Anesthesia depth monitoring method and system, readable storage medium and computer
By optimizing clinical EEG data and building neural network models, the problem that existing deep anesthesia monitoring methods cannot accurately control the dose of anesthetic drugs is solved, and accurate monitoring and control of the depth of anesthesia is achieved, and monitoring accuracy and reliability are improved. .
Patent Information
- Application Number
- CN202411008723.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-12-13
AI Technical Summary
The existing deep anesthesia monitoring methods cannot accurately control the dose of anesthetic, and are greatly affected by physical differences between different patients, making it difficult to accurately monitor and control during the anesthesia process.
By collecting clinical EEG data, performing data optimization and average processing, building a neural network model, using the EEG data matrix for deep learning processing, establishing anesthesia depth indicators and monitoring models, and optimizing the neural network model for in-depth monitoring of anesthesia.
It improves the accuracy and reliability of anesthesia depth monitoring, reduces environmental impact, enhances the accuracy of data processing, and realizes accurate monitoring and control of anesthesia depth.
Smart Images
Figure CN119112096A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of signal processing, and in particular to an anesthesia depth monitoring method, system, readable storage medium and computer. Background Art
[0002] With the improvement of people's living standards and the rapid development of medical technology, anesthesia is often the primary condition for clinical surgery. Usually, anesthesia is achieved by injecting drugs that block synaptic transmission into the patient's body, thereby putting the patient into an unconscious state. Therefore, anesthesia depth monitoring is an important factor in the success of the operation.
[0003] Currently, anesthesia depth monitoring is achieved by anesthesiologists processing the signals collected by a collection device installed on the patient's head based on the patient's relevant physiological signals and anesthesia experience. However, the monitoring equipment and the experience of anesthesiologists cannot completely overcome the impact of physical differences between different patients, resulting in the inability to accurately control the anesthetic dosage during anesthesia. Summary of the invention
[0004] Based on this, the object of the present invention is to provide an anesthesia depth monitoring method, system, readable storage medium and computer to at least solve the deficiencies in the above-mentioned technology.
[0005] The present invention provides a method for monitoring the depth of anesthesia, comprising: Collecting a number of clinical EEG data, and optimizing each of the clinical EEG data to obtain corresponding EEG optimized data; Performing corresponding averaging processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampling data; Performing segmented analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; Constructing a neural network model, and inputting the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; Performing depth monitoring on the data processing results and the sampling data to obtain corresponding depth monitoring results, and constructing corresponding anesthesia depth indicators based on each of the clinical EEG data; The neural network model is optimized according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and the anesthesia depth of the patient is monitored according to the anesthesia depth monitoring model.
[0006] Further, the steps of performing corresponding averaging processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampled data include: Performing data conversion on each of the EEG optimization data to obtain a corresponding bi-frequency EEG index; The dual-frequency EEG index is averaged using a mean absolute error algorithm to obtain a data sequence of each EEG optimization data, and the data sequence is resampled to obtain corresponding sampled data.
[0007] Further, the steps of performing segmented analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix include: Performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; The normalized sequence set is interpolated, and the time-frequency localized window function is optimized to obtain a window optimization function. The normalized sequence set is Fourier transformed using a Fourier transform optimization function constructed using the window optimization function. The window optimization function is moved along the time axis to obtain a corresponding EEG data matrix.
[0008] Furthermore, the expression of the window optimization function is: ; In the formula, represents the window function of time-frequency localization, represents the natural exponential function, Clinical EEG data The time domain transformation result is represents an imaginary unit; The expression of the Fourier transform is: ; In the formula, represents the length of the STFT sliding window, represents a time series, Represents the sampling frequency of clinical EEG data.
[0009] Furthermore, the steps of constructing a neural network model include: Acquire a preset deep learning model and a visual geometry model, and adjust model parameters of the deep learning model and the visual geometry model to obtain a corresponding deep learning optimization model and a visual geometry optimization model; Recombining the deep learning optimization model and the visual geometry optimization model to obtain a preliminary neural network model; The preliminary neural network model is validated using a cross-validation algorithm. If the model validation result meets the preset requirements, the acquired anesthesia monitoring data is input into the preliminary neural network model for model training to obtain a corresponding neural network model.
[0010] The present invention also provides an anesthesia depth monitoring system, comprising: A data optimization module, used for collecting a number of clinical EEG data, and optimizing each of the clinical EEG data to obtain corresponding EEG optimized data; An average processing module, used for performing corresponding average processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampled data; A segmentation analysis module, used for performing segmentation analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; A model building module is used to build a neural network model and input the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; A depth monitoring module, used for performing depth monitoring on the data processing result and the sampling data to obtain a corresponding depth monitoring result, and constructing a corresponding anesthesia depth index based on each of the clinical EEG data; The anesthesia monitoring module is used to optimize the neural network model according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and perform anesthesia depth monitoring on the patient according to the anesthesia depth monitoring model.
[0011] Furthermore, the average processing module includes: A data conversion unit, used for performing data conversion on each of the EEG optimization data to obtain a corresponding dual-frequency EEG index; The averaging processing unit is used to perform averaging processing on the dual-frequency EEG index using a mean absolute error algorithm to obtain a data sequence of each EEG optimization data, and resample the data sequence to obtain corresponding sampling data.
[0012] Furthermore, the segment analysis module includes: A normalization processing unit, used for performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; A Fourier transform unit is used to perform interpolation processing on the normalized sequence set, and optimize the time-frequency localized window function to obtain a window optimization function, perform Fourier transform on the normalized sequence set using the Fourier transform optimization function constructed by the window optimization function, and move the window optimization function along the time axis to obtain a corresponding EEG data matrix.
[0013] Furthermore, the model building module includes: A parameter adjustment unit, used to obtain a preset deep learning model and a visual geometry model, and adjust model parameters of the deep learning model and the visual geometry model to obtain a corresponding deep learning optimization model and a visual geometry optimization model; A model reorganization unit, used for reorganizing the deep learning optimization model and the visual geometry optimization model to obtain a preliminary neural network model; The model building unit is used to perform model verification on the preliminary neural network model using a cross-validation algorithm. If the model verification result meets the preset requirements, the acquired anesthesia monitoring data is input into the preliminary neural network model for model training to obtain a corresponding neural network model.
[0014] The present invention also provides a readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned anesthesia depth monitoring method is implemented.
[0015] The present invention also proposes a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned anesthesia depth monitoring method when executing the computer program.
[0016] The anesthesia depth monitoring method, system, readable storage medium and computer of the present invention collect a number of clinical EEG data, optimize the clinical EEG data and perform corresponding average processing, resample the obtained data sequence to obtain sampled data, and make the data more reliable through data optimization; convert the clinical EEG data into an EEG data matrix, and input the EEG data matrix into the constructed neural network model to obtain the corresponding data processing result, reduce the environmental impact of the data, and enhance the data processing accuracy; construct an anesthesia depth index based on the clinical EEG data, and use the anesthesia depth index and the depth monitoring result to optimize the neural network model to obtain an anesthesia depth monitoring model, and monitor the anesthesia depth of the patient according to the anesthesia depth monitoring model. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 is a flow chart of the anesthesia depth monitoring method in the first embodiment of the present invention; Figure 2 for Figure 1 Detailed flow chart of step S102; Figure 3 for Figure 1 Detailed flow chart of step S103; Figure 4 for Figure 1 Detailed flow chart of step S104; Figure 5 is a structural block diagram of an anesthesia depth monitoring system in a second embodiment of the present invention; Figure 6 FIG. 4 is a structural block diagram of a computer in a third embodiment of the present invention.
[0018] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION
[0019] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.
[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.
[0021] Embodiment 1 See also Figure 1 , which shows a method for monitoring the depth of anesthesia in a first embodiment of the present invention, and the method specifically includes steps S101 to S106: S101, collecting a number of clinical EEG data, and optimizing each of the clinical EEG data to obtain corresponding EEG optimized data; In the specific implementation, a number of clinical EEG data are collected, wherein the clinical EEG data are EEG data of known anesthesia results. The data are acquired in real time from the patient by the collection equipment and stored in the database. The data are collected before the patient's anesthesia surgery begins and are completed after the surgery until the patient's consciousness is restored. The collected relevant data are associated with the corresponding time. After the anesthesia surgery, the relevant vital sign data of the corresponding patient is recorded and optimized based on the relevant data to ensure the validity of the collected data.
[0022] S102, performing corresponding averaging processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampling data; For further information, see Figure 2 , the step S102 specifically includes steps S1021~S1022: S1021, performing data conversion on each of the EEG optimization data to obtain a corresponding bi-frequency EEG index; S1022, using a mean absolute error algorithm to average the dual-frequency EEG index to obtain a data sequence of each EEG optimization data, and resample the data sequence to obtain corresponding sampled data.
[0023] In the specific implementation, data conversion is performed on the various EEG optimization data obtained above, and they are converted into corresponding dual-frequency EEG indexes (i.e., BIS EEG indexes). The dual-frequency EEG indexes are averaged using the mean absolute error algorithm to obtain data sequences of various EEG optimization data. During the averaging process, an anesthesia level curve corresponding to the EEG optimization data is generated, and anesthesia level stages are defined, including a deep anesthesia stage, a moderate anesthesia stage, and a shallow anesthesia stage. The anesthesia level curve is matched with the three anesthesia level stages to obtain data sequences corresponding to various EEG optimization data. The data sequences are resampled using a resampling algorithm to obtain corresponding sampling data.
[0024] S103, performing segmented analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; For further information, see Figure 3 , the step S103 specifically includes steps S1031~S1032: S1031, performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; S1032, interpolating the normalized sequence set, optimizing the time-frequency localized window function to obtain a window optimization function, performing Fourier transform on the normalized sequence set using a Fourier transform optimization function constructed by the window optimization function, and moving the window optimization function along the time axis to obtain a corresponding EEG data matrix.
[0025] In the specific implementation, the above clinical EEG data are processed in a standardized manner, and the clinical EEG data are defined as , which have the same time interval (in this embodiment, the time interval is 5s) A set of EEG sequences is obtained, and the obtained set of EEG sequences is normalized to reduce the amplitude difference between EEG signals, thereby obtaining a normalized sequence set; Specifically, since during anesthesia, EEG signals may have different length outputs due to burst suppression during anesthesia, interpolation processing of the normalized sequence set obtained above can solve the problem of unbalanced data distribution, and the data resampling characteristics obtained by interpolation processing remain unchanged, so as to make the data smooth; Furthermore, the time-frequency localized window function in the conventional Fourier transform is obtained , for this window function Optimize and optimize the sequence length in the normalized sequence set to ensure the stability of the signal, and obtain the corresponding window optimization function: ; In the formula, represents the window function of time-frequency localization, represents the natural exponential function, Clinical EEG data The time domain transformation result is represents an imaginary unit; According to the window optimization function obtained above, the corresponding window width is selected, the conventional Fourier transform function is optimized, and the corresponding Fourier optimization function is obtained. The Fourier optimization function is used to perform Fourier transform on the model of each time period, and the window optimization function is moved along the time axis to convert the Fourier transform result into the corresponding EEG data matrix, where the horizontal quantity represents the time sampling point, and the vertical quantity represents the frequency of the sampling point corresponding to the signal. The calculation formula of Fourier change is: ; In the formula, represents the length of the STFT sliding window, represents a time series, Represents the sampling frequency of clinical EEG data.
[0026] S104, constructing a neural network model, and inputting the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; For further information, see Figure 4 , the step S104 specifically includes steps S1041 to S1043: S1041, obtaining a preset deep learning model and a visual geometry model, and adjusting model parameters of the deep learning model and the visual geometry model to obtain a corresponding deep learning optimization model and a visual geometry optimization model; S1042, reorganizing the deep learning optimization model and the visual geometry optimization model to obtain a preliminary neural network model; S1043, using a cross-validation algorithm to perform model verification on the preliminary neural network model. If the model verification result meets the preset requirements, the acquired anesthesia monitoring data is input into the preliminary neural network model for model training to obtain a corresponding neural network model.
[0027] In the specific implementation, a preset deep learning model and a visual geometry model are obtained. In this embodiment, the deep learning model uses the AlexNet model, and the visual geometry model uses the VGGNet model. The model parameters of the two models are adjusted, and a recognition classifier of EEG data is added to the classifiers of the two models. The output of the last output layer in the model is defined as four outputs, and the two convolutional network layers in the VGGNet model are deleted, and the corresponding activation function is modified from a linear rectification function to a hyperbolic tangent function, thereby obtaining the corresponding deep learning optimization model and visual geometry optimization model; Furthermore, the learners of the two models are combined to generate a new learner, and the new learner is used to construct a convolutional neural network model to obtain a preliminary neural network model. Professional anesthesiologists are used to perform anesthesia monitoring simulation on the several clinical EEG data obtained above to obtain corresponding anesthesia monitoring data. The anesthesia monitoring data is split into a training set and a validation set, wherein the ratio of the training set to the validation set is 4:1, and the cross-validation algorithm is used to perform model validation on the training set, validation set and preliminary neural network model. The model is trained with each training set. After each training is completed, the untrained training set is used to evaluate the trained model. The cycle is repeated three times in sequence, and the four results are averaged. The average processed value is compared with the fourth training result. If the fourth training result is greater than the average processed value, the validation set is input into the model for verification to output the corresponding neural network model.
[0028] Furthermore, the EEG data matrix is input into the neural network model obtained above for data processing to output a frequency spectrum corresponding to the EEG data matrix.
[0029] S105, performing depth monitoring on the data processing result and the sampling data to obtain a corresponding depth monitoring result, and constructing a corresponding anesthesia depth index based on each of the clinical EEG data; In the specific implementation, the above-mentioned sampling data and spectrum graph are deeply monitored, and the sampling data are matched with the spectrum graph to obtain the spectrum graph under the anesthesia level stage, and the consciousness level of the above-mentioned clinical EEG data is judged by a number of senior anesthesiologists to obtain the anesthesia depth index of each clinical EEG data. When the consciousness level score is between 0.8 and 1 (excluding the endpoints), it means that the anesthesia depth of the patient corresponding to the clinical EEG data is the shallow anesthesia stage; when the consciousness level score is between 0.5 and 0.8 (excluding the endpoints), it means that the anesthesia depth of the patient corresponding to the clinical EEG data is the moderate anesthesia stage, and when the consciousness level score is between 0 and 0.5 (excluding the endpoints), it means that the anesthesia depth of the patient corresponding to the clinical EEG data is the deep anesthesia stage.
[0030] S106, optimizing the neural network model according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and performing anesthesia depth monitoring on the patient according to the anesthesia depth monitoring model.
[0031] In the specific implementation, the anesthesia depth index and depth monitoring results obtained above are imported into the constructed neural network model so that the learner of the neural network model can be trained and learned, and the depth monitoring results, anesthesia depth index and neural network model are combined to form an anesthesia depth monitoring model. The anesthesia depth of the patient is monitored according to the anesthesia depth monitoring model. It can be understood that when the anesthesia data of a patient during an anesthesia operation is obtained, the anesthesia depth monitoring model is used to process the anesthesia data and predict the results, so as to obtain the depth monitoring result of the anesthesia data, and the patient's entire anesthesia process is effectively monitored according to the anesthesia monitoring result.
[0032] In summary, the anesthesia depth monitoring method in the above-mentioned embodiment of the present invention collects a number of clinical EEG data, optimizes the clinical EEG data and performs corresponding averaging processing, resamples the obtained data sequence to obtain sampled data, and makes the data more reliable through data optimization; converts the clinical EEG data into an EEG data matrix, and inputs the EEG data matrix into the constructed neural network model to obtain the corresponding data processing result, thereby reducing the environmental impact of the data and enhancing the data processing accuracy; constructs an anesthesia depth index based on the clinical EEG data, and uses the anesthesia depth index and the depth monitoring result to optimize the neural network model to obtain an anesthesia depth monitoring model, and performs anesthesia depth monitoring on the patient according to the anesthesia depth monitoring model.
[0033] Embodiment 2 Another aspect of the present invention is to provide an anesthesia depth monitoring system. Figure 5 , shown is an anesthesia depth monitoring system in a second embodiment of the present invention, the system comprising: The data optimization module 11 is used to collect a number of clinical EEG data and optimize the clinical EEG data to obtain corresponding EEG optimized data; An average processing module 12 is used to perform corresponding average processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resample the data sequence to obtain corresponding sampled data; Furthermore, the average processing module 12 includes: A data conversion unit, used for performing data conversion on each of the EEG optimization data to obtain a corresponding dual-frequency EEG index; The averaging processing unit is used to perform averaging processing on the dual-frequency EEG index using a mean absolute error algorithm to obtain a data sequence of each EEG optimization data, and resample the data sequence to obtain corresponding sampling data.
[0034] A segmentation analysis module 13 is used to perform segmentation analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; Furthermore, the segment analysis module 13 includes: A normalization processing unit, used for performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; A Fourier transform unit is used to perform interpolation processing on the normalized sequence set, and optimize the time-frequency localized window function to obtain a window optimization function, perform Fourier transform on the normalized sequence set using the Fourier transform optimization function constructed by the window optimization function, and move the window optimization function along the time axis to obtain a corresponding EEG data matrix.
[0035] A model building module 14 is used to build a neural network model and input the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; Furthermore, the model building module 14 includes: A parameter adjustment unit, used to obtain a preset deep learning model and a visual geometry model, and adjust model parameters of the deep learning model and the visual geometry model to obtain a corresponding deep learning optimization model and a visual geometry optimization model; A model reorganization unit, used for reorganizing the deep learning optimization model and the visual geometry optimization model to obtain a preliminary neural network model; The model building unit is used to perform model verification on the preliminary neural network model using a cross-validation algorithm. If the model verification result meets the preset requirements, the acquired anesthesia monitoring data is input into the preliminary neural network model for model training to obtain a corresponding neural network model.
[0036] A depth monitoring module 15, for performing depth monitoring on the data processing result and the sampling data to obtain a corresponding depth monitoring result, and constructing a corresponding anesthesia depth index based on each of the clinical EEG data; The anesthesia monitoring module 16 is used to optimize the neural network model according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and perform anesthesia depth monitoring on the patient according to the anesthesia depth monitoring model.
[0037] The functions or operation steps implemented when the above modules and units are executed are generally the same as those in the above method embodiments, and will not be repeated here.
[0038] The anesthesia depth monitoring system provided in the embodiment of the present invention has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the system embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.
[0039] Embodiment 3 The present invention also provides a computer, see Figure 6 , shown is a computer in the third embodiment of the present invention, including a memory 10, a processor 20, and a computer program 30 stored in the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, the above-mentioned anesthesia depth monitoring method is implemented.
[0040] The memory 10 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10 may be an internal storage unit of a computer, such as a hard disk of the computer. In other embodiments, the memory 10 may also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Further, the memory 10 may also include both an internal storage unit of the computer and an external storage device. The memory 10 may be used not only to store application software and various types of data installed in the computer, but also to temporarily store data that has been output or is to be output.
[0041] Among them, in some embodiments, the processor 20 can be an electronic control unit (Electronic Control Unit, abbreviated as ECU, also known as a vehicle computer), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor or other data processing chip, used to run the program code stored in the memory 10 or process data, such as executing access restriction programs, etc.
[0042] It should be pointed out that Figure 6 The structure shown does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.
[0043] The embodiment of the present invention further provides a readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the anesthesia depth monitoring method as described above is implemented.
[0044] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0045] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0046] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0047] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described 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.
[0048] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A method for monitoring the depth of anesthesia, It is characterized in that include: Collecting a number of clinical EEG data, and optimizing each of the clinical EEG data to obtain corresponding EEG optimized data; Performing corresponding averaging processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampling data; Performing segmented analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; Constructing a neural network model, and inputting the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; Performing depth monitoring on the data processing results and the sampling data to obtain corresponding depth monitoring results, and constructing corresponding anesthesia depth indicators based on each of the clinical EEG data; The neural network model is optimized according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and the anesthesia depth of the patient is monitored according to the anesthesia depth monitoring model.
2. The method for monitoring depth of anesthesia according to claim 1, It is characterized in that The steps of performing corresponding averaging processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampled data include: Performing data conversion on each of the EEG optimization data to obtain a corresponding bi-frequency EEG index; The dual-frequency EEG index is averaged using a mean absolute error algorithm to obtain a data sequence of each EEG optimization data, and the data sequence is resampled to obtain corresponding sampled data.
3. The method for monitoring depth of anesthesia according to claim 1, It is characterized in that The steps of performing segmented analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix include: Performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; The normalized sequence set is interpolated, and the time-frequency localized window function is optimized to obtain a window optimization function. The normalized sequence set is Fourier transformed using a Fourier transform optimization function constructed using the window optimization function. The window optimization function is moved along the time axis to obtain a corresponding EEG data matrix.
4. The method for monitoring depth of anesthesia according to claim 3, It is characterized in that The expression of the window optimization function is: ； In the formula, represents the window function of time-frequency localization, represents the natural exponential function, Clinical EEG data The time domain transformation result is: represents an imaginary unit; The expression of the Fourier transform is: ； In the formula, represents the length of the STFT sliding window, represents a time series, Represents the sampling frequency of clinical EEG data.
5. The method for monitoring depth of anesthesia according to claim 1, It is characterized in that The steps to build a neural network model include: Acquire a preset deep learning model and a visual geometry model, and adjust model parameters of the deep learning model and the visual geometry model to obtain a corresponding deep learning optimization model and a visual geometry optimization model; Recombining the deep learning optimization model and the visual geometry optimization model to obtain a preliminary neural network model; The preliminary neural network model is validated using a cross-validation algorithm. If the model validation result meets the preset requirements, the acquired anesthesia monitoring data is input into the preliminary neural network model for model training to obtain a corresponding neural network model.
6. An anesthesia depth monitoring system, It is characterized in that include: A data optimization module, used for collecting a number of clinical EEG data, and optimizing each of the clinical EEG data to obtain corresponding EEG optimized data; An average processing module, used for performing corresponding average processing on each of the EEG optimization data to obtain a data sequence of each of the EEG optimization data, and resampling the data sequence to obtain corresponding sampled data; A segmentation analysis module, used for performing segmentation analysis on each of the clinical EEG data to convert the data format of each of the clinical EEG data to obtain a corresponding EEG data matrix; A model building module is used to build a neural network model and input the EEG data matrix into the neural network model for data processing to obtain corresponding data processing results; A depth monitoring module, used for performing depth monitoring on the data processing result and the sampling data to obtain a corresponding depth monitoring result, and constructing a corresponding anesthesia depth index based on each of the clinical EEG data; The anesthesia monitoring module is used to optimize the neural network model according to the anesthesia depth index and the depth monitoring result to obtain an anesthesia depth monitoring model, and perform anesthesia depth monitoring on the patient according to the anesthesia depth monitoring model.
7. The anesthesia depth monitoring system according to claim 6, It is characterized in that The average processing module comprises: A data conversion unit, used for performing data conversion on each of the EEG optimization data to obtain a corresponding dual-frequency EEG index; The averaging processing unit is used to perform averaging processing on the dual-frequency EEG index using a mean absolute error algorithm to obtain a data sequence of each EEG optimization data, and resample the data sequence to obtain corresponding sampling data.
8. The anesthesia depth monitoring system according to claim 6, It is characterized in that The segment analysis module comprises: A normalization processing unit, used for performing data normalization processing on each of the clinical EEG data to obtain an EEG sequence set with the same time interval, and performing normalization processing on the EEG sequence set to obtain a corresponding normalized sequence set; A Fourier transform unit is used to perform interpolation processing on the normalized sequence set, and optimize the time-frequency localized window function to obtain a window optimization function, perform Fourier transform on the normalized sequence set using the Fourier transform optimization function constructed by the window optimization function, and move the window optimization function along the time axis to obtain a corresponding EEG data matrix.
9. A readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the anesthesia depth monitoring method as described in any one of claims 1 to 5 is implemented.
10. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the anesthesia depth monitoring method as described in any one of claims 1 to 5 is implemented.