Intelligent anesthesia state monitoring and feedback device

Abstract

The invention discloses an intelligent anesthesia state monitoring and feedback device, and belongs to the technical field of anesthesia state monitoring. An intelligent anesthesia state monitoring and feedback device comprises a parameter acquisition and monitoring module, a data fusion module, an anesthesia state evaluation module, a feedback regulation and control module and a man-machine interaction module. The problem that the anesthesia depth change trend of a patient cannot be predicted only by reducing the anesthetic infusion rate in the prior art is solved. According to the invention, the anesthesia state of a patient can be comprehensively reflected, the evaluation accuracy is improved, the operation process is simplified, the working efficiency of anesthetists is improved, a clear anesthesia state evaluation result can be output, specific medication adjustment parameters can be generated based on the evaluation result, the workload of doctors is reduced, the accuracy of anesthesia medication is improved, and the patient experience is improved. By predicting the anesthesia depth change trend of the patient, the anesthetic infusion rate is adjusted in advance, and abnormal fluctuation of the anesthesia depth is avoided.

Description

Technical Field

[0001] This invention relates to the field of anesthesia status monitoring technology, specifically to an intelligent anesthesia status monitoring and feedback device. Background Technology

[0002] In clinical surgery, inaccurate monitoring of anesthesia depth and improper dosage of anesthetic drugs can lead to an inability to accurately assess the patient's anesthetic status, exposing the patient to risks during anesthesia. Furthermore, the lack of effective monitoring during analgesia administration results in the rate of anesthetic drug administration not being scientifically and effectively controlled, affecting the progress of surgery and anesthesia. Finally, the lack of effective monitoring of the optimal depth of anesthesia required for surgery during sedation can easily lead to harm to the surgical patient from anesthetic drugs.

[0003] Chinese Patent Publication No. CN116491898B discloses an intelligent anesthesia target control device and drug delivery method suitable for surgical procedures. The device includes: an anesthesia drug delivery determination module, used to determine the anesthesia drug delivery plan for analgesia and sedation anesthesia for the patient according to the type of surgical procedure; an anesthesia control module, used to monitor the changes in the patient's brain nerve inhibition level index, and control and adjust the anesthesia drug delivery based on the abnormal results of the changes deviating from the normal value of the index; and a sedation control module, used to monitor the patient's vital signs data, and control and adjust the sedation drug delivery according to the optimal depth of anesthesia required for surgical anesthesia and the periodic score changes of the patient's vital signs data.

[0004] The aforementioned patents can only address the issue by reducing the infusion rate of anesthetic drugs during use, and cannot predict the changing trend of the patient's anesthesia depth. Therefore, they do not meet the existing needs. In response, we propose an intelligent anesthesia status monitoring and feedback device. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent anesthesia status monitoring and feedback device that can comprehensively reflect the patient's anesthesia status, provide rich data support for accurate assessment, continuously adapt to new clinical data, improve assessment accuracy, simplify operation procedures, and improve the work efficiency of anesthesiologists. It not only outputs clear anesthesia status assessment results but also generates specific medication adjustment parameters based on the assessment results. Through a medication control auxiliary module, it achieves linkage with anesthesia equipment, providing direct operational guidance to anesthesiologists, reducing their workload, and improving the accuracy of anesthetic medication. By predicting the trend of changes in the patient's anesthesia depth, it enables advance adjustment of the anesthetic drug infusion rate, avoiding abnormal fluctuations in anesthesia depth, thus solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent anesthesia state monitoring and feedback device, comprising:

[0007] The parameter acquisition and monitoring module is used to collect parameters from various dimensions, as well as individual patient characteristic parameters, and monitor environmental parameters.

[0008] The data fusion module is used to preprocess the parameters of each dimension. After the preprocessing is completed, the data is fused by improving the DS evidence theory and the fused feature vector is output.

[0009] The anesthesia status assessment module is used to construct a personalized anesthesia status assessment model based on individual patient characteristic parameters and fused feature vectors, and output anesthesia depth level, physiological function stability and anesthesia risk warning information.

[0010] The feedback control module is used to generate personalized anesthetic drug infusion control instructions based on the anesthetic status assessment results and individual patient characteristic parameters, and transmit the control instructions to the external anesthetic drug infusion device, while predicting the trend of changes in the patient's anesthetic depth.

[0011] The human-computer interaction module is used to set parameters and input commands for anesthesia equipment, display real-time monitoring data of patients, anesthesia status assessment results and drug infusion parameters, and provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

[0012] Preferably, the parameter acquisition and monitoring module specifically includes:

[0013] The patient's physiological parameters, electroencephalogram (EEG) parameters, and anesthetic drug metabolism parameters were collected, along with individual patient characteristic parameters.

[0014] The system collects data on temperature, humidity, air pressure, and anesthetic gas leakage concentration within the operating room. When the anesthetic gas leakage concentration exceeds a safety threshold, an alarm signal is issued, and the leakage information is transmitted to the operating room environmental control system.

[0015] Preferably, the data fusion module specifically includes:

[0016] The acquired parameters in each dimension are subjected to noise reduction, calibration, and format conversion.

[0017] The preprocessed parameters of each dimension are used as evidence bodies, and the basic probability allocation of each evidence body is calculated. Evidence weight coefficients are introduced to improve the DS evidence theory.

[0018] The basic probability assignments and weight coefficients of each piece of evidence are used to perform nonlinear fusion of multi-source data through a neural network. After the fusion is completed, the fused feature vector is output.

[0019] Preferably, the anesthesia status assessment module specifically includes:

[0020] The anesthesia status assessment model is constructed using a convolutional neural network and a long short-term memory network as a framework, and by utilizing individual patient characteristic parameters and feature vectors obtained from the fusion of multi-source data.

[0021] The risk warning unit is used to construct a multi-level risk warning mechanism, triggering the corresponding level of warning based on the risk warning level corresponding to the anesthesia status assessment result.

[0022] The threshold data adjustment unit is used to generate a personalized assessment threshold for the patient and match the type and concentration of anesthetic drug based on the personalized assessment threshold.

[0023] Preferably, the anesthesia status assessment model specifically includes:

[0024] The initial assessment results of the anesthesia status are output by analyzing individual patient characteristic parameters and static fusion data through convolutional neural networks.

[0025] By performing time-series analysis on dynamically fused data using a long short-term memory network, the dynamic changes in the anesthesia state are captured, and dynamic assessment results of the anesthesia state are output.

[0026] The initial assessment results are weighted and fused with the dynamic assessment results. The weighting coefficients are adaptively adjusted according to the surgical stage, and the final output is the anesthesia depth level, physiological function stability and the effect of anesthetic drugs.

[0027] Preferably, the risk warning unit specifically includes:

[0028] A multi-level risk early warning mechanism is constructed based on the results of anesthesia status assessment, with Level 1, Level 2, and Level 3 early warning levels set up.

[0029] A Level 1 warning is triggered when the depth of anesthesia is light or deep, the physiological function is basically stable, or the effect of the anesthetic drug is moderate.

[0030] A level 2 warning is triggered when the anesthesia depth is light and accompanied by heart rate fluctuations exceeding 20% ​​of the baseline value and blood pressure fluctuations exceeding 15% of the baseline value, or when the anesthesia depth is deep and accompanied by blood oxygen saturation below 92%.

[0031] A Level 3 warning is triggered when the anesthesia depth is at the level of over-anesthesia, or when serious abnormalities occur such as a heart rate below 40 beats / minute, blood pressure below 80 / 50 mmHg, or blood oxygen saturation below 85%.

[0032] Preferably, the threshold data adjustment unit specifically includes:

[0033] Construct a threshold database to store reference ranges of anesthesia depth parameters for patients of different ages and with different underlying diseases;

[0034] After the patient's individual information is entered, the personalized threshold setting unit automatically matches the corresponding reference range and, in conjunction with the anesthesiologist's manual adjustment instructions, generates the patient's personalized assessment threshold. The personalized assessment threshold includes the BIS threshold range, SEF95 threshold range, and BSR threshold range corresponding to each anesthesia depth level.

[0035] The system automatically matches the type of anesthetic drug and its corresponding effective concentration value based on the patient's individual assessment threshold.

[0036] When the assessment results do not match the actual situation, a correction command can be input via the touch screen, including the current actual anesthesia status, correction direction, and correction range.

[0037] The adjustment amount of each weight coefficient is calculated based on the degree of deviation. At the same time, all data in the correction process is recorded, including the monitoring parameters before correction, the evaluation results, the correction instructions and the adjusted weight coefficients.

[0038] Preferably, the intelligent control module specifically includes:

[0039] Based on individual patient characteristics and the PK model of anesthetic drugs, the pharmacokinetic parameters of anesthetic drugs are calculated.

[0040] By combining the PD model of anesthetic drugs and the fused feature vector, the target blood drug concentration that can maintain the patient's anesthesia depth at an appropriate level is calculated.

[0041] Based on the difference between the current blood concentration of the anesthetic drug and the target blood concentration, and in conjunction with the infusion characteristics of the drug infusion device, the infusion rate adjustment value of the anesthetic drug is calculated.

[0042] The calculated infusion parameters are converted into control commands that conform to the communication protocol of anesthetic drug infusion devices.

[0043] Based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module, the trend of changes in the patient's anesthesia depth can be predicted.

[0044] Preferably, the prediction of the patient's anesthesia depth trend based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module specifically includes:

[0045] Acquire anesthesia status assessment data and historical monitoring data, and extract key temporal features from the data;

[0046] The extracted key temporal features are input into the LSTM model, and the output is the prediction results of the anesthesia depth level and the corresponding confidence level at three time nodes: 5 minutes, 8 minutes and 10 minutes in the future. At the same time, a prediction curve of the continuous change of the anesthesia depth parameter is generated.

[0047] The predicted results at each time point are compared with the target anesthesia depth range. If the predicted anesthesia depth level at a certain time point will exceed the target range, or the rate of change of the anesthesia depth parameter exceeds the preset threshold, it is determined that there is a risk of anesthesia fluctuation. The pre-adjustment instruction generation process is immediately triggered, and the drug infusion parameters are adjusted in advance according to the pre-adjustment instruction.

[0048] Preferably, the human-computer interaction module specifically includes:

[0049] The display unit is used to set parameters and input commands for the anesthesia equipment, and to display the patient's real-time monitoring data, anesthesia status assessment results and drug infusion parameters.

[0050] The voice interaction module is used to recognize parameter setting commands, data query commands, and equipment control commands, and to provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] This invention, by setting up a multi-dimensional monitoring module, integrates monitoring content from multiple aspects such as electroencephalogram (EEG) signals, physiological parameters, analgesia-related signals, and anesthetic gas concentrations. It overcomes the limitations of existing single-dimensional monitoring technologies, comprehensively reflecting the patient's anesthetic state and providing rich data support for accurate assessment. By constructing an anesthetic state assessment model, it can continuously adapt to new clinical data, improve assessment accuracy, simplify operational procedures, and increase the work efficiency of anesthesiologists. It not only outputs clear anesthetic state assessment results but also generates specific medication adjustment parameters based on these results. Furthermore, through a medication control auxiliary module, it achieves linkage with anesthesia equipment, providing direct operational guidance to anesthesiologists, reducing their workload, and improving the precision of anesthetic medication. By predicting the trend of changes in the patient's anesthesia depth, it enables advance adjustment of the anesthetic drug infusion rate, avoiding abnormal fluctuations in anesthesia depth. Attached Figure Description

[0053] Figure 1 This is a block diagram of the intelligent anesthesia status monitoring and feedback device of the present invention;

[0054] Figure 2 This is a flowchart of the intelligent anesthesia status monitoring and feedback device of the present invention.

[0055] Specific implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] To address the issue that current technologies can only address the problem by reducing the infusion rate of anesthetic drugs and cannot predict the trend of changes in the patient's depth of anesthesia, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution:

[0058] An intelligent anesthesia status monitoring and feedback device, comprising:

[0059] The parameter acquisition and monitoring module is used to collect parameters from various dimensions, as well as individual patient characteristic parameters, and monitor environmental parameters.

[0060] The data fusion module is used to preprocess the parameters of each dimension. After the preprocessing is completed, the data is fused by improving the DS evidence theory and the fused feature vector is output.

[0061] The anesthesia status assessment module is used to construct a personalized anesthesia status assessment model based on individual patient characteristic parameters and fused feature vectors, and output anesthesia depth level, physiological function stability and anesthesia risk warning information.

[0062] The feedback control module is used to generate personalized anesthetic drug infusion control instructions based on the anesthetic status assessment results and individual patient characteristic parameters, and transmit the control instructions to the external anesthetic drug infusion device, while predicting the trend of changes in the patient's anesthetic depth.

[0063] The human-computer interaction module is used to set parameters and input commands for anesthesia equipment, display real-time monitoring data of patients, anesthesia status assessment results and drug infusion parameters, and provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

[0064] The parameter acquisition and monitoring module specifically includes:

[0065] The patient's physiological parameters, electroencephalogram (EEG) parameters, and anesthetic drug metabolism parameters were collected, along with individual patient characteristic parameters.

[0066] The system collects data on temperature, humidity, air pressure, and anesthetic gas leakage concentration within the operating room. When the anesthetic gas leakage concentration exceeds a safety threshold, an alarm signal is issued, and the leakage information is transmitted to the operating room environmental control system.

[0067] The data fusion module specifically includes:

[0068] The acquired parameters in each dimension are subjected to noise reduction, calibration, and format conversion.

[0069] The preprocessed parameters of each dimension are used as evidence bodies, and the basic probability allocation of each evidence body is calculated. The evidence weight coefficient is introduced to improve the DS evidence theory. The weight of EEG parameters and anesthetic drug metabolism parameters is 0.3, the weight of physiological parameters is 0.3, and the weight of environmental parameters is 0.1.

[0070] The basic probability assignments and weight coefficients of each piece of evidence are used to perform nonlinear fusion of multi-source data through a neural network. After the fusion is completed, the fused feature vector is output.

[0071] The anesthesia status assessment module specifically includes:

[0072] The anesthesia status assessment model is constructed using a convolutional neural network and a long short-term memory network as a framework, and by utilizing individual patient characteristic parameters and feature vectors obtained from the fusion of multi-source data.

[0073] The risk warning unit is used to construct a multi-level risk warning mechanism, triggering the corresponding level of warning based on the risk warning level corresponding to the anesthesia status assessment result.

[0074] The threshold data adjustment unit is used to generate a personalized assessment threshold for the patient and match the type and concentration of anesthetic drug based on the personalized assessment threshold.

[0075] The anesthesia status assessment model specifically includes:

[0076] The initial assessment results of the anesthesia status are output by analyzing individual patient characteristic parameters and static fusion data through convolutional neural networks.

[0077] By performing time-series analysis on dynamically fused data using a long short-term memory network, the dynamic changes in the anesthesia state are captured, and dynamic assessment results of the anesthesia state are output.

[0078] The initial assessment results are weighted and fused with the dynamic assessment results. The weighting coefficients are adaptively adjusted according to the surgical stage. For example, the weight of the dynamic assessment results during the induction period is 0.7, the weight of the dynamic assessment results during the maintenance period is 0.6, and the weight of the dynamic assessment results during the recovery period is 0.8. The final output includes the anesthesia depth level, physiological function stability, and the effect of anesthetic drugs. The anesthesia depth level is divided into five levels: awake, light anesthesia, appropriate anesthesia, deep anesthesia, and excessive anesthesia. The physiological function stability is divided into three levels: stable, basically stable, and unstable. The effect of anesthetic drugs is divided into three levels: good effect, moderate effect, and poor effect.

[0079] The risk warning unit specifically includes:

[0080] A multi-level risk early warning mechanism is constructed based on the results of anesthesia status assessment, with Level 1, Level 2, and Level 3 early warning levels set up.

[0081] When the depth of anesthesia is light or deep, the physiological function is basically stable, or the effect of the anesthetic drug is moderate, a level one warning is triggered, and a yellow warning signal is issued to remind the doctor to pay attention to changes in the patient's condition.

[0082] When the anesthesia depth is light and accompanied by heart rate fluctuations exceeding 20% ​​of the baseline value and blood pressure fluctuations exceeding 15% of the baseline value, or when the anesthesia depth is deep and accompanied by blood oxygen saturation below 92%, a level 2 warning is triggered, an orange warning signal is issued, and specific abnormal parameters and possible cause analysis are displayed.

[0083] When the depth of anesthesia is at the level of over-anesthesia, or when serious abnormalities occur such as heart rate below 40 beats / minute, blood pressure below 80 / 50 mmHg, or blood oxygen saturation below 85%, a level three warning is triggered, issuing a red warning signal, activating an audible and visual alarm, and automatically pushing the warning information to the surgical nurse station and the anesthesiologist's mobile phone to ensure that doctors can take timely intervention measures.

[0084] The threshold data adjustment unit specifically includes:

[0085] Construct a threshold database to store reference ranges of anesthesia depth parameters for patients of different ages and with different underlying diseases;

[0086] After the patient's individual information is entered, the personalized threshold setting unit automatically matches the corresponding reference range and, in conjunction with the anesthesiologist's manual adjustment instructions, generates the patient's personalized assessment threshold. The personalized assessment threshold includes the BIS threshold range, SEF95 threshold range, and BSR threshold range corresponding to each anesthesia depth level.

[0087] The system automatically matches the type of anesthetic drug and its corresponding effective concentration value based on the patient's individual assessment threshold.

[0088] When the assessment results do not match the actual situation, a correction command can be input via the touch screen, including the current actual anesthesia status, correction direction, and correction range.

[0089] The adjustment amount of each weight coefficient is calculated based on the degree of deviation, and the adjustment amount shall not exceed 10% of the initial value. The adjusted weight coefficients are then applied to the subsequent evaluation and analysis. At the same time, all data in the calibration process are recorded, including the monitoring parameters before calibration, the evaluation results, the calibration instructions, and the adjusted weight coefficients, for subsequent statistical analysis and optimization.

[0090] The intelligent control module specifically includes:

[0091] Based on individual patient characteristics and the PK model of anesthetic drugs, the pharmacokinetic parameters of anesthetic drugs are calculated.

[0092] By combining the PD model of anesthetic drugs and the fused feature vector, the target blood drug concentration that can maintain the patient's anesthesia depth at an appropriate level is calculated.

[0093] Based on the difference between the current blood concentration of the anesthetic drug and the target blood concentration, and combined with the infusion characteristics of the drug infusion device, the adjustment value of the infusion rate of the anesthetic drug is calculated. For example, when the current blood concentration of propofol is 2 μg / mL and the target blood concentration is 3 μg / mL, the infusion rate needs to be adjusted from 5 mg / (kg·h) to 8 mg / (kg·h) according to the PK model.

[0094] The calculated infusion parameters are converted into control commands conforming to the communication protocol of anesthetic drug infusion devices. These commands include drug type code, infusion rate (unit: mL / h or mg / (kg·h)), infusion dose (unit: mg), and control activation time. Simultaneously, the control command generation unit 42 establishes a command verification mechanism to logically verify the generated control commands (e.g., whether the infusion rate exceeds the device's maximum infusion rate, whether the infusion dose exceeds the safe dose range, etc.), ensuring the safety and effectiveness of the control commands.

[0095] Based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module, the trend of changes in the patient's anesthesia depth can be predicted.

[0096] Based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module, the trend of changes in the patient's depth of anesthesia is predicted, specifically including:

[0097] Acquire anesthesia status assessment data and historical monitoring data. The historical monitoring data includes data on patients undergoing similar surgeries matched with the current anesthesia stage, as well as monitoring data from the preceding 60 minutes of the patient's anesthesia process. Extract key temporal features from the data, such as the first-order difference sequence (reflecting the rate of change) and second-order difference sequence (reflecting the acceleration of change) of BIS values ​​within 10 minutes, and the cross-correlation coefficients between physiological parameters and the amount of anesthetic drugs infused (e.g., the time lag correlation between heart rate changes and the adjustment of propofol infusion rate).

[0098] The extracted key temporal features are input into the LSTM model, and the output is the prediction results of the anesthesia depth level and the corresponding confidence level at three time nodes: 5 minutes, 8 minutes and 10 minutes in the future. At the same time, a prediction curve of the continuous change of the anesthesia depth parameter is generated.

[0099] The predicted results at each time point are compared with the target anesthesia depth range. If the predicted anesthesia depth level at a certain time point will exceed the target range (e.g., if the current anesthesia period is predicted to drop to deep anesthesia period in 10 minutes), or if the rate of change of the anesthesia depth parameter exceeds the preset threshold (e.g., the BIS value decreases by more than 5 per minute), it is determined that there is a risk of anesthesia fluctuation. The pre-adjustment instruction generation process is immediately triggered. According to the pre-adjustment instruction, the drug infusion parameters are adjusted in advance to achieve early intervention in the anesthesia state, reduce anesthesia fluctuation, and ensure early intervention in the anesthesia state.

[0100] The human-computer interaction module specifically includes:

[0101] The display unit is used to set parameters and input commands for the anesthesia equipment, and to display the patient's real-time monitoring data, anesthesia status assessment results and drug infusion parameters.

[0102] The voice interaction module is used to recognize parameter setting commands, data query commands, and equipment control commands, and to provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

[0103] Working principle: When using the intelligent anesthesia state monitoring and feedback device of the present invention, according to Figure 1 and Figure 2 This includes the following steps:

[0104] S1: Simultaneously collect patient physiological parameters, EEG parameters, and anesthetic drug metabolism parameters, while also collecting individual patient characteristic parameters and monitoring environmental parameters;

[0105] S2: Preprocess the collected physiological parameters, EEG parameters, and anesthetic drug metabolism parameters, and output the fused feature vector by improving the DS evidence theory;

[0106] S3: Construct an anesthesia status assessment model, input the fused feature vector and individual feature parameters into the anesthesia assessment model, output the assessment results, and trigger corresponding level warnings based on the assessment results;

[0107] S4: Combine assessment results, individual characteristic parameters and PD model to calculate infusion parameters, generate control instructions, and transmit the control instructions to external anesthetic drug infusion equipment, while predicting the trend of changes in the patient's depth of anesthesia;

[0108] S5: Transmits the patient's real-time monitoring data, anesthesia status assessment results, and drug infusion parameters to the human-computer interaction module for real-time display.

[0109] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0110] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. An intelligent anesthesia status monitoring and feedback device, characterized in that, include: The parameter acquisition and monitoring module is used to collect parameters from various dimensions, as well as individual patient characteristic parameters, and monitor environmental parameters. The data fusion module is used to preprocess the parameters of each dimension. After the preprocessing is completed, the data is fused by improving the DS evidence theory and the fused feature vector is output. The anesthesia status assessment module is used to construct a personalized anesthesia status assessment model based on individual patient characteristic parameters and fused feature vectors, and output anesthesia depth level, physiological function stability and anesthesia risk warning information. The feedback control module is used to generate personalized anesthetic drug infusion control instructions based on the anesthetic status assessment results and individual patient characteristic parameters, and transmit the control instructions to the external anesthetic drug infusion device, while predicting the trend of changes in the patient's anesthetic depth. The human-computer interaction module is used to set parameters and input commands for anesthesia equipment, display real-time monitoring data of patients, anesthesia status assessment results and drug infusion parameters, and provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

2. The intelligent anesthesia state monitoring and feedback device according to claim 1, characterized in that, The parameter acquisition and monitoring module specifically includes: The patient's physiological parameters, electroencephalogram (EEG) parameters, and anesthetic drug metabolism parameters were collected, along with individual patient characteristic parameters. The system collects data on temperature, humidity, air pressure, and anesthetic gas leakage concentration within the operating room. When the anesthetic gas leakage concentration exceeds a safety threshold, an alarm signal is issued, and the leakage information is transmitted to the operating room environmental control system.

3. The intelligent anesthesia state monitoring and feedback device according to claim 1, characterized in that, The data fusion module specifically includes: The acquired parameters in each dimension are subjected to noise reduction, calibration, and format conversion. The preprocessed parameters of each dimension are used as evidence bodies, and the basic probability allocation of each evidence body is calculated. Evidence weight coefficients are introduced to improve the DS evidence theory. The basic probability assignments and weight coefficients of each piece of evidence are used to perform nonlinear fusion of multi-source data through a neural network. After the fusion is completed, the fused feature vector is output.

4. The intelligent anesthesia state monitoring and feedback device according to claim 1, characterized in that, The anesthesia status assessment module specifically includes: The anesthesia status assessment model is constructed using a convolutional neural network and a long short-term memory network as a framework, and by utilizing individual patient characteristic parameters and feature vectors obtained from the fusion of multi-source data. The risk warning unit is used to construct a multi-level risk warning mechanism, triggering the corresponding level of warning based on the risk warning level corresponding to the anesthesia status assessment result. The threshold data adjustment unit is used to generate a personalized assessment threshold for the patient and match the type and concentration of anesthetic drug based on the personalized assessment threshold.

5. The intelligent anesthesia state monitoring and feedback device according to claim 4, characterized in that, The anesthesia status assessment model specifically includes: The initial assessment results of the anesthesia status are output by analyzing individual patient characteristic parameters and static fusion data through convolutional neural networks. By performing time-series analysis on dynamically fused data using a long short-term memory network, the dynamic changes in the anesthesia state are captured, and dynamic assessment results of the anesthesia state are output. The initial assessment results are weighted and fused with the dynamic assessment results. The weighting coefficients are adaptively adjusted according to the surgical stage, and the final output is the anesthesia depth level, physiological function stability and the effect of anesthetic drugs.

6. The intelligent anesthesia state monitoring and feedback device according to claim 4, characterized in that, The risk warning unit specifically includes: A multi-level risk early warning mechanism is constructed based on the results of anesthesia status assessment, with Level 1, Level 2, and Level 3 early warning levels set up. A Level 1 warning is triggered when the depth of anesthesia is light or deep, the physiological function is basically stable, or the effect of the anesthetic drug is moderate. A level 2 warning is triggered when the anesthesia depth is light and accompanied by heart rate fluctuations exceeding 20% ​​of the baseline value and blood pressure fluctuations exceeding 15% of the baseline value, or when the anesthesia depth is deep and accompanied by blood oxygen saturation below 92%. A Level 3 warning is triggered when the anesthesia depth is at the level of over-anesthesia, or when serious abnormalities occur such as a heart rate below 40 beats / minute, blood pressure below 80 / 50 mmHg, or blood oxygen saturation below 85%.

7. The intelligent anesthesia state monitoring and feedback device according to claim 4, characterized in that, The threshold data adjustment unit specifically includes: Construct a threshold database to store reference ranges of anesthesia depth parameters for patients of different ages and with different underlying diseases; After the patient's individual information is entered, the personalized threshold setting unit automatically matches the corresponding reference range and, in conjunction with the anesthesiologist's manual adjustment instructions, generates the patient's personalized assessment threshold. The personalized assessment threshold includes the BIS threshold range, SEF95 threshold range, and BSR threshold range corresponding to each anesthesia depth level. The system automatically matches the type of anesthetic drug and its corresponding effective concentration value based on the patient's individual assessment threshold. When the assessment results do not match the actual situation, a correction command can be input via the touch screen, including the current actual anesthesia status, correction direction, and correction range. The adjustment amount of each weight coefficient is calculated based on the degree of deviation. At the same time, all data in the correction process is recorded, including the monitoring parameters before correction, the evaluation results, the correction instructions and the adjusted weight coefficients.

8. The intelligent anesthesia state monitoring and feedback device according to claim 1, characterized in that, The intelligent control module specifically includes: Based on individual patient characteristics and the PK model of anesthetic drugs, the pharmacokinetic parameters of anesthetic drugs are calculated. By combining the PD model of anesthetic drugs and the fused feature vector, the target blood drug concentration that can maintain the patient's anesthesia depth at an appropriate level is calculated. Based on the difference between the current blood concentration of the anesthetic drug and the target blood concentration, and in conjunction with the infusion characteristics of the drug infusion device, the infusion rate adjustment value of the anesthetic drug is calculated. The calculated infusion parameters are converted into control commands that conform to the communication protocol of anesthetic drug infusion devices; Based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module, the trend of changes in the patient's anesthesia depth can be predicted.

9. The intelligent anesthesia state monitoring and feedback device according to claim 8, characterized in that, The prediction of the patient's anesthesia depth trend based on the anesthesia status assessment data and historical monitoring data output by the anesthesia status assessment module specifically includes: Acquire anesthesia status assessment data and historical monitoring data, and extract key temporal features from the data; The extracted key temporal features are input into the LSTM model, and the output is the prediction results of the anesthesia depth level and the corresponding confidence level at three time nodes: 5 minutes, 8 minutes and 10 minutes in the future. At the same time, a prediction curve of the continuous change of the anesthesia depth parameter is generated. The predicted results at each time point are compared with the target anesthesia depth range. If the predicted anesthesia depth level at a certain time point will exceed the target range, or the rate of change of the anesthesia depth parameter exceeds the preset threshold, it is determined that there is a risk of anesthesia fluctuation. The pre-adjustment instruction generation process is immediately triggered, and the drug infusion parameters are adjusted in advance according to the pre-adjustment instruction.

10. The intelligent anesthesia state monitoring and feedback device according to claim 1, characterized in that, The human-computer interaction module specifically includes: The display unit is used to set parameters and input commands for the anesthesia equipment, and to display the patient's real-time monitoring data, anesthesia status assessment results and drug infusion parameters. The voice interaction module is used to recognize parameter setting commands, data query commands, and equipment control commands, and to provide feedback on abnormal anesthesia status through voice prompts and audible and visual alarms.

Images