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1It is a flow chart of the anesthesia state assessment method based on EEG data analysis according to an embodiment of the present application.

[0061] Figure 2 It is a block diagram of an anesthesia state assessment system based on EEG data analysis according to an embodiment of the present application.

[0062] Explanation of the accompanying drawings: 1. Data determination module; 2. Data calculation module; 3. Data acquisition module; 4. Data generation module. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0064] In order to improve the accuracy of the patient's anesthetic state assessment, this application, based on the premise of EEG data analysis, adds different evaluation methods to evaluate the patient's anesthetic depth during the anesthesia stage, and then establishes a mapping relationship between EEG data and evaluation data, so as to facilitate the subsequent use of the mapping relationship between the two to obtain accurate anesthetic state assessment results. The anesthetic state of this application refers to a medical state in which the patient enters a state of no pain, unconsciousness or coma, which is commonly used in surgery, pain management and medical procedures.

[0065] Figure 1 A flow chart of an anesthetic state assessment method based on EEG data analysis according to an embodiment of the present application is shown, and the main process of the method is described as follows.

[0066] Step S1: Determine an anesthesia assessment strategy based on the patient's attribute information.

[0067] Attribute information includes patient-related identity information and medical information such as age, medical history, surgery type, and gender.

[0068] The anesthesia assessment strategy is one or more of the human-computer dialogue evaluation, periodic instruction evaluation, handheld object falling evaluation and pressing device evaluation. Among them, the human-computer dialogue evaluation refers to: pre-setting an intelligent dialogue AI robot, the AI ​​robot outputs intelligent AI questions, and receives the patient's answer voice to the intelligent AI questions for evaluation. Periodic instruction evaluation refers to: the intelligent dialogue AI robot periodically outputs instructions, and detects the patient's response behavior to the instructions for evaluation. Handheld object falling evaluation refers to: evaluating by detecting the time it takes the patient to grab the fallen handheld object. Pressing device evaluation refers to: detecting the patient's pressing operation on the pressing device for evaluation, such as pressing time and pressing force and other pressing parameters as the data source for evaluation.

[0069] The process for determining anesthesia assessment strategy is:

[0070] First, the patient's historical anesthesia status assessment data is retrieved. The historical anesthesia status assessment data is the anesthesia assessment strategy used each time in the patient's past anesthesia experience and the age, medical history, and surgical category corresponding to each anesthesia assessment strategy. In this application, the anesthesia assessment strategy in the historical anesthesia status assessment data is referred to as the historical anesthesia assessment strategy, and the age, medical history, and surgical category corresponding to each historical anesthesia assessment strategy are referred to as the historical age, historical medical history, and historical surgical category, respectively.

[0071] Then, the first similarity value is calculated by the historical age and the age in the attribute information. For example, the value obtained by subtracting the age in the attribute information from the historical age can be used as the first similarity value. Alternatively, the difference between the age in the attribute information and the historical age can be calculated first, and then the ratio of the difference to the historical age can be used as the first similarity value. In this example, the ratio of the two is used as the first similarity value. The second similarity value is also calculated by the historical medical history and the medical history in the attribute information. Specifically, the second similarity value is based on the degree of correlation between the historical medical history and the medical history in the attribute information. For example, the similarity of coronary heart disease, heart disease, and non-atherosclerotic coronary heart disease that also require heart bypass surgery can be close to 1, that is, it can be considered that the above three medical histories have a very high degree of correlation. In actual application, the degree of correlation between different medical histories is obtained through big data training in advance. Therefore, after obtaining the historical medical history and the medical history in the attribute information, the degree of correlation between the two can be obtained through the big data model trained in advance, that is, the second similarity value is obtained. At the same time, a third similarity value is calculated through the historical surgical categories and the surgical categories in the attribute information. Specifically, the third similarity value is taken as the degree of correlation between the historical surgical categories and the surgical categories in the attribute information. Similarly, calculating the third similarity value is similar to calculating the second similarity value. Both are achieved by training the big data model in advance, and then obtaining the degree of correlation between the historical surgical categories and the surgical categories in the attribute information based on the big data model, that is, obtaining the third similarity value.

[0072] Based on the obtained first similarity value, second similarity value and third similarity value, the sum of the three is calculated to obtain the total similarity value, and it is determined whether the total similarity value is greater than the similarity threshold. If so, the historical anesthesia assessment strategy is determined to be the current anesthesia assessment strategy; if not, the anesthesia assessment strategy is matched according to the attribute information.

[0073] It should be noted that, by separately setting score thresholds for the first similarity value, the second similarity value, and the third similarity value, when any similarity value exceeds the corresponding score threshold, the historical anesthesia assessment strategy is determined to be the current anesthesia assessment strategy; otherwise, it is necessary to match the anesthesia assessment strategy based on the attribute information.

[0074] The above process of matching anesthesia assessment strategies according to attribute information is as follows:

[0075] If the patient's age reaches the age threshold, the human-computer dialogue assessment is selected;

[0076] If the patient's surgical category and medical history do not affect the patient's response to instructions, select the periodic instruction assessment;

[0077] If the patient's hands can move normally, choose the hand-held object drop assessment;

[0078] If the patient's surgical type and medical history do not affect the patient's ability to perform compressions, select the compression device assessment;

[0079] Finally, the selected evaluation items are output as the anesthesia assessment strategy. Therefore, the anesthesia assessment strategy may be one of the human-computer dialogue evaluation, periodic instruction evaluation, handheld object falling evaluation and compression device evaluation, or a combination of multiple items.

[0080] Step S2: During the anesthesia process of the patient, an anesthesia assessment strategy is used to assess the patient to obtain a consciousness score.

[0081] For each patient whose anesthesia status is to be evaluated, after determining the corresponding anesthesia evaluation strategy according to the above step S1, the corresponding evaluation device is controlled in turn to perform the anesthesia status evaluation. The evaluation device, such as the above-mentioned intelligent dialogue AI robot, handheld object, pressing device, etc., is an intelligent device for executing the anesthesia evaluation strategy.

[0082] Before using the test device to evaluate the patient's anesthetic state, the evaluation time point is first determined. The evaluation time point is the moment when the patient is evaluated. There are usually multiple evaluation time points. This is because the patient needs to go through the normal state and the sedation state before entering the anesthetic state. Therefore, multiple evaluation time points are set to better monitor the patient's anesthesia process. In a specific example, the evaluation time point can be periodic, such as 1 second / time, 2 seconds / time, etc., or it can be non-periodic, because with the injection of anesthetic drugs, the time for the patient to enter the anesthetic state from the sedation state may be relatively short, so the evaluation time point can be multiple moments that are successively accelerated, such as gradually decreasing from 10 seconds / time to 1 second / time.

[0083] Based on the set evaluation time points, the patient's anesthetic status is evaluated during the anesthesia process.

[0084] When the human-computer dialogue assessment is determined as an anesthesia assessment strategy, the intelligent dialogue AI robot outputs dialogue instructions at the evaluation time point, and collects the actual audio emitted by the patient after outputting the dialogue instructions, and then calls the reference audio corresponding to the dialogue instructions, that is, the reference audio corresponding to the dialogue instructions is set in advance. Then, the similarity between the actual audio and the reference audio is calculated. The similarity between the two can be calculated based on signal processing methods, such as cross-correlation, peak signal-to-noise ratio (PSNR), and phase correlation (PC). Cross-correlation measures the similarity between two audio signals by calculating the cross-correlation coefficient between the two audio signals, PSNR evaluates the similarity between the two audio signals by comparing the ratio between the peak signal and noise of the two audio signals, and PC measures the similarity between the two audio signals by calculating the phase correlation between the two audio signals. Of course, calculating the similarity between the actual audio and the reference audio can also be based on feature extraction methods, such as mel frequency cepstral coefficients (MFCCs), short-time Fourier transform (STFT), deep learning models such as convolutional neural networks (CNN) or recurrent neural networks (RNN), etc., to extract audio features and calculate the similarity between the actual audio and the reference audio, and use the calculated similarity value as the awareness score.

[0085] When the periodic instruction evaluation is determined as the anesthesia assessment strategy, the intelligent dialogue AI robot outputs the instruction at the evaluation time point, and after outputting the instruction, collects the actual response behavior of the patient through the camera, and then calls the reference response behavior corresponding to the instruction, that is, the reference response behavior corresponding to the instruction is set in advance. For example, if the instruction is the voice of raising the right hand, the reference response behavior is the behavior of raising the right hand. Then, the similarity between the actual response behavior and the reference response behavior is calculated. The similarity algorithm used can be a hash algorithm, such as mean hash, interpolation hash, and perceptual hash method, that is, the hash values ​​of the actual response behavior and the reference response behavior are calculated respectively, and then the Hamming distance is used to calculate the difference between the two pictures. Of course, the square graph algorithm can also be used to calculate the similarity value between the two, such as the grayscale histogram Q algorithm, the RGB histogram algorithm, etc. Finally, the calculated similarity value is used as the consciousness score.

[0086] When the handheld object falling assessment is determined as an anesthesia assessment strategy, a manual or intelligent device drives the handheld object to fall at the assessment time point, and collects the actual time when the patient grasps the handheld object. Specifically, a height sensor and a timer are set in the handheld object. After the handheld object is grasped, the time when it is grasped is uploaded, and this time is the actual time. Then call the reference time, calculate the difference between the actual time and the reference time, and then use the difference to match the corresponding consciousness score. Specifically, the present application sets in advance a consciousness score corresponding to the difference between the actual time and the reference time. The corresponding relationship between the difference and the consciousness score can be set in advance by experts in related fields, or it can be generated after training through a big data model. Therefore, after obtaining the difference between the actual time and the reference time, the consciousness score corresponding to the difference can be matched.

[0087] When the pressing device evaluation is determined as an anesthesia assessment strategy, the intelligent dialogue AI robot issues a pressing command. After hearing the pressing command, the patient presses the pressing device. The pressing device uploads the patient's actual pressing time and the pressing force change trend during the actual pressing time in real time, and obtains the consciousness score according to the pressing force change trend of the patient during the actual pressing time. In a specific example, the consciousness score is obtained according to the pressing force change trend of the patient during the actual pressing time: first determine whether the pressing force of the patient during the actual pressing time is a decreasing trend. If so, a preset initial score is obtained, and then the decreasing change amount is calculated to match the additional score, that is, the corresponding relationship between the decreasing change amount and the additional score is set in advance. Finally, the sum of the initial score and the additional score is calculated as the consciousness score. If not, the consciousness score is directly output as 0 points. In other examples, it is first determined whether the patient's pressing force during the actual pressing time is decreasing. If so, a preset initial score is obtained, and then the decreasing change is calculated to match the weight corresponding to the change, and the larger the change, the larger the weight. Finally, the product of the initial score and the weight is calculated to obtain the consciousness score. In actual use, the specific calculation method used is not limited in this application.

[0088] The above is the consciousness score calculated when a single item among the human-computer dialogue assessment, periodic instruction assessment, hand-held object falling assessment and pressing device assessment is used as an anesthesia assessment strategy. When multiple items among the human-computer dialogue assessment, periodic instruction assessment, hand-held object falling assessment and pressing device assessment are used as an anesthesia assessment strategy, the consciousness score corresponding to each item is still calculated separately according to the above calculation method, and then the average of the multiple consciousness scores is used as the consciousness score of the anesthesia assessment strategy.

[0089] Step S3: During the anesthesia of the patient, the patient's electroencephalogram (EEG) signal is collected.

[0090] Similarly, during the process of anesthetizing a patient, that is, during the process of evaluating the patient using an anesthesia assessment strategy, the EEG signal of the patient is collected in real time by a plurality of EEG electrodes provided on the EEG device worn on the patient's head, thereby obtaining the EEG signal of the patient during the anesthesia process. In order to facilitate subsequent analysis, the obtained EEG signal is converted to the required frequency band, thereby obtaining the intensity value of the EEG signal in the frequency band, such as converting to the θ EEG frequency band (4-8Hz) or converting to the β EEG frequency band (14-100Hz).

[0091] Step S4: Establish a mapping relationship between the consciousness score and the EEG signal to obtain a consciousness assessment model, which is used to assess whether the patient has entered an anesthetized state.

[0092] First, the consciousness score and the intensity value of the EEG signal at the same evaluation time point are retrieved, and the consciousness score and intensity value at the same evaluation time point are used as a piece of associated data, that is, a mapping relationship between the consciousness score and the intensity value of the EEG signal is established according to the evaluation time point. Since multiple evaluation time points are set, multiple pieces of associated data are obtained, and each piece of associated data is input into the training model, which is iteratively trained to obtain a consciousness assessment model.

[0093] Based on the obtained consciousness assessment model, it is convenient for subsequent real-time access to the electroencephalogram (EEG) machine, and the corresponding consciousness score and the anesthesia depth corresponding to the consciousness score are obtained based on the EEG signal matching uploaded by the EEG machine. That is, the consciousness score is used to determine whether the patient is still able to respond, thereby judging whether the patient has entered an anesthetized state.

[0094] In summary, the implementation principle of the anesthesia state assessment method based on EEG data analysis in the embodiment of the present application is as follows: first, select an anesthesia assessment strategy corresponding to the patient, and use the selected anesthesia assessment strategy to assess the patient's anesthesia depth during the anesthesia process, thereby obtaining a consciousness score. At the same time, the patient's EEG signal is also monitored in real time during the anesthesia process, and then the consciousness score calculated at the same evaluation time point and the intensity value corresponding to the collected EEG signal are used as a pair of associated data, and the associated data are used for model training to obtain a consciousness assessment model, so that after the EEG signal is obtained, the consciousness score corresponding to the EEG signal and the anesthesia depth corresponding to the consciousness score can be matched by the consciousness assessment model. It can be seen from this that the present application no longer uses EEG signals alone as the basis for assessing whether the patient has entered an anesthesia state, but rather judges the patient's anesthesia depth (i.e., obtains a consciousness score) based on whether the patient can respond during anesthesia, and then assesses the patient's anesthesia depth together with the intensity value of the EEG signal, thereby ensuring the accuracy of assessing whether the patient has entered an anesthesia state.

[0095] Figure 2 A block diagram of an anesthesia state assessment system based on EEG data analysis according to an embodiment of the present application is shown. The system includes a data determination module 1, a data calculation module 2, a data acquisition module 3 and a data generation module 4.

[0096] The data determination module 1 is used to determine the anesthesia assessment strategy based on the patient's attribute information, the attribute information including age, medical history and surgery type, and the anesthesia assessment strategy is one or more of human-computer dialogue assessment, periodic instruction assessment, handheld object falling assessment and compression device assessment.

[0097] The data calculation module 2 is used to evaluate the patient using an anesthesia assessment strategy to obtain a consciousness score during the anesthesia process of the patient, and the consciousness score is used to reflect the anesthesia depth of the patient.

[0098] The data acquisition module 3 is used to collect the patient's electroencephalogram signal during the anesthesia process.

[0099] The data generation module 4 is used to establish a mapping relationship between the consciousness score and the EEG signal to obtain a consciousness assessment model, and the consciousness assessment model is used to assess whether the patient has entered an anesthetized state.

[0100] The modules involved in the embodiments described in the present application may be implemented by software or by hardware. The modules described may also be set in a processor, for example, it may be described as: a processor includes a data determination module 1, a data calculation module 2, a data acquisition module 3 and a data generation module 4. The names of these modules do not constitute a limitation on the modules themselves in some cases, for example, the data determination module 1 may also be described as "a module for determining an anesthesia assessment strategy based on the patient's attribute information".

[0101] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0102] In order to better execute the procedure of the above method, the present application also provides an anesthesia status assessment device based on EEG data analysis, which includes a memory and a processor.

[0103] The memory may be used to store instructions, programs, codes, code sets or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function, and instructions for implementing the above-mentioned method for assessing anesthesia state based on EEG data analysis, etc.; the data storage area may store data involved in the above-mentioned method for assessing anesthesia state based on EEG data analysis, etc.

[0104] The processor may include one or more processing cores. The processor calls the data stored in the memory by running or executing the instructions, programs, code sets or instruction sets stored in the memory, performs various functions of the present application and processes data. The processor may be at least one of a special purpose integrated circuit, a digital signal processor, a digital signal processing device, a programmable logic device, a field programmable gate array, a central processing unit, a controller, a microcontroller and a microprocessor. It is understandable that for different devices, the electronic device used to implement the above-mentioned processor function can also be other, and the embodiments of the present application are not specifically limited.

[0105] The present application also provides a computer-readable storage medium, for example, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes. The computer-readable storage medium stores a computer program that can be loaded by a processor and execute the above-mentioned anesthesia state assessment method based on EEG data analysis.

[0106] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the aforementioned disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present application (but not limited to) by each other to form a technical solution.

Claims

1. An anesthetic state assessment device based on EEG data analysis, characterized in that: The device comprises a memory and a processor: The memory is used to store instructions, programs, codes, code sets or instruction sets; The processor is used to execute instructions, programs, codes, code sets or instruction sets in the memory to implement the following functions: Determine an anesthesia assessment strategy according to the patient's attribute information, wherein the attribute information includes age, medical history, and surgery type, and the anesthesia assessment strategy is one or more of a human-machine dialogue assessment, a periodic instruction assessment, a handheld object falling assessment, and a pressing device assessment; During the process of anesthetizing the patient, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, wherein the consciousness score is used to reflect the anesthesia depth of the patient; During the process of anesthetizing the patient, collecting the patient's electroencephalogram signal; A mapping relationship between the consciousness score and the EEG signal is established to obtain a consciousness assessment model, and a consciousness score corresponding to the EEG signal and an anesthetic depth corresponding to the consciousness score are obtained through matching the consciousness assessment model, and whether the patient can still respond verbally or with hand-held or pressing movements is determined through the consciousness score, so as to assess whether the patient has entered an anesthetized state; When the human-computer dialogue evaluation is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including: Determine the timing of the assessment; When the patient is anesthetized to the evaluation time point, outputting a dialogue instruction; collecting actual audio emitted by the patient; A consciousness score is obtained according to the similarity between the actual audio and the reference audio, wherein the reference audio is the audio corresponding to the dialogue instruction.

2. The anesthesia state assessment device based on EEG data analysis according to claim 1, characterized in that: When the periodic instruction evaluation is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including: Determine the timing of the assessment; When the patient is anesthetized to the evaluation time point, outputting an instruction; collecting actual response behaviors performed by the patient; An awareness score is obtained according to the similarity between the actual response behavior and the reference response behavior, wherein the reference response behavior is a behavior corresponding to the instruction.

3. The anesthesia state assessment device based on EEG data analysis according to claim 1, characterized in that: When the handheld object falling assessment is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including: Determine the timing of the assessment; When the patient is anesthetized to the evaluation time point, driving the handheld object to fall; collecting the actual time the patient holds the handheld object; An awareness score is obtained according to the difference between the actual time and the reference time.

4. The anesthesia state assessment device based on EEG data analysis according to claim 1, characterized in that: When the pressing device evaluation determines that the anesthesia evaluation strategy is used, the anesthesia evaluation strategy is used to evaluate the patient to obtain a consciousness score, including: Determine the timing of the assessment; When the patient is anesthetized to the evaluation time point, outputting a compression instruction; Collecting the actual pressing time of the patient on the pressing device and the changing trend of the pressing force during the actual pressing time; The consciousness score is obtained according to the changing trend of the pressing force during the actual pressing time.

5. The anesthesia state assessment device based on EEG data analysis according to claim 4, characterized in that: The step of obtaining the consciousness score according to the changing trend of the pressing force during the actual pressing time includes: When it is determined that the pressing force of the patient during the actual pressing time is in a decreasing trend, a preset initial score is obtained; Calculate the change in the decreasing pressure intensity to get the additional score; The sum of the initial score and the additional score is calculated and determined as the awareness score.

6. The anesthesia state assessment device based on EEG data analysis according to any one of claims 2 to 4, characterized in that: The step of establishing a mapping relationship between the consciousness score and the electroencephalogram signal to obtain a consciousness assessment model includes: The consciousness score calculated and the EEG signal collected at the same evaluation time point are used as a pair of associated data, and the consciousness score and the EEG signal in the pair of associated data have a mapping relationship; The associated data is input into a training model for iterative training to obtain a consciousness assessment model.

7. The anesthesia state assessment device based on EEG data analysis according to claim 1, characterized in that: Determining the anesthesia assessment strategy according to the patient's attribute information includes: Retrieving historical anesthesia status assessment data of the patient, wherein the historical anesthesia status assessment data includes a historical anesthesia assessment strategy and a historical age, historical medical history, and historical surgery category corresponding to the historical anesthesia assessment strategy; Calculating a total similarity value between the historical anesthesia state assessment data and the attribute information; Determine whether the total similarity value is greater than the similarity threshold; If yes, determining the historical anesthesia assessment strategy as the anesthesia assessment strategy; If not, the anesthesia assessment strategy is obtained according to the attribute information matching.

8. The anesthesia state assessment device based on EEG data analysis according to claim 7, characterized in that: The calculating the total similarity value of the historical anesthesia state assessment data and the attribute information comprises: Obtaining a first similarity value according to the historical age and the age; obtaining a second similarity value based on the historical medical history and the medical history; Obtaining a third similarity value according to the historical surgery category and the surgery category; A total similarity value is obtained according to the first similarity value, the second similarity value and the third similarity value.

9. An anesthetic state assessment system based on EEG data analysis, characterized in that: include: A data determination module (1) is used to determine an anesthesia assessment strategy based on the patient's attribute information, wherein the attribute information includes age, medical history and surgery type, and the anesthesia assessment strategy is one or more of a human-machine dialogue assessment, a periodic instruction assessment, a handheld object falling assessment and a pressing device assessment; A data calculation module (2) is used to evaluate the patient using the anesthesia evaluation strategy to obtain a consciousness score during anesthesia of the patient, wherein the consciousness score is used to reflect the anesthesia depth of the patient; A data acquisition module (3) is used to collect the patient's electroencephalogram signal during the anesthesia of the patient; A data generation module (4) is used to establish a mapping relationship between the consciousness score and the electroencephalogram signal to obtain a consciousness assessment model, and to obtain a consciousness score corresponding to the electroencephalogram signal and an anesthetic depth corresponding to the consciousness score through matching with the consciousness assessment model, and to determine whether the patient is still able to respond verbally or by hand-holding or pressing movements, so as to assess whether the patient has entered an anesthetic state. When the human-computer dialogue evaluation is determined as the anesthesia assessment strategy, the anesthesia assessment strategy is used to assess the patient to obtain a consciousness score, including: Determine the timing of the assessment; When the patient is anesthetized to the evaluation time point, outputting a dialogue instruction; collecting actual audio emitted by the patient; A consciousness score is obtained according to the similarity between the actual audio and the reference audio, wherein the reference audio is the audio corresponding to the dialogue instruction.

Images