High-precision anesthesia depth intelligent monitoring method and device

Abstract

This application discloses a high-precision intelligent monitoring method and device for anesthesia depth, relating to the field of anesthesia monitoring technology. The method includes acquiring the raw electroencephalogram (EEG) signal of the target subject; preprocessing the raw EEG signal to obtain a preprocessed EEG signal; identifying a first anesthesia spindle event in the preprocessed EEG signal to obtain a first feature vector; performing multi-dimensional pattern matching between the anesthetic agent type and concentration information corresponding to the first feature vector and a constructed database model to determine the multi-dimensional feature interval model with the highest matching degree; the database model is a model constructed based on feature vectors of different anesthesia stages of several anesthetic drugs; determining the corresponding anesthesia stage based on the multi-dimensional feature interval model with the highest matching degree and displaying it on an interactive interface. This application can achieve high-precision intelligent monitoring of anesthesia depth.

Description

Technical Field

[0001] This application relates to the field of anesthesia monitoring technology, and in particular to a high-precision intelligent monitoring method and device for anesthesia depth. Background Technology

[0002] In studies of anesthesia mechanisms, electroencephalography (EEG) is an important monitoring tool. However, there is currently a lack of convenient and objective EEG indicators to accurately describe changes in the depth of anesthesia induced by different anesthetics, which poses a challenge to related research. The main shortcomings are as follows: Limitations of monitoring indicators: Existing studies often rely on burst inhibition activity in EEG as an indicator of deep anesthesia, but this is mainly limited to the deep anesthesia period and cannot accurately identify the induction, light anesthesia, and recovery stages. The clinically used bispectral index (BIS) has limited applicability. Advantages compared to BIS: 1. Widely applicable to all general anesthetics (especially ketamine) and sedative-hypnotic drugs; 2. Accurately characterizes the onset of anesthesia; 3. Spindle wave and slow wave coupling parameters can predict the risk of postoperative cognitive impairment in elderly patients. The widely used righting reflex, however, only provides binary information on the onset and end of anesthesia and cannot reflect dynamic changes in depth during the anesthesia process.

[0003] High equipment costs: Some studies use multi-site EEG recording technology and complex algorithms to assess the depth of anesthesia, but this requires high-level EEG recording equipment, surgical techniques and advanced data analysis capabilities. The equipment cost often exceeds 300,000 yuan, which is unaffordable for most small and medium-sized laboratories.

[0004] Overall analysis lacks temporal resolution: Conventional EEG analysis methods, such as time-frequency plots, frequency band analysis, and power spectral density, mainly capture overall energy changes over a period of time, but lack sufficiently high temporal resolution to explain the continuous dynamic changes at different stages of anesthesia (such as from light anesthesia to deep anesthesia).

[0005] Therefore, there is a need for a low-cost, easy-to-operate monitoring method and device that can reflect changes in the depth of general anesthesia cycle with high temporal resolution. Summary of the Invention

[0006] The purpose of this application is to provide a high-precision intelligent monitoring method and device for anesthesia depth, which can realize high-precision intelligent monitoring of anesthesia depth.

[0007] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a high-precision intelligent monitoring method for anesthesia depth, specifically including the following steps: Obtain the raw electroencephalogram (EEG) signals of the target subject.

[0008] The original EEG signal is preprocessed to obtain the preprocessed EEG signal.

[0009] The preprocessed EEG signal is subjected to first anesthesia spindle event identification to obtain a first feature vector; the first feature vector includes: the average density of the first anesthesia spindle, the first amplitude, and the first duration.

[0010] Based on the first feature vector, the anesthetic type and concentration information corresponding to the first feature vector are matched with the constructed database model in a multi-dimensional pattern to determine the multi-dimensional feature interval model with the highest matching degree; the database model is a model constructed based on the feature vectors of different anesthetic stages of several anesthetic drugs.

[0011] The corresponding anesthesia stage is determined based on the multidimensional feature interval model with the highest matching degree and displayed on the interactive interface.

[0012] Optionally, the construction of the database model specifically includes the following steps: Acquire characteristic electroencephalogram (EEG) signal data of known subjects during different anesthetic stages; the anesthetic stages include: induction period, light anesthesia period, deep anesthesia period, and recovery period.

[0013] The characteristic EEG signal data of each anesthesia stage were denoised and baseline corrected to obtain the preprocessed characteristic EEG signal data of each anesthesia stage.

[0014] The preprocessed EEG signal data for each anesthesia stage are subjected to second anesthesia spindle event identification to obtain a second feature vector; the second feature vector includes: the average density of the second anesthesia spindle, the second amplitude, and the second duration.

[0015] Based on the second feature vector corresponding to the anesthesia stage when using different anesthetic drugs, a multidimensional feature interval model is established for each anesthesia stage of each anesthetic drug.

[0016] The database model is constructed based on all the multidimensional feature interval models.

[0017] Optionally, the original EEG signal is preprocessed to obtain a preprocessed EEG signal, specifically including the following steps: The original EEG signal is subjected to common-mode interference suppression to obtain an amplified EEG signal.

[0018] The amplified EEG signal is then converted from analog to digital to obtain a digitized EEG signal.

[0019] The digitized EEG signal is bandpass filtered, followed by denoising and baseline correction to obtain the preprocessed EEG signal.

[0020] Optionally, the preprocessed EEG signal is subjected to first anesthesia spindle wave event recognition to obtain a first feature vector, specifically including the following steps: The preprocessed EEG signal was subjected to Hilbert transform to obtain the analytical signal.

[0021] The signal envelope is calculated based on the analytic signal.

[0022] Calculate the standard deviation and mean of the signal envelope within the current time window of a first preset length.

[0023] The dynamic detection threshold is set based on the standard deviation and the mean.

[0024] A second preset length sliding time window is set. Within the sliding time window, when the duration of the signal envelope continuously exceeding the dynamic detection threshold reaches a preset duration threshold, the continuous segment of the preprocessed EEG signal corresponding to the duration is identified as a potential first anesthesia spindle wave event.

[0025] The first anesthesia spindle event is determined based on the frequency of the potential first anesthesia spindle, and the first feature vector is obtained.

[0026] Optionally, the formula for calculating the signal envelope is as follows: ; in, For the signal envelope; The preprocessed EEG signal; The result is the Hilbert transform of the preprocessed EEG signal.

[0027] Optionally, the expression for the dynamic detection threshold is as follows: ; in, This is a set dynamic detection threshold; The mean value of the signal envelope within the current time window of the first preset length; The standard deviation of the signal envelope within the current time window of the first preset length; k This is an empirical coefficient.

[0028] Optionally, the multi-dimensional pattern matching process is based on the following logic: Match degree = sim ( V_real-time , V_model ); in, V_real-time =[density, amplitude, duration]; V_modelIt is the feature vector of a certain anesthetic drug in the database model at a certain stage; sim ( ) is a similarity function.

[0029] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the high-precision intelligent monitoring method for anesthesia depth as described above.

[0030] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the high-precision intelligent monitoring method for anesthesia depth described above.

[0031] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the high-precision intelligent monitoring method for anesthesia depth described above.

[0032] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a high-precision intelligent monitoring method and device for anesthesia depth. By acquiring the raw EEG signals of the target subject, it provides raw physiological data support for the assessment of anesthesia status. By preprocessing the raw EEG signals, it solves the problem of inaccurate feature extraction caused by noise and artifacts in the raw signals, and realizes the purification and standardization of effective EEG signals. By identifying the first anesthesia spindle wave event and extracting the first feature vector containing average density, amplitude and duration from the preprocessed signals, it solves the problem that the anesthesia status is difficult to directly quantify from the raw signals, and realizes the digital extraction and quantitative description of key anesthesia-related physiological features. By performing multi-dimensional pattern matching between the first feature vector, the anesthetic agent type and concentration information corresponding to the first feature vector and a database model constructed based on different anesthesia stages of multiple anesthetic drugs, it overcomes the limitation of traditional indicators such as burst suppression that can only identify the deep anesthesia stage, and realizes dynamic and accurate determination of the entire cycle of anesthesia induction, light anesthesia, deep anesthesia and recovery. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1This is an application environment diagram of a high-precision intelligent monitoring method for anesthesia depth in one embodiment of this application; Figure 2 A flowchart illustrating a high-precision intelligent monitoring method for anesthesia depth provided in an embodiment of this application; Figure 3 A schematic diagram of a single-channel electroencephalogram (EEG) recording device used in a high-precision intelligent monitoring method for anesthesia depth provided in an embodiment of this application; Figure 4 A schematic diagram of the human-computer interaction interface corresponding to a high-precision intelligent monitoring method for anesthesia depth provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0035] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0036] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] The high-precision intelligent monitoring method for anesthesia depth provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send the acquired raw EEG signal of the target object to server 104. Server 104 receives the raw EEG signal of the target object and preprocesses it to obtain a preprocessed EEG signal. It then performs first anesthesia spindle wave event recognition on the preprocessed EEG signal to obtain a first feature vector. The first feature vector includes: the average density, first amplitude, and first duration of the first anesthesia spindle wave. Based on the first feature vector, the anesthetic agent type and concentration information corresponding to the first feature vector are matched with a constructed database model in a multi-dimensional pattern to determine the multi-dimensional feature interval model with the highest matching degree. The database model is a model constructed based on feature vectors of different anesthesia stages of several anesthetic drugs. The corresponding anesthesia stage is determined based on the multi-dimensional feature interval model with the highest matching degree. Server 104 can feed back the corresponding anesthesia stage to terminal 102 and display it on the interactive interface. Furthermore, in some embodiments, the high-precision intelligent anesthesia depth monitoring method can also be implemented by either server 104 or terminal 102. For example, terminal 102 can directly process the raw EEG signal of the target subject to obtain the corresponding anesthesia stage, or server 104 can obtain the raw EEG signal of the target subject from the data storage system and process the raw EEG signal to obtain the corresponding anesthesia stage.

[0038] The terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, and tablets. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0039] In one exemplary embodiment, such as Figure 2 As shown, a high-precision intelligent monitoring method for anesthesia depth is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the following steps are taken: S201 to S205.

[0040] S201: Acquire the raw EEG signals of the target subject.

[0041] S202: The original EEG signal is preprocessed to obtain the preprocessed EEG signal.

[0042] S203: Perform first anesthesia spindle event identification on the preprocessed EEG signal to obtain a first feature vector; the first feature vector includes: the average density of the first anesthesia spindle, the first amplitude, and the first duration.

[0043] S204: Based on the first feature vector, the anesthetic type and concentration information corresponding to the first feature vector are matched with the constructed database model in a multi-dimensional pattern to determine the multi-dimensional feature interval model with the highest matching degree; the database model is a model constructed based on the feature vectors of different anesthetic stages of several anesthetic drugs.

[0044] S205: Determine the corresponding anesthesia stage based on the multidimensional feature interval model with the highest matching degree, and display it on the interactive interface.

[0045] By implementing steps S201 to S205 above, this application provides raw physiological data support for anesthetic state assessment by acquiring the raw EEG signals of the target subject; by preprocessing the raw EEG signals, it solves the problem of inaccurate feature extraction caused by noise and artifacts in the raw signals, and realizes the purification and standardization of effective EEG signals; by identifying the first anesthetic spindle wave event and extracting the first feature vector containing average density, amplitude and duration from the preprocessed signals, it solves the problem that the anesthetic state is difficult to directly quantify from the raw signals, and realizes the digital extraction and quantitative description of key physiological features related to anesthesia; by performing multi-dimensional pattern matching between the first feature vector, the anesthetic agent type and concentration information corresponding to the first feature vector and a database model constructed based on different anesthetic stages of multiple anesthetic drugs, it overcomes the limitation that traditional indicators such as burst suppression can only identify the deep anesthesia stage, and realizes dynamic and accurate determination of the entire cycle of anesthesia induction, light anesthesia, deep anesthesia and awakening.

[0046] In another exemplary embodiment of this application, in order to accurately preprocess the original EEG signal to obtain a preprocessed EEG signal, the above S202 is replaced by the following S2021~S2023: S2021: Suppress common-mode interference on the original EEG signal to obtain an amplified EEG signal.

[0047] S2022: Perform analog-to-digital conversion on the amplified EEG signal to obtain a digitized EEG signal.

[0048] S2023: The digitized EEG signal is subjected to bandpass filtering, followed by denoising and baseline correction to obtain the preprocessed EEG signal.

[0049] In another exemplary embodiment of this application, in order to accurately identify the first anesthesia spindle wave event from the preprocessed EEG signal and obtain the first feature vector, the above S203 is replaced by the following S2031~S2036: S2031: Perform Hilbert transform on the preprocessed EEG signal to obtain the analytical signal.

[0050] S2032: Calculate the signal envelope based on the analyzed signal.

[0051] S2033: Calculate the standard deviation and mean of the signal envelope within the current time window of a first preset length (e.g., the first 30 seconds).

[0052] S2034: Set a dynamic detection threshold based on the standard deviation and the mean.

[0053] S2035: Set a second preset sliding time window (window length 1ms). Within the sliding time window, when the signal envelope continuously exceeds the dynamic detection threshold for a duration... When the preset duration threshold (0.5 seconds ≤ Δt ≤ 3.0 seconds) is reached, the continuous segments of preprocessed EEG signals corresponding to the duration are identified as potential first anesthesia spindle wave events.

[0054] S2036: Determine the first anesthesia spindle event based on the frequency of the potential first anesthesia spindle, and obtain the first feature vector.

[0055] It should be noted that determining the first anesthesia spindle event based on the frequency of the potential first anesthesia spindle specifically includes: performing a fast Fourier transform on the preprocessed EEG signal within the potential first anesthesia spindle event, calculating its power spectral density, and taking the frequency corresponding to the peak value of the power spectral density in the 7-15Hz frequency band as the dominant frequency; if the dominant frequency is within the preset physiological frequency band of the anesthesia spindle, then the segment is formally identified as the first anesthesia spindle event; wherein, the frequency of the anesthesia spindle is the core physiological basis for defining and identifying the anesthesia spindle.

[0056] The average density of the first anesthesia spindle wave is the number of times the first anesthesia spindle wave event occurs per minute; the first duration is the length of time from the start to the end of the first anesthesia spindle wave event; and the first amplitude is the root mean square value of the signal envelope within the first anesthesia spindle wave event.

[0057] As an optional implementation, the construction of the database model specifically includes the following steps: S301: Acquire characteristic EEG signal data of known subjects when using different anesthetic drugs for the corresponding anesthesia stages; the anesthesia stages include: induction period, light anesthesia period, deep anesthesia period and recovery period.

[0058] S302: Denoise and baseline correction are performed on the characteristic EEG signal data of each anesthesia stage to obtain preprocessed characteristic EEG signal data of each anesthesia stage.

[0059] S303: Perform second anesthesia spindle event identification on the characteristic EEG signal data of each anesthesia stage after preprocessing to obtain a second feature vector; the second feature vector includes: the average density of the second anesthesia spindle, the second amplitude, and the second duration.

[0060] S304: Based on the second feature vector of the corresponding anesthesia stage when using different anesthetic drugs, a multidimensional feature interval model is established for each anesthesia stage of each anesthetic drug.

[0061] S305: Construct a database model based on all multidimensional feature interval models.

[0062] It should be noted that a second anesthesia spindle event identification is performed on the characteristic EEG signal data of each anesthesia stage after preprocessing to obtain a second feature vector. The specific processing flow of this step is the same as that of the first anesthesia spindle event identification to obtain the first feature vector, and will not be described again here.

[0063] In one example, to implement the high-precision intelligent monitoring method for anesthesia depth described above, the following device is used: 1. Single-channel electroencephalography (EEG) recording device: Used to collect cortical electroencephalogram (EEG) signals from organisms. Compared to complex multi-channel devices, it greatly simplifies hardware requirements and reduces costs.

[0064] 2. Signal processing module: connected to the EEG recording device, used to perform bandpass filtering processing of the acquired raw EEG signal from 0.1 to 60 Hz.

[0065] The single-channel electroencephalography (EEG) recording device used in this application (such as...) Figure 3 The image shows an improved data acquisition device, whose specific components and connection relationship with the signal processing module are as follows: Acquisition Electrode: Contains two miniature stainless steel cranial screws, serving as a recording electrode and a reference electrode, respectively. The recording electrode includes a miniature cranial screw with its tip designed to contact the dura mater or a biomimetic dura mater layer; the electrode body has standard threads for engagement and fixation with a nut pre-embedded in the cranial interface. This application does not include the step of implanting the electrode into a living organism, but only protects the electrode itself and its connection to the signal processing module.

[0066] Connection cable: A miniature flexible shielded cable is used, with one end connecting to the skull screw via a gold-plated pin, and the other end connecting to the signal processing module via a standard interface. The cable is designed to be detachable for easy electrical connection to the recording system during signal acquisition.

[0067] Preamplifier and Analog-to-Digital Converter: The connecting cable transmits the acquired weak electrical signal to an integrated bioelectrical signal amplifier. This amplifier uses a differential input structure to suppress common-mode interference (such as 50Hz power frequency noise). The amplifier output is connected to a high-precision analog-to-digital converter to convert the analog signal into a digital signal. The sampling rate is set to 1000Hz to meet the time resolution requirements of subsequent spindle wave detection.

[0068] Bandpass filtering: The digital EEG signal is subjected to bandpass filtering of 0.1 to 60 Hz to remove out-of-band noise.

[0069] System Integration: All the above components (except for the electrode components used for fixing) are integrated into a portable, electromagnetically shielded metal shell to form a complete and independent EEG signal acquisition front end, which communicates with the host computer (the computer running the data analysis module) via a USB interface.

[0070] 3. Data Analysis Module: The core innovation, containing a self-developed MATLAB script algorithm. This algorithm is used to detect and quantify "anesthesia spindle waves" in filtered EEG signals, extracting their dynamic characteristic parameters such as frequency, amplitude, and duration. The data analysis module performs the following specific steps to achieve automatic detection and parameter extraction of spindle waves: (1) Signal preprocessing: The EEG signal filtered by the signal processing module is further denoised and baseline corrected to eliminate slow drift and residual noise.

[0071] (2) Spindle wave detection: Signal envelope extraction: First, the preprocessed EEG signal... Perform a Hilbert transform to obtain the analytic signal, and then calculate the signal envelope. The calculation formula is as follows: ; in, For the signal envelope; The preprocessed EEG signal; Preprocessed EEG signals The Hilbert transform. (Source: The Hilbert transform is a standard method in signal processing used to extract the envelope of oscillating signals).

[0072] Dynamic threshold setting: Set a dynamic detection threshold. This threshold is based on the standard deviation of the signal envelope within the current time window (e.g., the first 30 seconds). and mean To calculate: ; in, This is a set dynamic detection threshold; The mean value of the signal envelope within the current time window of the first preset length; The standard deviation of the signal envelope within the current time window of the first preset length. k The coefficient is an empirical value, set to 1.5 based on the team's research findings.

[0073] Event recognition: Within a sliding time window (window length 1ms), when the signal envelope... Continuously exceeding the dynamic threshold Duration Satisfying 0.5 seconds ≤ If the EEG signal segment within this time period is ≤3.0 seconds, it is identified as a potential “anesthesia spindle wave” event.

[0074] (3) Feature parameter extraction: For each identified spindle wave event, the following feature parameters are calculated: Duration: The duration calculated during the event recognition step. , which represents the length of time from the start to the end of a spindle wave.

[0075] Amplitude: Extracts the signal envelope within the event. Root mean square value: ; in, Amplitude The amplitude of the first anesthesia spindle wave; The duration of the first anesthesia spindle event, characterizing the length of time from the start to the end of the spindle wave; This is the moment when the first anesthesia spindle wave event begins; This is the moment when the first anesthesia spindle wave event ends; This represents the signal envelope within the first anesthesia spindle wave event. (Source: Root mean square amplitude is defined by standard mathematical definition).

[0076] Frequency: The raw EEG signal within the event. A Fast Fourier Transform (FFT) was performed to calculate the power spectral density. The frequency corresponding to the peak value of the power spectral density in the 7-15Hz band was taken as the dominant frequency of the spindle wave. Simultaneously, the total power within this band was recorded for subsequent analysis. This frequency is the core physiological basis for defining and identifying the "anesthesia spindle wave."

[0077] 4. Anesthesia Depth Calculation Module: Based on the spindle wave feature parameters extracted by the data analysis module, this module compares them with a pre-constructed database. This database contains spindle wave feature threshold models corresponding to different anesthetic stages (such as isoflurane, propofol, ketamine, etc.) during induction, light anesthesia, deep anesthesia, and recovery. Based on this, this application calculates and displays the current anesthetic stage of the organism in real time. The threshold model database can be constructed based on data from existing published literature, simulated electroencephalograms of isolated brain slices, or clinical data collected through non-invasive methods; this application does not limit its specific source. The core of the anesthesia depth calculation module lies in a pre-constructed threshold model database based on multi-dimensional spindle wave features. Its construction and real-time calculation process are as follows: (1) Database construction (offline stage): First, through a series of preliminary experiments (based on standardized ex vivo brain tissue slices or publicly available animal EEG databases), characteristic EEG data were collected for different anesthetic drugs at corresponding anesthesia stages (induction, light anesthesia, deep anesthesia, and recovery). For each stage, the data analysis module described above was used to extract features such as spindle wave density (number of occurrences per minute), amplitude (root mean square value), and duration. Based on these experimental data, a multidimensional feature interval model was established for each anesthetic drug and each anesthesia stage. For example, based on the team's research findings, the following simplified model can be preliminarily constructed.

[0078] etomidate: Induction period: extremely high density (>30 times / minute) and extremely high amplitude (>80μV).

[0079] During light anesthesia: density and amplitude remain at high levels.

[0080] During deep anesthesia: density decreases significantly (<10 times / minute), but amplitude remains high.

[0081] Isoflurane: During deep anesthesia: density, amplitude, and duration all increase, reaching their peak.

[0082] Recovery phase: Density, amplitude, and duration all continue to increase until recovery.

[0083] Propofol: During light anesthesia: the density reaches its peak (>25 times / minute).

[0084] During deep anesthesia: density decreases significantly (<10 times / minute), while amplitude remains at a high level.

[0085] (2) Real-time calculation (online stage): During the actual monitoring process, the system calculates the average density, amplitude, and duration of spindle waves in the current time window (such as 30 seconds) in real time. Then, these three features are used as a feature vector and input into the inference module. The module performs multi-dimensional pattern matching between the real-time feature vector and the preset database model. The matching process can follow the following logic: Matching degree = sim ( V_real-time , V_model ); Among them, V_real-time =[density, amplitude, duration]; V_model is the feature vector of a certain anesthetic at a certain stage in the database model; sim ( ) is a similarity function.

[0086] The system will select the anesthetic stage corresponding to the model with the highest matching degree as the current output and display it on the interactive interface.

[0087] (3) Drug information association: Input the type and concentration information of the anesthetic used, and the system will only call the threshold model corresponding to this drug for comparison, thereby improving the accuracy and specificity of the inference.

[0088] 5. Human-machine interaction interface (as shown in Figure 4 ):Used to display the real-time EEG waveform, the analyzed spindle wave features, and the inferred anesthetic stage, and can receive the anesthetic drug information input by the researcher.

[0089] Figure 4 shows the human-machine interaction interface, which is used to display the real-time EEG waveform, the analyzed spindle wave features, and the inferred anesthetic stage, and can receive the anesthetic drug information input by the researcher. The determination result of the anesthetic stage and the time information are displayed in the upper right area of the interface. The lower right area of the interface is the real-time display area of the spindle wave waveform. The "Anesthetic Drugs" function button at the bottom of the interface can realize the linkage control with the drug administration process. The overall interface can synchronously display the electroencephalogram feature waveform, the determination result of the anesthetic stage, and the conventional vital sign parameters, realizing the intuitive and multi-dimensional monitoring of the anesthetic process.

[0090] In summary, this application uses electrodes implanted in the skull to acquire single-channel EEG signals, which are first transmitted to a signal processing module for bandpass filtering. The filtered signal then enters a data analysis module, running a self-developed MATLAB script. This script independently identifies the "anesthesia spindle wave" as a key EEG event and automatically calculates its frequency, amplitude, and duration. Next, an anesthesia depth estimation module compares and performs logical judgments with a built-in database of spindle wave characteristics of different anesthetics at different stages, constructed based on extensive prior animal experiments. This allows for real-time and accurate estimation of the anesthesia stage (induction, light anesthesia, deep anesthesia, or recovery) of the organism. Finally, all results (waveforms, parameters, anesthesia stage) are visualized for researchers through a human-computer interface, providing precise guidance for experimental procedures (e.g., timely administration of anesthetic drugs).

[0091] This application achieves the following beneficial effects by using "anesthesia spindle waves" as independent EEG markers and employing algorithms to monitor and analyze their dynamic characteristics (frequency, amplitude, and duration) in real time: 1. Achieved precise monitoring of the entire anesthesia cycle: Overcame the limitation of traditional indicators (such as burst suppression) in the background technology that can only identify the deep anesthesia period, and for the first time achieved dynamic, precise and visual monitoring of the entire cycle of anesthesia induction, light anesthesia, deep anesthesia and recovery.

[0092] 2. Significantly reduced equipment costs and lower barriers to entry: By adopting single-channel EEG hardware and self-developed algorithms, the expensive multi-channel equipment and complex analysis techniques are replaced, keeping the equipment cost at around 23,000 yuan, which is only 10% of the cost of similar international products. This makes it affordable for small and medium-sized laboratories and solves the high cost problem in the background technology.

[0093] 3. Provides objective indicators with high temporal resolution: Overcomes the shortcomings of traditional macroscopic methods such as EEG band analysis, by tracking the real-time changes of spindle waves as independent events, it objectively quantifies the dynamic evolution of anesthesia depth with higher temporal resolution, providing a more accurate tool for the study of anesthesia mechanisms.

[0094] It should be noted that this application does not limit the specific types of anesthetic drugs. Therefore, this application can identify characteristic EEG activity through algorithms, and can monitor the degree of loss of consciousness from light to deep under propofol, etomidate, ketamine, sevoflurane and isoflurane anesthesia in real time. It can also be used to continuously monitor the sedation depth of dexmedetomidine in the intensive care unit to avoid oversedation.

[0095] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 5As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data from the database model and data generated during the intelligent monitoring process. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a high-precision intelligent monitoring method for anesthesia depth.

[0096] Those skilled in the art will understand that Figure 5 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0097] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0098] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0099] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of the relevant data are carried out in compliance with the relevant data protection laws and policies of the country where the location is located, and with the authorization granted by the owner of the corresponding device.

[0100] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0101] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A high-precision intelligent monitoring method for anesthesia depth, characterized in that, The high-precision intelligent monitoring method for anesthesia depth includes: Acquire the raw electroencephalogram (EEG) signals of the target subject; The original EEG signal is preprocessed to obtain the preprocessed EEG signal; The preprocessed EEG signal is subjected to first anesthesia spindle event identification to obtain a first feature vector; the first feature vector includes: the average density of the first anesthesia spindle, the first amplitude, and the first duration. Based on the first feature vector, the anesthetic agent type and concentration information corresponding to the first feature vector are matched with the constructed database model in a multi-dimensional pattern to determine the multi-dimensional feature interval model with the highest matching degree; the database model is a model constructed based on the feature vectors of different anesthetic stages of several anesthetic drugs. The corresponding anesthesia stage is determined based on the multidimensional feature interval model with the highest matching degree and displayed on the interactive interface.

2. The high-precision intelligent monitoring method for anesthesia depth according to claim 1, characterized in that, The construction of the database model specifically includes: Acquire characteristic electroencephalogram (EEG) signal data of known subjects during different anesthetic stages; the anesthetic stages include: induction phase, light anesthesia phase, deep anesthesia phase, and recovery phase; Denoising and baseline correction were performed on the characteristic EEG signal data of each anesthesia stage to obtain preprocessed characteristic EEG signal data of each anesthesia stage. The preprocessed EEG signal data of each anesthesia stage are subjected to second anesthesia spindle event identification to obtain a second feature vector; the second feature vector includes: the average density of the second anesthesia spindle, the second amplitude, and the second duration. Based on the second feature vector of the corresponding anesthesia stage when using different anesthetic drugs, a multidimensional feature interval model is established for each anesthesia stage of each anesthetic drug; The database model is constructed based on all the multidimensional feature interval models.

3. The high-precision intelligent monitoring method for anesthesia depth according to claim 1, characterized in that, The raw EEG signal is preprocessed to obtain a preprocessed EEG signal, specifically including: The original EEG signal was subjected to common-mode interference suppression to obtain an amplified EEG signal; The amplified EEG signal is then converted from analog to digital to obtain a digitized EEG signal. The digitized EEG signal is bandpass filtered, followed by denoising and baseline correction to obtain the preprocessed EEG signal.

4. The high-precision intelligent monitoring method for anesthesia depth according to claim 1, characterized in that, The preprocessed EEG signal is subjected to first anesthesia spindle wave event recognition to obtain a first feature vector, specifically including: The preprocessed EEG signal was subjected to Hilbert transform to obtain the analytical signal; Calculate the signal envelope based on the analyzed signal; Calculate the standard deviation and mean of the signal envelope within the current time window of a first preset length; A dynamic detection threshold is set based on the standard deviation and the mean; A second preset length sliding time window is set. Within the sliding time window, when the duration of the signal envelope continuously exceeding the dynamic detection threshold reaches a preset duration threshold, the continuous segment of the preprocessed EEG signal corresponding to the duration is identified as a potential first anesthesia spindle wave event. The first anesthesia spindle wave event is determined based on the frequency of the potential first anesthesia spindle wave, and the first feature vector is obtained.

5. The high-precision intelligent monitoring method for anesthesia depth according to claim 4, characterized in that, The formula for calculating the signal envelope is: ； in, For the signal envelope; The preprocessed EEG signal; The result is the Hilbert transform of the preprocessed EEG signal.

6. The high-precision intelligent monitoring method for anesthesia depth according to claim 4, characterized in that, The expression for the dynamic detection threshold is: ； in, This is a set dynamic detection threshold; The mean value of the signal envelope within the current time window of the first preset length; The standard deviation of the signal envelope within the current time window of the first preset length; k This is an empirical coefficient.

7. The high-precision intelligent monitoring method for anesthesia depth according to claim 1, characterized in that, The multi-dimensional pattern matching process is based on the following logic: Match degree = sim ( V_real-time , V_model ); in, V_real-time =[density, amplitude, duration]; V_model It is the feature vector of a certain anesthetic drug in the database model at a certain stage; sim ( ) is a similarity function.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the high-precision intelligent monitoring method for anesthesia depth according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the high-precision intelligent monitoring method for anesthesia depth as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the high-precision intelligent monitoring method for anesthesia depth as described in any one of claims 1-7.

Images