Non-invasive body surface anesthesia level monitoring device
Through the non-invasive body surface monitoring device, the evoked potential characteristics of the anesthetic site are monitored and analyzed in real time, and the prediction model is used to generate real-time anesthesia plane, which solves the problems of large monitoring delay and poor accuracy in the prior art, and realizes high-precision anesthesia plane monitoring and timely anesthesia adjustment.
Patent Information
- Application Number
- CN202510007394.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-03
- Publication Date
- 2025-06-03
AI Technical Summary
The existing anesthesia plane monitoring technology has large monitoring delays and poor accuracy, which affects the anesthesia effect and surgical safety.
The non-invasive body surface monitoring device is used to obtain the induced potential timing through the body surface electrode monitoring module. The feature collection module extracts real-time potential characteristic parameters. The anesthesia plane prediction module uses the prediction model to analyze these parameters, generates a real-time anesthesia plane, and displays it in real time through the view rendering module.
It has achieved the reduction of monitoring delay, improved monitoring accuracy, and timely adjusted anesthesia plan to ensure the appropriate anesthesia level of patients during the operation, and improved surgical safety and patient comfort.
Smart Images
Figure CN120078364A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of anesthesia monitoring, and particularly to a non-invasive body surface device for monitoring the anesthesia plane. Background Art
[0002] Regional anesthesia is widely used in clinical surgeries. In the application of regional anesthesia, accurately controlling the anesthesia plane is crucial. Existing anesthesia plane monitoring mostly relies on subjective testing methods such as acupuncture and cold stimulation, which depend on the experience of medical staff, have low monitoring efficiency, large errors, are difficult to perform real-time dynamic monitoring, easily lead to insufficient or excessive anesthesia, affect the anesthesia effect and the progress of the surgery, and endanger the safety of patients.
[0003] In summary, the existing technology has technical problems such as large monitoring delay, poor accuracy, and affecting the anesthesia effect. Summary of the Invention
[0004] The present invention provides a non-invasive body surface device for monitoring the anesthesia plane to solve the technical problems of large monitoring delay, poor accuracy, and affecting the anesthesia effect in the existing technology, and achieve the technical effects of reducing the monitoring delay, improving the monitoring accuracy, and thus timely adding anesthetic drugs.
[0005] The non-invasive body surface device for monitoring the anesthesia plane provided by the present invention includes:
[0006] Anesthesia site acquisition module, configured to acquire a preset key anesthesia site and extract a first key site from the preset key anesthesia site.
[0007] Body surface electrode monitoring module, configured to attach a first body surface electrode to the first key site and monitor a first evoked potential time series of the first key site through the first body surface electrode.
[0008] Feature collection module, configured to read a predetermined potential feature and collect features of the real-time potential corresponding to the real-time moment in the first evoked potential time series based on the predetermined potential feature to obtain real-time potential feature parameters.
[0009] Anesthesia plane prediction module, configured to activate an anesthesia plane prediction model and analyze the real-time potential feature parameters through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane.
[0010] Visual rendering module, configured to render the predicted real-time anesthesia plane to the first key site on the anesthesia plane visual view.
[0011] In a feasible implementation manner, the execution steps of the body surface electrode monitoring module include:
[0012] Read the predetermined time-domain features, and collect the time-domain features of the first evoked potential time series based on the predetermined time-domain features to obtain the first time-domain feature parameter set.
[0013] Perform a fast Fourier transform on the first evoked potential time series to obtain the first evoked potential frequency domain.
[0014] Read the predetermined frequency-domain features, and collect the frequency-domain features of the first evoked potential frequency domain based on the predetermined frequency-domain features to obtain the first frequency-domain feature parameter set.
[0015] Use the multi-domain feature parameter set formed based on the first time-domain feature parameter set and the first frequency-domain feature parameter set as the input information of the anesthesia plane disappearance prediction model.
[0016] Obtain the predicted time for the disappearance of the anesthesia plane at the first key site through the anesthesia plane disappearance prediction model.
[0017] Perform an additional anesthesia treatment on the first key site according to the predicted time.
[0018] In a feasible implementation manner, the anesthesia plane disappearance prediction model refers to an intelligent model obtained by performing machine learning on a historical data set formed based on historical anesthesia plane disappearance records.
[0019] In a feasible implementation manner, the execution steps of the body surface electrode monitoring module further include:
[0020] Form an anesthesia disappearance factor set, and construct a preset anesthesia disappearance factor structure diagram in combination with the anesthesia disappearance factor set.
[0021] Monitor and obtain a target factor parameter set based on the anesthesia disappearance factor set.
[0022] Obtain an anesthesia disappearance coefficient by combining the target factor parameter set and the preset anesthesia disappearance factor structure diagram.
[0023] Adjust the predicted time with the anesthesia disappearance coefficient as the weight coefficient.
[0024] In a feasible implementation manner, the execution steps of the body surface electrode monitoring module further include:
[0025] Form a user factor set in the user dimension.
[0026] Form a stimulus factor set in the stimulus dimension.
[0027] Form an environment factor set in the environment dimension.
[0028] The user factor set, the stimulus factor set, and the environment factor set together constitute the anesthesia disappearance factor set.
[0029] In a feasible implementation, the execution steps of the anesthesia plane prediction module include:
[0030] Obtain an evoked potential data set.
[0031] Extract a first data set of the evoked potential of the first historical site in the evoked potential data set.
[0032] Extract a first anesthesia plane from the first data set and obtain a first anesthesia feature set of the first anesthesia plane.
[0033] Randomly extract a first anesthesia feature from the first anesthesia feature set.
[0034] Based on the first point feature parameter set in the first data set and the first anesthesia feature, construct a first training data set.
[0035] Perform supervised learning and testing on the first training data set to obtain a first prediction model.
[0036] Perform integrated fusion construction on the first prediction model to obtain the anesthesia plane prediction model.
[0037] In a feasible implementation, when performing integrated fusion construction on the first prediction model to obtain the anesthesia plane prediction model, the execution steps of the anesthesia plane prediction module further include:
[0038] Use the first prediction model as a primary prediction unit.
[0039] Based on the first prediction result of the first prediction model, construct a first prediction result vector.
[0040] Store the prediction result vector data set constructed based on the first prediction result vector into a meta-prediction unit.
[0041] Based on the primary prediction unit and the meta-prediction unit, construct the anesthesia plane prediction model.
[0042] In a feasible implementation, rendering the predicted real-time anesthesia plane to the first key site on the anesthesia plane viewable image includes obtaining the first position coordinates of the first key site and forming the anesthesia plane viewable image based on the first position coordinates.
[0043] The present invention discloses a non-invasive body surface anesthesia plane monitoring device, comprising: an anesthetic site acquisition module for acquiring a preset key anesthetic site and extracting a first key site from the preset key anesthetic site; a body surface electrode monitoring module for attaching a first body surface electrode to the first key site and monitoring a first evoked potential time series of the first key site through the first body surface electrode; a feature collection module for reading a predetermined potential feature and collecting features of a real-time potential corresponding to a real-time moment in the first evoked potential time series based on the predetermined potential feature to obtain a real-time potential feature parameter; an anesthesia plane prediction module for activating an anesthesia plane prediction model and analyzing the real-time potential feature parameter through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane; and a view rendering module for rendering the predicted real-time anesthesia plane correspondingly to the first key site on an anesthesia plane view. The non-invasive body surface anesthesia plane monitoring device disclosed by the present invention solves the technical problems of large monitoring delay, poor accuracy, and affecting the anesthesia effect, and realizes the technical effects of reducing the monitoring delay, improving the monitoring accuracy, and thus timely adding anesthetic drugs. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 FIG. is a schematic structural diagram of a non-invasive body surface anesthesia plane monitoring device of the present invention;
[0045] Figure 2 FIG. is a schematic flow diagram of a non-invasive body surface anesthesia plane monitoring device of the present invention.
[0046] Description of reference numerals: anesthetic site acquisition module 11, body surface electrode monitoring module 12, feature collection module 13, anesthesia plane prediction module 14, view rendering module 15. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0047] The above technical solutions will be described in detail below in combination with the accompanying drawings of the specification and specific embodiments to better understand the above technical solutions. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments used to explain the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention. In addition, it should be noted that only parts related to the present invention are shown in the drawings for the convenience of description.
[0048] Embodiment
[0049] Figure 1 FIG. is a schematic structural diagram of a non-invasive body surface anesthesia plane monitoring device of the present invention, wherein the device comprises:
[0050] Anesthesia site acquisition module 11, configured to acquire a preset key anesthesia site and extract a first key site from the preset key anesthesia site.
[0051] Specifically, first determine the preset key anesthesia site in the target scenario. The preset key anesthesia site refers to the key anesthesia site of the human body corresponding to the individual to be anesthetized, such as a spinal anesthesia point, a nerve plexus anesthesia point, or a central nerve point, etc. Among them, the preset key anesthesia site is obtained by analyzing the surgical plan.
[0052] Body surface electrode monitoring module 12, configured to attach a first body surface electrode to the first key site and monitor a first evoked potential time series of the first key site through the first body surface electrode.
[0053] Specifically, the body surface electrode monitoring module is configured to continuously collect evoked potential information of the site through the first body surface electrode attached to the first key site and output it as a first evoked potential time series; among them, the evoked potential information refers to the somatosensory evoked potential on the body surface, which is used to evaluate the response of the central nervous system to external stimuli. Specifically, the somatosensory evoked potential on the body surface is the potential change of the cerebral cortex caused by stimulating the sensory nerves on the body surface. Preferably, it is recorded by placing electrodes on the scalp.
[0054] Specifically, before attachment, first clean the target site to remove grease and impurities to ensure good contact between the electrode and the skin, and then attach the first body surface electrode to the surface of the first key site and fix the electrode to avoid displacement during monitoring.
[0055] In some embodiments, as Figure 2 shown, after attaching the first body surface electrode to the first key site and monitoring a first evoked potential time series of the first key site through the first body surface electrode, the execution steps of the body surface electrode monitoring module 12 include:
[0056] Read preset time-domain features, and collect time-domain features of the first evoked potential time series based on the preset time-domain features to obtain a first time-domain feature parameter set; perform a fast Fourier transform on the first evoked potential time series to obtain a first evoked potential frequency domain; read preset frequency-domain features, and collect frequency-domain features of the first evoked potential frequency domain based on the preset frequency-domain features to obtain a first frequency-domain feature parameter set; use the multi-domain feature parameter set formed based on the first time-domain feature parameter set and the first frequency-domain feature parameter set as input information for the anesthesia plane disappearance prediction model; obtain the predicted time for the anesthesia plane of the first key site to disappear through the anesthesia plane disappearance prediction model; perform anesthesia supplementation processing on the first key site according to the predicted time.
[0057] Specifically, according to predetermined time-domain characteristics, such as signal amplitude, latency, peak interval, etc., feature engineering analysis is performed on the evoked potential time series to extract the first set of time-domain feature parameters, which contains all the time-domain features related to anesthesia plane prediction; then, a fast Fourier transform is performed on the first evoked potential time series to decompose the time-domain components into frequency components, generating the first evoked potential frequency domain, and then corresponding frequency-domain feature indicators (such as main frequency, frequency band energy distribution) are extracted from the first evoked potential frequency domain to generate the first set of frequency-domain feature parameters.
[0058] Specifically, using the first set of time-domain feature parameters and the first set of frequency-domain feature parameters as the multi-domain feature parameter set input into the anesthesia plane disappearance prediction model, the anesthesia plane disappearance prediction model is activated to perform disappearance node prediction, and then the predicted time of the anesthesia plane disappearance at the first key site is obtained. Then, based on the predicted time, it is judged whether additional anesthesia is needed. The multi-domain feature parameter set contains information in two different domains, time domain and frequency domain, and can more comprehensively reflect the anesthesia depth and changes.
[0059] Specifically, if the predicted time is close to the current time or insufficient to complete the current operation, additional anesthesia treatment is performed on the first key site to ensure the continuity of the anesthesia effect and the safety of the operation.
[0060] Through the above process, the evoked potential time series of the key site is monitored in real time, and the disappearance time of the anesthesia plane is predicted, which helps to timely adjust the anesthesia plan, ensure that the patient maintains an appropriate anesthesia level throughout the operation, improve the surgical safety and patient comfort, and at the same time reduce the risk of excessive use of anesthetic drugs.
[0061] In some embodiments, the anesthesia plane disappearance prediction model refers to an intelligent model obtained by performing machine learning on a historical data set constructed based on historical anesthesia plane disappearance records.
[0062] Specifically, the anesthesia plane disappearance prediction model is an intelligent model based on machine learning. This model is trained and optimized through analyzing the historical dataset composed of historical anesthesia plane disappearance records. Exemplarily, first, a large number of historical anesthesia records are collected, including data such as anesthesia dosage, patient physiological parameters, surgical types, anesthesia plane changes, etc. The collected data is organized into a structured format for machine learning analysis. Then, features related to the disappearance of the anesthesia plane are identified from the historical data, including time-domain and frequency-domain features, and the features are processed, such as normalization, encoding conversion, etc., to improve the training efficiency and prediction accuracy of the model. Next, a suitable machine learning algorithm, such as random forest, support vector machine, neural network, etc., is selected to construct the anesthesia plane disappearance prediction model, and the selected model is trained using the historical dataset. The cross-validation method is used to evaluate the stability and accuracy of the model, and indicators such as accuracy, recall rate, F1 score, etc. are used to measure the prediction performance of the model. Furthermore, the model parameters are adjusted to optimize the prediction performance.
[0063] Through the above method, an intelligent model that can accurately predict the anesthesia plane disappearance time can be constructed, which helps to adjust the anesthesia plan in a timely manner, ensure patient safety, and improve surgical efficiency.
[0064] In some implementation manners, after obtaining the predicted time of the anesthesia plane disappearance of the first key part through the anesthesia plane disappearance prediction model, the execution steps of the body surface electrode monitoring module 12 further include:
[0065] Form an anesthesia disappearance factor set, and construct a preset anesthesia disappearance factor structure diagram in combination with the anesthesia disappearance factor set; monitor and obtain a target factor parameter set based on the anesthesia disappearance factor set; obtain an anesthesia disappearance coefficient by combining the target factor parameter set with the preset anesthesia disappearance factor structure diagram; adjust the predicted time with the anesthesia disappearance coefficient as the weight coefficient.
[0066] Specifically, first, through correlation analysis or knowledge base, various factors affecting the disappearance of the anesthesia plane are determined, such as drug type, dosage, patient physiological state, surgical type, etc. Then, these factors are formed into an anesthesia disappearance factor set. Then, a large amount of historical anesthesia data is analyzed through statistical regression analysis methods to obtain the relationships between different factors and their impacts on the anesthesia disappearance time, and a preset anesthesia disappearance factor structure diagram is constructed. Exemplarily, this structure diagram is a graph relationship network, where the graph nodes respectively represent anesthesia effects and various factors, and the weights of the node edges indicate the influence between factors or the influence of factors on anesthesia effects. The greater the influence, the higher the weight of the node edge.
[0067] Specifically, the predicted time refers to the time point when the aforementioned anesthesia plane is expected to disappear. Since the disappearance of the anesthesia plane is a dynamic process and is affected by multiple factors, it is necessary to dynamically adjust the predicted time according to the real-time monitored target factor parameter set and the anesthesia disappearance coefficient, that is, to perform weighted correction on the weights of each influencing factor and the predicted anesthesia disappearance time.
[0068] Exemplarily, when the concentration of the anesthetic drug is low, the disappearance time of the anesthesia plane may be advanced, so the predicted time needs to be adjusted forward; on the contrary, if the drug concentration is high or the patient's metabolism is slow, the anesthesia plane may last for a longer time and the predicted time needs to be adjusted backward.
[0069] By real-time monitoring the anesthesia disappearance factors and dynamically adjusting the predicted time in combination with the anesthesia disappearance coefficient, the time point when the anesthesia plane disappears can be accurately predicted, which helps to timely adjust the anesthesia or add anesthetic drugs, improve the safety and personalization of anesthesia, ensure the continuity of the anesthesia effect, and effectively guarantee the comfort and safety of the patient.
[0070] In some implementation manners, an anesthesia disappearance factor set is formed, and a preset anesthesia disappearance factor structure diagram is constructed in combination with the anesthesia disappearance factor set. The execution steps of the body surface electrode monitoring module 12 further include:
[0071] Form a user factor set in the user dimension; form a stimulation factor set in the stimulation dimension; form an environmental factor set in the environmental dimension; the user factor set, the stimulation factor set and the environmental factor set together form the anesthesia disappearance factor set.
[0072] Specifically, the anesthesia disappearance factor set includes a user factor set, a stimulation factor set, and an environmental factor set. Among them, user factors refer to the characteristics related to the anesthesia object (i.e., the patient), which are usually related to the patient's physiological state, individual differences, health status, and response to anesthetic drugs, including age, weight, gender, cardiovascular status, metabolic status, drug allergy history, or chronic disease status, etc.; stimulation factors refer to external or internal factors that have a stimulating effect on the patient during anesthesia, mainly including the type, dose, administration method of anesthetic drugs, and the interaction between drugs and other drugs, etc.; environmental factors refer to various external influences on the environment where the patient is located during anesthesia, such as temperature, humidity, oxygen concentration, type of surgery (such as open surgery or minimally invasive surgery), etc.
[0073] By combining the user factor set, the stimulation factor set, and the environmental factor set, the key influencing factors during the anesthesia process can be comprehensively and meticulously characterized, which helps to achieve a more accurate prediction of the anesthesia plane disappearance time.
[0074] A feature collection module 13, configured to read predetermined potential features, and collect features of the real-time potential corresponding to the real-time moment in the first evoked potential time series based on the predetermined potential features, so as to obtain real-time potential feature parameters.
[0075] Specifically, the predetermined potential features refer to potential features obtained from previous research, clinical experience, or historical data analysis, and are used to characterize the behavior of evoked potentials under different anesthesia states. The set of predetermined potential features includes: the peak amplitude of the evoked potential, which usually reflects the response intensity of the nervous system. For the anesthesia process, the amplitude of the evoked potential usually decreases as the depth of anesthesia increases; the latency of the evoked potential, that is, the time delay when the potential waveform appears after stimulation. Under anesthesia, the latency will be correspondingly prolonged; the frequency components of the evoked potential waveform, such as the power distribution of the frequency band, are used to identify the changes of the potential in the frequency domain and reflect different states of nerve activity; the shape of the evoked potential waveform, such as whether it is symmetric, the smoothness of the waveform, etc., is related to the health status of the patient's nervous system.
[0076] Specifically, at each real-time moment, real-time data is extracted based on the predetermined potential features, including directly processing the time-domain waveform of the evoked potential and analyzing the time-domain features such as amplitude and latency; through frequency-domain transformation (such as FFT), the spectral features of the evoked potential signal are extracted; the morphology of the evoked potential is analyzed, and by calculating the symmetry, kurtosis, etc. of the waveform, the response mode of the nervous system is evaluated.
[0077] Specifically, the feature parameters of different dimensions are combined together to form real-time potential feature parameters to reflect the response of the patient's nervous system under anesthesia. By extracting real-time potential features from the evoked potential time series, the changes of the potential during the anesthesia process can be accurately and real-time collected and processed, providing key information for anesthesia depth assessment, anesthesia management optimization, and intelligent anesthesia prediction, and having technical effects such as improving anesthesia safety, accuracy, and personalized management.
[0078] An anesthesia plane prediction module 14, configured to activate an anesthesia plane prediction model, and analyze the real-time potential feature parameters through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane.
[0079] Specifically, using the real-time potential feature parameters as input data, the real-time anesthesia plane is analyzed and obtained through the anesthesia plane prediction model. Among them, the anesthesia plane is defined as a continuous range of anesthesia states, which is used to describe the patient's response to anesthetic drugs.
[0080] In some embodiments, the anesthesia plane prediction model is activated, and the real-time potential feature parameters are analyzed through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane. The execution steps of the anesthesia plane prediction module 14 include:
[0081] Obtain an evoked potential dataset; extract a first dataset of the evoked potential of the first historical site in the evoked potential dataset; extract the first anesthesia plane in the first dataset, and obtain the first anesthesia feature set of the first anesthesia plane; randomly extract the first anesthesia feature in the first anesthesia feature set; construct a first training dataset based on the first point feature parameter set in the first dataset and the first anesthesia feature; perform supervised learning and testing on the first training dataset to obtain a first prediction model; perform integrated fusion construction on the first prediction model to obtain the anesthesia plane prediction model.
[0082] Specifically, first, select the evoked potential signals of the first historical site from the evoked potential dataset. This evoked potential dataset contains the evoked potential signals obtained by the patient at multiple different sites. By combining with the labels of the anesthesia plane states (light anesthesia, deep anesthesia, etc.), a data subset related to the first historical site can be extracted; then, determine the anesthesia plane information from the extracted first dataset, and analyze and extract the feature set related to the anesthesia plane. For example, potential amplitude, latency, waveform shape, etc.
[0083] Specifically, randomly extract the first anesthesia feature from the extracted first anesthesia feature set, and combine the first point feature parameter set and the first anesthesia feature to construct a first training dataset. This first training dataset is a specific training dataset for the first anesthesia feature; then, use the first training dataset to perform supervised learning on the prediction model, and test the accuracy and reliability of the model by comparing the prediction results of the model with the actual results.
[0084] Optionally, traverse the first anesthesia feature set for random extraction to obtain multiple prediction models corresponding to different features. When the performances of the multiple prediction models all meet the preset requirements, perform integrated fusion construction based on the integrated learning method to obtain the anesthesia plane prediction model.
[0085] Through this process, an intelligent model that can accurately predict the anesthesia plane can be constructed. Furthermore, it can realize the prediction of the anesthesia effect based on the characteristic parameters of the evoked potential. Among them, the application of the integrated learning method further improves the generalization ability and prediction accuracy of the model, making the anesthesia management more precise and personalized.
[0086] In some implementation manners, perform integrated fusion construction on the first prediction model to obtain the anesthesia plane prediction model. The execution steps of the anesthesia plane prediction module 14 further include:
[0087] Take the first prediction model as the primary prediction unit; based on the first prediction result of the first prediction model, form a first prediction result vector; store the prediction result vector dataset formed based on the first prediction result vector in the meta-prediction unit; build the anesthesia plane prediction model according to the primary prediction unit and the meta-prediction unit.
[0088] Specifically, first, define the trained first prediction model as the primary prediction unit, and use the primary prediction unit to predict a new dataset or a test dataset to generate a first prediction result; then, form the prediction results into a first prediction result vector, and this first prediction result vector contains the prediction outputs of all test samples.
[0089] Specifically, gather multiple first prediction result vectors into a prediction result vector dataset and store it in the meta-prediction unit, and this meta-prediction unit will use the result vector dataset for further analysis and prediction.
[0090] Specifically, select a suitable ensemble learning method, such as voting, stacking, weighted average, etc., combine the direct prediction of the primary prediction unit and the analysis result of the meta-prediction unit to build a more comprehensive anesthesia plane prediction model, and use the validation set to check the prediction accuracy and evaluate the performance of the integrated anesthesia plane prediction model; if the performance evaluation result meets the expectation, then deploy the verified anesthesia plane prediction model to actual applications.
[0091] Through the above process, a more powerful and accurate anesthesia plane prediction model can be constructed. Among them, the primary prediction unit provides fast prediction results, while the meta-prediction unit provides deeper analysis and correction on this basis. Such an integrated fusion method improves the prediction accuracy and robustness of the model.
[0092] The viewable rendering module 15 is used to render the predicted real-time anesthesia plane to the first key part on the anesthesia plane viewable.
[0093] Specifically, obtain the predicted anesthesia plane data from the anesthesia plane prediction model, including the level, range or other relevant parameters of the anesthesia plane, and then map the predicted anesthesia plane data to the corresponding position (i.e., the first key part) of the pre-constructed anesthesia plane viewable template.
[0094] Optionally, represent different levels of the anesthesia plane on the anesthesia plane viewable with different colors or marks to ensure that relevant personnel can clearly see the anesthesia status and distribution of this part, so as to perform more precise anesthesia management.
[0095] In some embodiments, rendering the predicted real-time anesthesia plane onto the first key part on the anesthesia plane viewable image includes obtaining the first position coordinates of the first key part and forming the anesthesia plane viewable image based on the first position coordinates.
[0096] Specifically, through medical imaging technology or a surgical navigation system, determine the precise position coordinates of the first key part on the patient's body, and then input these coordinates into the management system of the anesthesia plane viewable image to provide accurate spatial position information for the subsequent rendering process.
[0097] In summary, a non-invasive body surface monitoring anesthesia plane device provided by the present invention has the following technical effects:
[0098] Through an anesthesia site acquisition module for obtaining a preset key anesthesia site and extracting the first key part from the preset key anesthesia site; a body surface electrode monitoring module for attaching the first body surface electrode to the first key part and monitoring the first evoked potential time series of the first key part through the first body surface electrode; a feature collection module for reading a predetermined potential feature and collecting the feature of the real-time potential corresponding to the real-time moment in the first evoked potential time series based on the predetermined potential feature to obtain a real-time potential feature parameter; an anesthesia plane prediction module for activating an anesthesia plane prediction model and analyzing the real-time potential feature parameter through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane; a viewable image rendering module for rendering the predicted real-time anesthesia plane onto the first key part on the anesthesia plane viewable image, thereby achieving the technical effects of reducing monitoring latency, improving monitoring accuracy, and thus timely adding anesthetic drugs.
[0099] It should be understood that the disclosed embodiments of the present invention and the above descriptions enable those skilled in the art to implement the present invention using the present invention. At the same time, the present invention is not limited to the above-mentioned part of the embodiments. It should be understood that those of ordinary skill in the art can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.
Claims
1. A device for non-invasively monitoring the anesthesia level on the body surface, characterized in that: include: An anesthesia site acquisition module, used to acquire preset key anesthesia sites and extract the first key site among the preset key anesthesia sites; A body surface electrode monitoring module, used for attaching a first body surface electrode to the first key part, and obtaining a first evoked potential timing of the first key part through the first body surface electrode monitoring; a feature collection module, used for reading a predetermined potential feature, and collecting features of a real-time potential corresponding to a real-time moment in the first evoked potential time sequence based on the predetermined potential feature, to obtain a real-time potential feature parameter; An anesthesia plane prediction module is used to activate the anesthesia plane prediction model and analyze the real-time potential characteristic parameters through the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane; The visible graph rendering module is used to render the predicted real-time anesthesia plane to the first key part on the anesthesia plane visible graph.
2. The device for non-invasively monitoring the anesthesia level on the body surface according to claim 1, characterized in that: A first body surface electrode is attached to the first key part, and a first evoked potential timing of the first key part is obtained by monitoring the first body surface electrode. The execution steps of the body surface electrode monitoring module include: Reading a predetermined time domain feature, and collecting time domain features of the first evoked potential time series based on the predetermined time domain feature to obtain a first time domain feature parameter set; Performing fast Fourier transformation on the first evoked potential time series to obtain a first evoked potential frequency domain; Reading a predetermined frequency domain feature, and collecting frequency domain features of the first evoked potential frequency domain based on the predetermined frequency domain feature to obtain a first frequency domain feature parameter set; Using a multi-domain feature parameter set formed based on the first time-domain feature parameter set and the first frequency-domain feature parameter set as input information of an anesthesia plane disappearance prediction model; Obtaining a predicted time of disappearance of the anesthesia plane of the first key part by using the anesthesia plane disappearance prediction model; The first key part is subjected to additional anesthesia treatment according to the predicted time.
3. The device for non-invasively monitoring the anesthesia level on the body surface according to claim 2, characterized in that: The anesthesia plane disappearance prediction model refers to an intelligent model obtained by machine learning of a historical data set composed of historical anesthesia plane disappearance records.
4. The device for non-invasively monitoring the anesthesia level on the body surface according to claim 2, characterized in that: After obtaining the predicted time of disappearance of the anesthesia plane of the first key part through the anesthesia plane disappearance prediction model, the execution step of the body surface electrode monitoring module further includes: Establishing an anesthetic disappearance factor set, and constructing a preset anesthetic disappearance factor structure diagram in combination with the anesthetic disappearance factor set; Obtaining a target factor parameter set based on the anesthesia disappearance factor set monitoring; Combining the target factor parameter set with the preset anesthesia disappearance factor structure diagram to obtain an anesthesia disappearance coefficient; The predicted time is adjusted using the anesthesia disappearance coefficient as a weight coefficient.
5. The device for non-invasively monitoring the anesthesia level on the body surface according to claim 4, characterized in that: A set of anesthesia disappearance factors is formed, and a preset anesthesia disappearance factor structure diagram is constructed in combination with the set of anesthesia disappearance factors. The execution steps of the body surface electrode monitoring module also include: Constructing a user factor set of user dimension; To form a set of stimulus factors that form stimulus dimensions; Construct a set of environmental factors in the environmental dimension; The user factor set, the stimulation factor set and the environmental factor set together constitute the anesthesia disappearance factor set.
6. The device for non-invasively monitoring anesthesia level on body surface according to claim 1, characterized in that: The anesthesia plane prediction model is activated, and the real-time potential characteristic parameters are analyzed by the anesthesia plane prediction model to obtain a predicted real-time anesthesia plane. The execution steps of the anesthesia plane prediction module include: Acquire evoked potential data sets; extracting a first data set of evoked potentials of a first historical part from the evoked potential data sets; Extracting a first anesthesia plane from the first data set, and acquiring a first anesthesia feature set of the first anesthesia plane; Randomly extracting a first anesthesia feature from the first anesthesia feature set; Building a first training data set based on a first point feature parameter set in the first data set and the first anesthesia feature; Performing supervised learning and testing on the first training data set to obtain a first prediction model; The first prediction model is integrated and fused to obtain the anesthesia plane prediction model.
7. The device for non-invasively monitoring anesthesia level on body surface according to claim 6, characterized in that: The first prediction model is integrated and fused to obtain the anesthesia plane prediction model, and the execution steps of the anesthesia plane prediction module also include: Using the first prediction model as a primary prediction unit; Based on the first prediction result of the first prediction model, forming a first prediction result vector; storing the prediction result vector data set formed based on the first prediction result vector into a meta-prediction unit; The anesthesia plane prediction model is constructed based on the primary prediction unit and the meta-prediction unit.
8. The device for non-invasively monitoring anesthesia level on body surface according to claim 1, characterized in that: Rendering the predicted real-time anesthesia plane corresponding to the first key part on the anesthesia plane visual map includes acquiring first position coordinates of the first key part and forming the anesthesia plane visual map based on the first position coordinates.