Anesthesia Depth Dynamic Identification and Auxiliary Analysis Method, Device, Server and Medium

Abstract

The present invention discloses a method, device, server and medium for dynamically identifying and assisting in the analysis of anesthesia depth. The method includes: collecting multiple signals within one cycle, preprocessing the signals to form a signal sequence; respectively inputting the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model and a convolutional neural network model to respectively obtain four kinds of anesthesia-assisted analysis results output, and fusing the four kinds of anesthesia-assisted analysis results according to the weight value corresponding to each model to obtain an anesthesia-assisted analysis result; before inputting the signals of the next cycle into the four models, according to the current anesthesia scenario, evaluating the anesthesia physiological signals, calculating an evaluation value according to the model evaluation method, and adjusting each weight value according to the evaluation value to form an anesthesia-assisted analysis result for the new cycle. It can obtain a more accurate assisted analysis result, providing an accurate assisted judgment result for medical workers to analyze the anesthesia state.

Description

Technical Field

[0001] The present invention relates to the technical field of auxiliary analysis, and particularly to an auxiliary analysis method, device, server and medium for dynamic identification of anesthesia depth. Background Art

[0002] The monitoring of anesthesia depth plays a crucial role in anesthesiology. The accurate assessment of anesthesia depth can effectively avoid excessive or insufficient anesthesia and ensure that patients are in a safe and comfortable state during surgery. In recent years, some anesthesia depth monitoring methods based on conventional physiological signals have been gradually proposed and relevant research has been carried out. These methods infer the anesthesia depth by monitoring conventional physiological signals such as heart rate, blood oxygen saturation, respiratory rate, blood pressure, etc. These signals are relatively easy to obtain and have high clinical universality, which can provide an indirect reference for anesthesia depth. However, the change of a single physiological signal often cannot fully reflect the depth state of anesthesia. Especially when the anesthesia depth is shallow or changes rapidly, traditional methods are difficult to provide accurate judgments.

[0003] Most of the existing anesthesia depth monitoring systems are based on the analysis of single or a few physiological signals, and their monitoring accuracy and real-time performance still face great challenges. During the operation, the anesthesia depth is affected by various factors, such as individual differences of patients, metabolism of anesthetic drugs, types of surgery, etc. These factors make it difficult for the change of a single physiological signal to fully reflect the anesthesia depth, and thus an accurate reference cannot be given. Summary of the Invention

[0004] Embodiments of the present invention provide an auxiliary analysis method, device, server and medium for dynamic identification of anesthesia depth to solve the technical problem of low accuracy of the inference result for anesthesia depth in the prior art.

[0005] In a first aspect, embodiments of the present invention provide an auxiliary analysis method for dynamic identification of anesthesia depth, including:

[0006] Collecting a variety of anesthesia physiological signals by using a wearable device within a cycle, preprocessing the variety of anesthesia physiological signals to form a signal sequence;

[0007] Inputting the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model and a convolutional neural network model respectively, and obtaining the first anesthesia auxiliary analysis result, the second anesthesia auxiliary analysis result, the third anesthesia auxiliary analysis result and the fourth anesthesia auxiliary analysis result output respectively;

[0008] Fusing the first anesthesia auxiliary analysis result, the second anesthesia auxiliary analysis result, the third anesthesia auxiliary analysis result and the fourth anesthesia auxiliary analysis result according to the weight value corresponding to each model to obtain an anesthesia auxiliary analysis result;

[0009] Before inputting the signals of the next cycle into the four models, according to the current anesthesia scenario, evaluate and select the anesthesia physiological signals, determine the model evaluation method based on the evaluated anesthesia physiological signals, calculate the evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form the anesthesia assistance analysis result of the new cycle.

[0010] In a second aspect, an embodiment of the present invention further provides an anesthesia depth dynamic identification assistance analysis device, including:

[0011] A preprocessing module, configured to collect a variety of anesthesia physiological signals by using a wearable device within one cycle, and preprocess the variety of anesthesia physiological signals to form a signal sequence;

[0012] An input module, configured to input the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model respectively, and obtain the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result output respectively;

[0013] A fusion module, configured to fuse the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result according to the weight value corresponding to each model to obtain an anesthesia assistance analysis result;

[0014] An adjustment module, configured to, before inputting the signals of the next cycle into the four models, evaluate and select the anesthesia physiological signals according to the current anesthesia scenario, determine the model evaluation method based on the evaluated anesthesia physiological signals, calculate the evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form the anesthesia assistance analysis result of the new cycle.

[0015] In a third aspect, an embodiment of the present invention further provides a server, including:

[0016] One or more processors;

[0017] A storage device, configured to store one or more programs,

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the anesthesia depth dynamic identification assistance analysis method provided in the above embodiment.

[0019] In a fourth aspect, an embodiment of the present invention further provides a storage medium including computer-executable instructions, and the computer-executable instructions are used to execute the anesthesia depth dynamic identification assistance analysis method provided in the above embodiment when executed by a computer processor.

[0020] The anesthesia depth dynamic identification auxiliary analysis method, device and storage medium provided by the embodiments of the present invention collect a variety of anesthesia physiological signals through a wearable device within one cycle, preprocess the variety of anesthesia physiological signals to form a signal sequence; respectively input the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model and a convolutional neural network model to respectively obtain the first anesthesia auxiliary analysis result, the second anesthesia auxiliary analysis result, the third anesthesia auxiliary analysis result and the fourth anesthesia auxiliary analysis result output; fuse the first anesthesia auxiliary analysis result, the second anesthesia auxiliary analysis result, the third anesthesia auxiliary analysis result and the fourth anesthesia auxiliary analysis result according to the weight value corresponding to each model to obtain an anesthesia auxiliary analysis result; before inputting the signals of the next cycle into the four models, select and evaluate anesthesia physiological signals according to the current anesthesia scenario, determine the model evaluation method according to the evaluated anesthesia physiological signals, calculate an evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form a new cycle anesthesia auxiliary analysis result. The four models can output corresponding auxiliary analysis results according to their respective characteristics, can make full use of a variety of anesthesia physiological signals, respectively analyze the correlation relationships therein, and can analyze the prediction results of the four models according to the actual scenario, determine the prediction accuracy of each model, and can further adjust the weights of the prediction results of each model dynamically according to the prediction accuracy, so as to obtain a more accurate auxiliary analysis result, providing an accurate auxiliary judgment result for medical staff to analyze the anesthesia state. Description of the Drawings

[0021] Other features, objects and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments read with reference to the accompanying drawings:

[0022] Figure 1 is a flowchart of the anesthesia depth dynamic identification auxiliary analysis method provided by Embodiment 1 of the present invention;

[0023] Figure 2 is a flowchart of the anesthesia depth dynamic identification auxiliary analysis method provided by Embodiment 2 of the present invention;

[0024] Figure 3 is a structural schematic diagram of the anesthesia depth dynamic identification auxiliary analysis device provided by Embodiment 3 of the present invention;

[0025] Figure 4 is a structural schematic diagram of the server provided by Embodiment 4 of the present invention. Detailed Embodiments

[0026] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only for explaining the present invention and not for limiting the present invention. Additionally, it should be noted that for the sake of description, only the parts related to the present invention rather than all the structures are shown in the drawings.

[0027] Embodiment 1

[0028] Figure 1 FIG. is a flowchart of the method for dynamically identifying and assisting in analyzing the depth of anesthesia provided in Embodiment 1 of the present invention. This embodiment is applicable to the situation of dynamically identifying and assisting in analyzing the depth of anesthesia during surgical anesthesia. This method can be executed by a device for dynamically identifying and assisting in analyzing the depth of anesthesia, and specifically includes the following steps:

[0029] Step 110, collect a variety of anesthesia physiological signals using a wearable device within a period, and preprocess the variety of anesthesia physiological signals to form a signal sequence.

[0030] In this embodiment, a collection period can be preset in advance. The period can be preset. Exemplarily, 30 s can be adopted, taking into account the integrity and flexibility of the collected data. Exemplarily, a wearable physiological signal collector can be used to collect various physiological signals during anesthesia. The overall design of the wearable multi-sensor module takes flexible materials as the core, combines flexible circuit boards and sensor technologies, realizes the bendable fitting of the biological signal collection area, and reduces signal artifacts caused by body surface movement. Using a multi-channel physiological signal sensing unit, a flexible integrated sensor structure, a multi-modal signal acquisition and dynamic gain control circuit, signals corresponding to conventional physiological parameters such as the patient's heart rate, respiratory rate, blood oxygen saturation, and skin temperature can be collected in real time.

[0031] After the collection is completed, various collected signals can be preprocessed to achieve noise reduction. Exemplarily, by performing a weighted calculation on the signal quality indicators (SQI): the QRS wave power spectrum distribution pSQI, kurtosis kSQI, and baseline relative power basSQI. Use a global time reference to synchronize all collected signals to achieve timestamp alignment. Form a variety of physiological signal sequences.

[0032] Step 120, input the signal sequence into the trained random forest model, support vector machine model, gradient boosting machine model, and convolutional neural network model respectively, and obtain the first anesthesia-assisted analysis result, the second output anesthesia-assisted analysis result, the third anesthesia-assisted analysis result, and the fourth anesthesia-assisted analysis result output respectively.

[0033] The signal sequence is , each sample , four different models (RF, SVM, GBM, CNN) can be used for the prediction of anesthetic depth, and four different prediction results are obtained.

[0034] Among them, is the prediction result of the random forest model; is the prediction result of the support vector machine model; is the prediction result of the gradient boosting machine model; is the prediction result of the convolutional neural network model.

[0035] Step 130, fuse the first anesthetic assistance analysis result, the second output anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result according to the weight value corresponding to each model to obtain the anesthetic assistance analysis result.

[0036] In this embodiment, the predicted value output by the fusion model is the weighted average of the prediction results of each model. The fusion model uses the weighted average method to combine the outputs of the four models. The final predicted value output by the fusion model is the weighted average of the prediction results of each model, and the calculation formula is:

[0037] ,

[0038] where the weight satisfies the following constraint conditions:

[0039] ,

[0040] is the weight of the random forest model; is the weight of the support vector machine model; is the weight of the gradient boosting machine model; is the weight of the convolutional neural network model.

[0041] Map the prediction result to a dimensionless anesthetic depth index of 0 - 100. Initially, the same weight can be uniformly set for each model.

[0042] Step 140, before inputting the signals of the next cycle into the four models, select the evaluation anesthetic physiological signals according to the current anesthetic scenario, determine the model evaluation method according to the evaluation anesthetic physiological signals, calculate the evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form the anesthetic assistance analysis result of the new cycle.

[0043] The anesthetic physiological process is a complex process. Therefore, it is necessary to dynamically monitor the changes in physiological signals and dynamically return the anesthetic analysis results. Therefore, in this embodiment, before the data collected in the next cycle is input into the four models, it is necessary to readjust the weight values of each model according to the current scenario and in combination with the scenario requirements, so as to obtain the anesthetic-assisted analysis results that conform to the current state. Exemplarily, the core anesthetic physiological signals can be determined according to the anesthetic scenario, and the model evaluation method can be determined by using the core anesthetic physiological signals, and the weight values can be adjusted.

[0044] Exemplarily, it may include: when anesthetizing for a vascular surgery scenario, calculating the sensitivity of each model to the heart rate signal; calculating the prediction stability of each model when the heart rate signal changes; determining the model whose sensitivity is greater than the preset sensitivity threshold and whose prediction stability is less than the preset stability threshold, and increasing the weight of the model.

[0045] During the anesthesia process of vascular surgery, the heart rate is an important evaluation index. Therefore, whether the model can have a sensitive response to the heart rate change can determine whether the model can accurately grasp these changes and give relatively accurate evaluation results.

[0046] Therefore, in this embodiment, first calculate the sensitivity of each model to the heart rate, and evaluate the applicability and accuracy of each model in this scenario according to the sensitivity.

[0047] Optionally, the sensitivity of each model to the heart rate can be calculated in the following manner:

[0048] ; where is the sensitivity index, which is used to measure the response degree of the model to the heart rate change;

[0049] is the change rate calculation formula, and the differential calculation method can be adopted, is the output of the model, which refers to the anesthetic depth index predicted by the model;

[0050] is the heart rate signal.

[0051] Through the above method, by calculating the ratio of the change rate of the model output to the change rate of the heart rate, the sensitivity of each model to the heart rate can be calculated.

[0052] Although the sensitivity of the model to the heart rate can be used to determine that the model can perceive the heart rate change and make a quick adjustment. However, these models may also be overly sensitive, resulting in incorrect evaluation results.

[0053] Therefore, in this embodiment, it is also necessary to calculate the prediction stability of each model when the heart rate signal changes. Exemplarily, the prediction stability of each model when the heart rate signal changes can be calculated in the following manner:

[0054] , where is the prediction stability of the model; The standard deviation of the model output.

[0055] Specifically, the standardization can be calculated in the following manner:

[0056] ;

[0057] where is the number of samples of the model data; is the standard deviation representing the model output, reflecting the stability of the model output; is the output value of different models; is the average value of all model outputs.

[0058] For each model, after calculating the heart rate sensitivity and prediction stability, determining the model with the sensitivity greater than the preset sensitivity threshold and the prediction stability less than the preset stability threshold, and increasing the weight of the determined model can improve the accuracy of the anesthesia assistance analysis result.

[0059] And in the next cycle, the above process can be repeated to readjust the weight of each model to achieve a dynamic anesthesia assistance analysis result.

[0060] This embodiment uses a wearable device to collect Collect a variety of anesthesia physiological signals, preprocess the variety of anesthesia physiological signals to form a signal sequence; input the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model respectively, and obtain the first anesthesia-assisted analysis result, the second anesthesia-assisted analysis result, the third anesthesia-assisted analysis result, and the fourth anesthesia-assisted analysis result output respectively; fuse the first anesthesia-assisted analysis result, the second anesthesia-assisted analysis result, the third anesthesia-assisted analysis result, and the fourth anesthesia-assisted analysis result according to the weight value corresponding to each model to obtain an anesthesia-assisted analysis result; before inputting the signals of the next cycle into the four models, according to the current anesthesia scenario, select and evaluate the anesthesia physiological signals, determine the model evaluation method according to the evaluated anesthesia physiological signals, calculate the evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form a new cycle anesthesia-assisted analysis result. The four models can output corresponding assisted analysis results according to their respective characteristics, can make full use of a variety of anesthesia physiological signals, analyze the correlation relationships therein respectively, and can analyze the prediction results of the four models according to the actual scenario, determine the prediction accuracy of each model, and can further adjust the weights of the prediction results of each model dynamically according to the prediction accuracy, and can obtain a more accurate assisted analysis result, providing an accurate auxiliary judgment result for medical staff to analyze the anesthesia state.

[0061] In a preferred implementation manner of this embodiment, the method may further add the following steps: when the duration of the vascular surgery exceeds a preset duration threshold, increase the weights of skin temperature and heart rate variability. During long surgeries, skin temperature (ST) and heart rate variability (HRV) are key indicators because they can reflect the patient's metabolic state and the stability of the autonomic nervous system. Therefore, it is necessary to increase the weights of these two features. Exemplarily, by adjusting the model contribution values of the above two signals and inversely adjusting the acquired signal values of the two signals, the weights of skin temperature and heart rate variability can be increased. Exemplarily, the following method can be adopted:

[0062] ;

[0063] where, is the weight of the heart rate feature;

[0064] 0.35 is the initial weight of the heart rate feature;

[0065] is the SHAP value of the heart rate feature;

[0066] is the average value of the SHAP values of all features.

[0067] .

[0068] wherein, is the weight of the blood pressure feature;

[0069] 0.30 is the initial weight of the blood pressure feature;

[0070] is the SHAP value of the blood pressure feature;

[0071] is the average value of the SHAP values of all features.

[0072] By using the above method, the accuracy of the model's anesthesia-assisted analysis can be further improved during long-term surgeries.

[0073] In another preferred implementation manner obtained in this embodiment, the method may further include the following steps: calculating the contribution degree of each anesthesia physiological signal in each model, determining the anesthesia physiological signals whose contribution degrees exceed a preset contribution degree threshold in all models, and increasing the feature weight of the anesthesia physiological signal according to a preset ratio. SHAP is a feature importance analysis method based on game theory, which can explain the contribution of each feature to the model prediction. By calculating the SHAP value of each anesthesia physiological signal in each model, determining the anesthesia physiological signals whose contribution degrees exceed a preset contribution degree threshold in all models, and further determining the importance of the anesthesia physiological signal at the current stage. Therefore, it is necessary to increase the weight value of the anesthesia physiological signal. Exemplarily, the contribution value can still be used to increase the anesthesia physiological signal by adjusting the contribution value according to a preset ratio.

[0074] Embodiment 2

[0075] Figure 2 is a schematic flowchart of the anesthesia depth dynamic identification and auxiliary analysis method provided in Embodiment 2 of the present invention. This embodiment is optimized based on the above embodiment. The step of selecting an evaluation anesthesia physiological signal according to the current anesthesia scenario and determining the model evaluation method is specifically optimized as follows: in the case of local anesthesia for non-vascular surgery, using the contribution values of all anesthesia physiological signals in each model, calculating the diversity of each model using all anesthesia physiological signals; calculating the correlation between the first anesthesia-assisted analysis result, the second output anesthesia-assisted analysis result, the third anesthesia-assisted analysis result, and the fourth anesthesia-assisted analysis result to obtain a correlation score; determining the models whose diversity exceeds the diversity threshold and whose correlation is within a preset correlation range, and increasing the weight value of the models.

[0076] See Figure 2 , the anesthesia depth dynamic identification and auxiliary analysis method includes:

[0077] Step 210, collect a variety of anesthetic physiological signals using a wearable device within one cycle, preprocess the variety of anesthetic physiological signals, and form a signal sequence.

[0078] Step 220, input the signal sequence into the trained random forest model, support vector machine model, gradient boosting machine model, and convolutional neural network model respectively, and obtain the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result output respectively.

[0079] Step 230, fuse the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result according to the weight value corresponding to each model to obtain an anesthetic assistance analysis result.

[0080] Step 240, before inputting the signals of the next cycle into the four models, when it is a local anesthesia scenario for non-vascular surgery, use the contribution values of all anesthetic physiological signals in each model to calculate the diversity of each model using all anesthetic physiological signals.

[0081] In local anesthesia or non-cardiovascular surgery, heart rate is not a key indicator. Therefore, it is necessary to keep the model weights relatively balanced to avoid over-reliance on heart rate. By dynamically adjusting the model weights and feature weights, the overall prediction performance is optimized.

[0082] Therefore, in this embodiment, through feature diversity, the comprehensive processing ability of the model for a variety of physiological signals is evaluated. Feature diversity is measured by calculating the SHAP value distribution of the model for different features.

[0083] Exemplarily, it can be calculated in the following manner:

[0084] , the standard deviation calculation method can also be calculated by the method provided in the above embodiment.

[0085] Wherein, is the feature diversity score of the model; is the SHAP value of all features.

[0086] Step 250, calculate the correlation between the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result to obtain a correlation score.

[0087] In this embodiment, the model prediction consistency (Model Consistency Score, MCS) is used to evaluate the consistency of the model prediction results. MCS measures its consistency by calculating the correlation between the model output and the outputs of other models:

[0088] ;

[0089] Among them, is the model prediction consistency score, is the output of the current model; is the output of other models.

[0090] Step 260: Determine the models whose diversity exceeds the diversity threshold and whose correlation is within the preset correlation range, increase the weight value of the models, and adjust each weight value according to the evaluation value to form a new cycle of anesthesia assistance analysis results.

[0091] In this embodiment, select the models whose feature diversity is greater than the preset diversity threshold and whose correlation is within the preset correlation range, and increase the weight value of the models during fusion.

[0092] In this embodiment, according to the current anesthesia scenario, select to evaluate the anesthesia physiological signals, and determine the model evaluation method, which is specifically optimized as follows: in the case of local anesthesia for non-vascular surgery, use the contribution values of all anesthesia physiological signals in each model to calculate the diversity of each model using all anesthesia physiological signals; calculate the correlation of the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result to obtain a correlation score; determine the models whose diversity exceeds the diversity threshold, and judge whether the correlation of the models is within the preset correlation range, and increase the weight value of the models. For other anesthesia scenarios, standardized and consistent evaluations can be used to determine which model is more accurate in this scenario, and then dynamically adjust its weight to further improve the accuracy of anesthesia assistance analysis in different anesthesia scenarios.

[0093] In a preferred implementation manner of this embodiment, the method may further add the following steps: calculate the contribution value of each anesthesia physiological signal in each cycle, and when the contribution value of the respiratory rate or blood oxygen saturation is higher than that of other anesthesia victory signals by more than a preset multiple, increase the feature weight of the respiratory rate or blood oxygen saturation. Pediatric patients usually have a faster metabolism rate of anesthetic drugs and a more vulnerable respiratory system. Therefore, the respiratory rate (RR) and blood oxygen saturation (SpO2) are key indicators. Increasing the weights of these two features can better reflect the tolerance of pediatric patients to anesthesia and the changes in their physiological states. According to the real-time SHAP contribution value analysis, if the SHAP value of the respiratory rate or blood oxygen saturation is significantly higher than other features, further increase its weight.

[0094] Exemplarily, it can be achieved in the following way:

[0095] , is the weight of the respiratory rate feature;

[0096] 0.35 is the initial weight of the respiratory rate feature;

[0097] is the SHAP value of the respiratory rate feature;

[0098] is the average value of the SHAP values of all features.

[0099] ,

[0100] is the weight of the blood oxygen saturation feature;

[0101] 0.30 is the initial weight of the blood oxygen saturation feature;

[0102] is the SHAP value of the blood oxygen saturation feature;

[0103] is the average value of the SHAP values of all features.

[0104] Through the above method, more accurate anesthesia assistance analysis results can be given for the pediatric anesthesia scenario.

[0105] In another preferred embodiment of this embodiment, the method may further include the following steps: determining a model with a correlation score less than a preset correlation score threshold, and reducing the weight of the model. If the correlation score is too low, it may indicate that its prediction result is too different from other models and its prediction accuracy is not high. Therefore, it is necessary to reduce the weight of this model.

[0106] Embodiment III

[0107] Figure 3 is a schematic structural diagram of the anesthesia depth dynamic identification assistance analysis device provided in Embodiment III of the present invention. Refer to Figure 3 , the anesthesia depth dynamic identification assistance analysis device includes:

[0108] A preprocessing module 310, configured to collect a variety of anesthesia physiological signals using a wearable device within one cycle, preprocess the variety of anesthesia physiological signals, and form a signal sequence;

[0109] An input module 320, configured to respectively input the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model, and respectively obtain a first anesthesia assistance analysis result, a second output anesthesia assistance analysis result, a third anesthesia assistance analysis result, and a fourth anesthesia assistance analysis result;

[0110] The fusion module 330 is configured to fuse the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result according to the weight value corresponding to each model to obtain an anesthesia assistance analysis result;

[0111] The adjustment module 340 is configured to, before inputting the signals of the next cycle into the four models, select an evaluation anesthesia physiological signal according to the current anesthesia scenario, determine a model evaluation method according to the evaluation anesthesia physiological signal, calculate an evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form a new cycle anesthesia assistance analysis result.

[0112] The anesthesia depth dynamic identification assistance analysis device provided in this embodiment collects a variety of anesthesia physiological signals through a wearable device within one cycle, preprocesses the variety of anesthesia physiological signals to form a signal sequence; inputs the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model respectively, and obtains the output first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result respectively; fuses the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result according to the weight value corresponding to each model to obtain an anesthesia assistance analysis result; before inputting the signals of the next cycle into the four models, select an evaluation anesthesia physiological signal according to the current anesthesia scenario, determine a model evaluation method according to the evaluation anesthesia physiological signal, calculate an evaluation value according to the model evaluation method, and adjust each weight value according to the evaluation value to form a new cycle anesthesia assistance analysis result. By using four models, corresponding assistance analysis results can be output according to their respective characteristics, making full use of a variety of anesthesia physiological signals, respectively analyzing the correlation relationships therein, and analyzing the prediction results of the four models according to the actual scenario to determine the prediction accuracy of each model, and further dynamically adjusting the weights of the prediction results of each model according to the prediction accuracy, so as to obtain a more accurate assistance analysis result, providing an accurate assistance judgment result for medical staff to analyze the anesthesia state.

[0113] Based on the above embodiments, the adjustment module includes:

[0114] The sensitivity calculation unit is configured to calculate the sensitivity of each model to the heart rate signal when in an anesthesia scenario for vascular surgery;

[0115] The stability calculation unit is configured to calculate the prediction stability of each model when the heart rate signal changes;

[0116] A first weight increasing unit, configured to determine a model whose sensitivity is greater than a preset sensitivity threshold and whose prediction stability is less than a preset stability threshold, and increase the weight of the model.

[0117] Based on the above embodiments, the adjustment module further includes:

[0118] A diversity calculation unit, configured to select all anesthesia physiological signals as evaluation anesthesia physiological signals in the case of local anesthesia for non-vascular surgery, and calculate the diversity of each model using all anesthesia physiological signals by using the contribution values of all anesthesia physiological signals in each model;

[0119] A correlation calculation unit, configured to calculate the correlation of the first anesthesia assistance analysis result, the second output anesthesia assistance analysis result, the third anesthesia assistance analysis result, and the fourth anesthesia assistance analysis result, and obtain a correlation score;

[0120] A second weight increasing unit, configured to determine a model whose diversity exceeds a diversity threshold and whose correlation is within a preset correlation range, and increase the weight value of the model.

[0121] Based on the above embodiments, the device further includes:

[0122] A weight reduction module, configured to determine a model whose correlation score is less than a preset correlation score threshold, and reduce the weight of the model.

[0123] Based on the above embodiments, the device further includes:

[0124] A first growth module, configured to calculate the contribution degree of each anesthesia physiological signal in each model, determine the anesthesia physiological signals whose contribution degrees exceed a preset contribution degree threshold in all models, and increase the feature weights of the anesthesia physiological signals according to a preset ratio.

[0125] Based on the above embodiments, the device further includes:

[0126] A second growth module, configured to calculate the contribution value of each anesthesia physiological signal in each period, and increase the feature weight of the respiratory rate or blood oxygen saturation when the contribution value of the respiratory rate or blood oxygen saturation is higher than the contribution values of other anesthesia physiological signals by more than a preset multiple.

[0127] Based on the above embodiments, the device further includes:

[0128] A third growth module, configured to increase the weights of skin temperature and heart rate variability when the duration of vascular surgery exceeds a preset duration threshold.

[0129] The anesthesia depth dynamic identification and auxiliary analysis device provided by the embodiments of the present invention can execute the anesthesia depth dynamic identification and auxiliary analysis method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0130] Embodiment 4

[0131] Figure 4 It is a schematic structural diagram of a server provided by Embodiment 6 of the present invention. Figure 4 It shows a block diagram of an exemplary server 12 suitable for implementing the embodiments of the present invention. Figure 4 The shown server 12 is only an example and should not impose any limitation on the functions and usage scope of the embodiments of the present invention.

[0132] As Figure 4 shown, the server 12 is presented in the form of a general-purpose computing device. The components of the server 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0133] The bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus structures. For example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0134] The server 12 typically includes a variety of computer system readable media. These media can be any available media accessible by the server 12, including volatile and non-volatile media, removable and non-removable media.

[0135] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The server 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 34 may be used for reading and writing non-removable, non-volatile magnetic media ( Figure 4 not shown, commonly referred to as a "hard disk drive"). Although Figure 4Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (such as a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM, or other optical medium) can be provided. In these cases, each drive can be connected to the bus 18 through one or more data medium interfaces. The memory 28 may include at least one program product having a set (such as at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0136] A program / utility 40 having a set (at least one) of program modules 42 can be stored, for example, in the memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment. The program modules 42 generally perform the functions and / or methods in the embodiments described in the present invention.

[0137] The server 12 can also communicate with one or more external devices 14 (such as a keyboard, a pointing device, a display 24, etc.), and can also communicate with one or more devices that enable a user to interact with the server 12, and / or communicate with any device that enables the server 12 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 22. Also, the server 12 can communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 20. As shown in the figure, the network adapter 20 communicates with other modules of the server 12 through the bus 18. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the server 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0138] The processing unit 16 executes various functional applications and data processing by running the programs stored in the system memory 28, such as implementing the anesthesia depth dynamic identification and auxiliary analysis method provided by the embodiments of the present invention.

[0139] Embodiment Five

[0140] Embodiment Five of the present invention also provides a storage medium containing computer-executable instructions, and the computer-executable instructions are used to execute any one of the anesthesia depth dynamic identification and auxiliary analysis methods provided by the above embodiments when executed by a computer processor.

[0141] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable media may be computer-readable signal media or computer-readable storage media. The computer-readable storage media may be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, the computer-readable storage media may be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component.

[0142] The computer-readable signal media may include data signals propagated in a baseband or as part of a carrier wave, which carry computer-readable program codes. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal media may also be any computer-readable media other than the computer-readable storage media, and this computer-readable media can send, propagate, or transmit a program for use by or in combination with an instruction execution system, device, or component.

[0143] The program codes contained on the computer-readable media can be transmitted by any appropriate media, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0144] The computer program codes for performing the operations of the present invention can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program codes can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or device. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0145] Note that the above is only a preferred embodiment of the present invention and the technical principles applied. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, re-adjustments and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, more other equivalent embodiments can be included, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for dynamically identifying and assisting in analyzing the depth of anesthesia, characterized in that Including: Collecting a variety of anesthetic physiological signals using a wearable device within one cycle, preprocessing the variety of anesthetic physiological signals to form a signal sequence; Inputting the signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model respectively, and obtaining the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result output respectively; Fusing the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result according to the weight value corresponding to each model to obtain an anesthetic assistance analysis result; Before inputting the signals of the next cycle into the four models, according to the current anesthetic scenario, selecting an evaluation anesthetic physiological signal, determining a model evaluation method according to the evaluation anesthetic physiological signal, calculating an evaluation value according to the model evaluation method, and adjusting each weight value according to the evaluation value to form a new cycle anesthetic assistance analysis result; The step of selecting an evaluation anesthetic physiological signal according to the current anesthetic scenario and determining a model evaluation method according to the evaluation anesthetic physiological signal includes: When in an anesthetic scenario for vascular surgery, calculating the sensitivity of each model to the heart rate signal; Calculating the prediction stability of each model when the heart rate signal changes; Determining the model whose sensitivity is greater than a preset sensitivity threshold and whose prediction stability is less than a preset stability threshold, and increasing the weight of the model; When in a local anesthetic scenario for non-vascular surgery, selecting all anesthetic physiological signals as the evaluation anesthetic physiological signal, and calculating the diversity of each model using all anesthetic physiological signals by using the contribution values of all anesthetic physiological signals in each model; Calculating the correlation of the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result to obtain the correlation score of each model; Determining the model whose diversity exceeds the diversity threshold and whose correlation score is within a preset score range, and increasing the weight value of the model.

2. The method according to claim 1, wherein The method further includes: Determining the model whose correlation score is less than a preset correlation score threshold, and reducing the weight of the model.

3. The method according to claim 1, wherein The method further includes: Calculating the contribution degree of each anesthetic physiological signal in each model, determining the anesthetic physiological signals whose contribution degrees exceed a preset contribution degree threshold in all models, and increasing the feature weight of the anesthetic physiological signals according to a preset ratio.

4. The method according to claim 1, wherein The method further includes: Calculating the contribution value of each anesthetic physiological signal in each cycle, and when the contribution value of the respiratory rate or blood oxygen saturation is higher than that of other anesthetic physiological signals by more than a preset multiple, increasing the feature weight of the respiratory rate or blood oxygen saturation.

5. The method according to claim 1, wherein The method further includes: When the duration of vascular surgery exceeds a preset duration threshold, increasing the weights of skin temperature and heart rate variability.

6. An apparatus for dynamically identifying and assisting in analyzing the depth of anesthesia, characterized in that, Including: A preprocessing module for collecting a variety of anesthetic physiological signals using a wearable device within one cycle, preprocessing the variety of anesthetic physiological signals to form a signal sequence; An input module for respectively inputting a signal sequence into a trained random forest model, a support vector machine model, a gradient boosting machine model, and a convolutional neural network model to obtain the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result output respectively; A fusion module for fusing the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result according to the weight value corresponding to each model to obtain an anesthetic assistance analysis result; An adjustment module for, before inputting the signals of the next cycle into the four models, selecting an evaluation anesthetic physiological signal according to the current anesthetic scenario, determining a model evaluation method according to the evaluation anesthetic physiological signal, calculating an evaluation value according to the model evaluation method, and adjusting each weight value according to the evaluation value to form a new cycle anesthetic assistance analysis result; The adjustment module includes: A sensitivity calculation unit for calculating the sensitivity of each model to the heart rate signal when the anesthetic scenario is for vascular surgery; A stability calculation unit for calculating the prediction stability of each model when the heart rate signal changes; A first weight increasing unit for determining a model whose sensitivity is greater than a preset sensitivity threshold and whose prediction stability is less than a preset stability threshold, and increasing the weight of the model; A diversity calculation unit for, when the anesthetic scenario is for non-vascular surgery local anesthesia, selecting all anesthetic physiological signals as the evaluation anesthetic physiological signal, and calculating the diversity of each model using all anesthetic physiological signals by using the contribution values of all anesthetic physiological signals in each model; A correlation calculation unit for calculating the correlation of the first anesthetic assistance analysis result, the second anesthetic assistance analysis result, the third anesthetic assistance analysis result, and the fourth anesthetic assistance analysis result to obtain a correlation score; A second weight increasing unit for determining a model whose diversity exceeds the diversity threshold and whose correlation is within a preset correlation range, and increasing the weight value of the model.

7. A server, characterized in that, The server includes: One or more processors; A storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the anesthetic depth dynamic identification assistance analysis method as described in any one of claims 1-5.

8. A storage medium containing computer-executable instructions, characterized in that, The computer-executable instructions are used to execute the anesthetic depth dynamic identification assistance analysis method as described in any one of claims 1-5 when executed by a computer processor.

Images