Data processing method for predicting neurocognitive impairment in patients undergoing extracorporeal circulation cardiac surgery

Abstract

This application provides a data processing method for predicting neurocognitive disorders in patients undergoing cardiopulmonary bypass surgery. This method obtains physiological monitoring data from a target patient during cardiopulmonary bypass, then uses a preset neurocognitive disorder prediction model to determine a neurocognitive disorder prediction feature value based on the physiological monitoring data. When the neurocognitive disorder prediction feature value is determined to be greater than a preset feature threshold, the method outputs neurocognitive disorder warning information, thereby predicting the risk of neurocognitive disorders in patients undergoing cardiopulmonary bypass surgery and providing early warning.

Description

Technical Field

[0001] The present application relates to data processing technology, and in particular to a method for processing data to predict neurocognitive disorders in patients undergoing extracorporeal circulation heart surgery. Background Art

[0002] Cardiac surgery with extracorporeal circulation is an important treatment for certain heart diseases, but patients may experience neurocognitive disorders (NCDs) after such surgery. These symptoms include memory loss and difficulty concentrating, which can severely impact their quality of life and recovery.

[0003] Early identification and prediction of neurocognitive impairment after extracorporeal circulation cardiac surgery remain challenging. Although some studies have explored relevant factors, there is still a lack of effective, reliable, and easily implemented prediction tools to assist clinical decision-making. Clinicians often rely on patients' clinical presentations and subjective reports to assess their risk of neurocognitive impairment, a method that is often inaccurate and inefficient. Summary of the Invention

[0004] The present application provides a method for processing predictive data of neurocognitive disorders in patients undergoing extracorporeal circulation cardiac surgery, which is used to predict neurocognitive disorders based on the patient's physiological monitoring data, thereby helping clinicians identify high-risk patients earlier and take corresponding preventive measures.

[0005] In a first aspect, the present application provides a method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal circulation cardiac surgery, comprising:

[0006] Acquiring physiological monitoring data of a target patient during extracorporeal circulation, the physiological monitoring data including arterial pressure difference monitoring value, blood glucose monitoring value, body temperature monitoring value, and electroencephalogram monitoring signal, wherein the target patient is a patient undergoing cardiac surgery;

[0007] Using a preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction feature value based on the physiological monitoring data;

[0008] If it is determined that the neurocognitive disorder prediction characteristic value is greater than a preset characteristic threshold, neurocognitive disorder warning information is output.

[0009] In the above scheme, by obtaining the patient's physiological monitoring data during extracorporeal circulation (including arterial pressure difference monitoring values, blood glucose monitoring values, body temperature monitoring values and EEG monitoring signals), a preset neurocognitive disorder prediction model is used for analysis to predict the patient's possible neurocognitive disorder risk, thereby timely discovering potential risk factors, and thus helping to take early intervention measures to reduce the incidence and severity of neurocognitive disorders.

[0010] Optionally, the preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer, the input layer is used to receive the physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging and classification prediction on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

[0011] In this solution, a neurocognitive disorder prediction model structure consisting of an input layer, a convolutional layer, a fully connected layer, and an output layer is provided. This allows the model to effectively extract features from EEG monitoring signals and integrate them with other physiological monitoring data, improving prediction accuracy. Furthermore, the convolutional layer extracts useful feature vectors from the EEG monitoring signals, while the fully connected layer combines these features with other physiological data for comprehensive analysis, thereby more accurately predicting the likelihood of neurocognitive disorder.

[0012] Optionally, the fully connected layer is used to perform feature merging, classification and prediction on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector, including:

[0013] The fully connected layer uses formula 1 and generates the neurocognitive disorder prediction feature value based on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector , the formula 1 is:

[0014]

[0015] in, is the arterial pressure difference monitoring value, is the blood glucose monitoring value, is the body temperature monitoring value, is the EEG monitoring signal, is the configuration weight matrix of the fully connected layer, The EEG monitoring signal The EEG signal feature vector processed by a convolutional neural network; is the convolution kernel of the convolutional neural network, is the bias term of the fully connected layer, It is a vector concatenation operation.

[0016] In the above scheme, the fully connected layer uses Equation 1 to comprehensively analyze multiple physiological monitoring data. Through vector concatenation and weight matrix operations, the reliability and accuracy of the prediction results are improved. This formula can better quantify the impact of physiological monitoring data on neurocognitive disorder prediction and help identify key risk factors. By integrating information from different physiological monitoring data sources (arterial pressure gradient, blood glucose level, body temperature, and electroencephalogram), a more comprehensive feature vector is formed through vector concatenation, which helps capture the interrelationships and influences between multimodal data. Subsequently, the feature vectors of the EEG monitoring signals are extracted using a convolutional neural network. This can capture complex patterns in the EEG signals that are associated with neurocognitive disorders. This allows for the handling of nonlinear feature relationships, improving the flexibility and accuracy of the prediction model. The neurocognitive disorder prediction feature values generated in this way can more accurately reflect the likelihood of a patient developing neurocognitive disorder during cardiopulmonary bypass surgery, thereby facilitating the issuance of timely early warning information. It can be seen that integrating multiple physiological monitoring data sources and processing them through deep learning models can improve the ability to predict neurocognitive disorders, thereby facilitating early diagnosis and prevention of such complications. In addition, the multimodal data fusion method can capture complex associations that are difficult to reveal with a single monitoring data, thereby improving the accuracy and reliability of the prediction.

[0017] Optionally, after acquiring the physiological monitoring data of the target patient during extracorporeal circulation, the method further includes:

[0018] Generate a physiological monitoring data sequence based on the physiological monitoring data of the target patient at each time point within a preset period range, the physiological monitoring data sequence including an arterial pressure difference monitoring value sequence, a blood glucose monitoring value sequence, a body temperature monitoring value sequence, and an electroencephalogram monitoring signal sequence;

[0019] Correspondingly, after using the preset neurocognitive disorder prediction model and determining the neurocognitive disorder prediction feature value according to the physiological monitoring data, the method further includes:

[0020] Using the preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic curve based on the physiological monitoring data sequence;

[0021] Correspondingly, if it is determined that the neurocognitive disorder prediction characteristic value is greater than the preset characteristic threshold, outputting neurocognitive disorder warning information includes:

[0022] In the preset first coordinate system, if it is determined that at least a partial area of the neurocognitive disorder prediction characteristic curve is above the preset characteristic threshold, the neurocognitive disorder warning information is output.

[0023] In the above scheme, by generating physiological monitoring data sequences and determining neurocognitive disorder prediction characteristic curves based on these sequences, a more comprehensive assessment of a patient's risk trends throughout treatment can be achieved. Furthermore, the generation of characteristic curves enables physicians to visualize changes in a patient's physiological state, thereby identifying potential problems earlier.

[0024] Optionally, after determining that at least a portion of the neurocognitive disorder prediction characteristic curve is above the preset characteristic threshold, the method further includes:

[0025] In the preset first coordinate system, a neurocognitive disorder prediction characteristic area is determined based on the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, wherein the neurocognitive disorder prediction characteristic area is the area of the portion of the neurocognitive disorder prediction characteristic curve above the preset characteristic threshold, wherein the abscissa axis of the preset first coordinate system is used to represent time, and the ordinate axis is used to represent the neurocognitive disorder prediction characteristic value;

[0026] If the neurocognitive disorder prediction feature area is greater than a preset feature area threshold, the neurocognitive disorder warning information includes a first warning level;

[0027] If the neurocognitive disorder prediction feature area is less than or equal to the preset feature area threshold, the neurocognitive disorder warning information includes a second warning level, wherein the first warning level is higher than the second warning level.

[0028] In this approach, by plotting the neurocognitive impairment prediction curve in the first coordinate system and calculating the area between it and the preset threshold, we can further differentiate between different levels of warning information. This classification of warning levels helps the medical team take different response measures based on the severity of the risk and optimize resource allocation.

[0029] Furthermore, a patient's risk status can be assessed by monitoring the neurocognitive impairment prediction curve, rather than relying solely on data from a single point in time. This approach provides continuous time-series data, enabling better tracking of patient status changes. By calculating the area of the neurocognitive impairment prediction curve above a preset threshold (i.e., the neurocognitive impairment prediction area), the patient's risk level can be quantified. This allows physicians to determine whether a patient is at high risk based on the quantified results. If the neurocognitive impairment prediction area exceeds the preset threshold, a higher-level warning (first warning level) is triggered. Conversely, a smaller area triggers a lower-level warning (second warning level). This tiered warning mechanism helps clinicians take appropriate measures based on different warning levels and effectively allocate medical resources. Providing different levels of warning information can serve as part of a clinical decision support system, helping physicians make more timely and appropriate treatment decisions. For example, for patients at higher risk, physicians may arrange more frequent monitoring or implement preventive interventions. Furthermore, early identification of neurocognitive impairment risk allows for early intervention to reduce the likelihood of adverse consequences. The establishment of this warning system aims to mitigate issues such as cognitive decline after surgery. In addition, different warning thresholds and standards can be further set for different patients. This approach allows for personalized assessment of each patient, thereby improving the pertinence and effectiveness of the warning.

[0030] Optionally, after determining the neurocognitive disorder prediction characteristic area according to the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, the method further includes:

[0031] Generating an EEG monitoring curve within the preset period range according to the EEG monitoring signal sequence, and generating a first characteristic curve according to the EEG monitoring curve, wherein the first characteristic curve is used to characterize the first-order derivative of each point on the EEG monitoring curve;

[0032] In a preset second coordinate system, determining a monitoring feature region sequence based on the first characteristic curve, wherein each monitoring feature region in the monitoring feature region sequence is a closed region formed by each portion of the first characteristic curve below the abscissa axis of the preset second coordinate system and the abscissa axis, the abscissa axis of the preset second coordinate system is used to represent time, and the ordinate axis is used to represent the first-order derivative of the EEG monitoring curve;

[0033] Determine, in the monitoring feature region sequence, a monitoring feature region whose area is greater than a preset region area threshold as a target detection feature region, and determine a characteristic time range corresponding to the target detection feature region on the abscissa axis of the preset second coordinate system;

[0034] generating a predicted time range according to the characteristic time range, and determining a neurocognitive disorder prediction characteristic segment from the neurocognitive disorder prediction characteristic curve using the predicted time range, the predicted time range including the characteristic time range;

[0035] generating a second characteristic curve according to the neurocognitive disorder prediction feature segment, wherein the second characteristic curve is used to represent the second-order derivative of each point on the neurocognitive disorder prediction feature segment;

[0036] In a preset third coordinate system, a second characteristic area is determined based on the second characteristic curve, where the second characteristic area is the area of a portion of a second characteristic region enclosed by the second characteristic curve and the abscissa axis of the preset third coordinate system, which is above the abscissa axis of the preset third coordinate system, wherein the boundaries of the second characteristic region at the two endpoints of the second characteristic curve are boundary lines perpendicular to the abscissa axis of the preset third coordinate system, the abscissa axis of the preset third coordinate system is used to represent time, and the ordinate axis is used to represent the second-order derivative of the neurocognitive disorder prediction feature segment;

[0037] If it is determined that the second characteristic area meets the preset characteristic area condition, the neurocognitive disorder warning information includes a third warning level, which is higher than the first warning level.

[0038] In this approach, by analyzing the EEG signal sequence to generate a first characteristic curve, and based on this, determining the monitoring characteristic area, we can identify time periods that may be associated with neurocognitive impairment. Determining the monitoring characteristic area helps narrow the scope of attention, thereby focusing on high-risk time periods.

[0039] Monitoring the first-order derivative of EEG signals and the second-order derivative of the neurocognitive impairment prediction curve provides a more refined risk assessment tool. Changes in these derivatives can reveal the changing trend and speed of potential cognitive impairment risk. By analyzing the first-order derivative of the EEG signal sequence and combining it with the second-order derivative analysis of the neurocognitive impairment prediction curve, a multidimensional assessment framework is provided that captures more information about the patient's condition. By identifying specific regions within the sequence of monitored feature regions (i.e., target detection feature regions), events or time points that may lead to neurocognitive impairment can be more precisely located. The relevant features of these specific regions may be caused by specific conditions during surgery. A predicted time range is then generated based on the characteristic time range corresponding to the target detection feature region. Analysis of the second characteristic curve is then introduced to calculate the second characteristic area, further enhancing the accuracy of the early warning system. When the second characteristic area meets specific conditions, a higher-level warning (the third warning level) is triggered.

[0040] Optionally, determining whether the second characteristic area satisfies a preset characteristic area condition includes:

[0041] In the preset third coordinate system, a third characteristic area is determined according to the second characteristic curve, where the third characteristic area is the area of a region enclosed by a portion of the second characteristic curve above the abscissa axis of the preset third coordinate system and the abscissa axis;

[0042] It is determined that a ratio of the second characteristic area to the third characteristic area is greater than a preset ratio threshold.

[0043] In the above scheme, by determining whether the second characteristic area meets the preset conditions, the accuracy of the warning information can be further improved. The introduction of the third warning level helps to take more urgent measures in extremely high-risk situations. Specifically, by calculating the ratio of the second characteristic area to the third characteristic area, the risk level of a patient developing neurocognitive disorders can be more accurately determined, thereby providing more accurate warning information. In cases where the preset characteristic area conditions are met, that is, when the third warning level is triggered, further intervention measures can be taken to reduce or alleviate the patient's possible neurocognitive disorder symptoms after surgery and improve the patient's overall recovery.

[0044] Optionally, the start time of the predicted time range is earlier than the start time of the characteristic time range, and the end time of the predicted time range is the same as the end time of the characteristic time range.

[0045] Optionally, generating a predicted time range according to the characteristic time range includes:

[0046] Using formula 2, and according to the characteristic time range Determine the forecast timeframe , wherein the formula 2 is:

[0047]

[0048] in, is the starting time of the characteristic time range, is the end time of the characteristic time range, is the starting time of the forecast time range, is the end time of the forecast time range, is the area of the monitoring feature region, is the preset area threshold, Please book in advance.

[0049] In the above scheme, since the change of the neurocognitive disorder prediction characteristic curve will be ahead of the change of the EEG monitoring signal curve, the starting time of the prediction time range is set to be before the starting time of the characteristic time range, and the ending time of the prediction time range is set to be the same as the ending time of the characteristic time range. This can more effectively determine the change correlation between the neurocognitive disorder prediction characteristic curve and the EEG monitoring signal curve, and generate the prediction time range through the above formula 2, and dynamically adjust the length of the prediction time range to ensure that the above correlation between the two can be reflected while reducing the amount of data processing calculations.

[0050] In a second aspect, the present application provides a device for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery, comprising:

[0051] an acquisition module, configured to acquire physiological monitoring data of a target patient during extracorporeal circulation, the physiological monitoring data including arterial pressure difference monitoring values, blood glucose monitoring values, body temperature monitoring values, and electroencephalogram (EEG) monitoring signals, wherein the target patient is a patient undergoing cardiac surgery;

[0052] a processing module, configured to utilize a preset neurocognitive disorder prediction model and determine a neurocognitive disorder prediction feature value based on the physiological monitoring data;

[0053] The output module is used to output neurocognitive disorder warning information when it is determined that the neurocognitive disorder prediction characteristic value is greater than a preset characteristic threshold.

[0054] Optionally, the preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer, the input layer is used to receive the physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging and classification prediction on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

[0055] Optionally, the processing module is specifically configured to:

[0056] The fully connected layer uses formula 1 and generates the neurocognitive disorder prediction feature value based on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector , the formula 1 is:

[0057]

[0058] in, is the arterial pressure difference monitoring value, is the blood glucose monitoring value, is the body temperature monitoring value, is the EEG monitoring signal, is the configuration weight matrix of the fully connected layer, The EEG monitoring signal The EEG signal feature vector processed by a convolutional neural network; is the convolution kernel of the convolutional neural network, is the bias term of the fully connected layer, It is a vector concatenation operation.

[0059] Optionally, the processing module is specifically configured to:

[0060] Generate a physiological monitoring data sequence based on the physiological monitoring data of the target patient at each time point within a preset period range, the physiological monitoring data sequence including an arterial pressure difference monitoring value sequence, a blood glucose monitoring value sequence, a body temperature monitoring value sequence, and an electroencephalogram monitoring signal sequence;

[0061] Using the preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic curve based on the physiological monitoring data sequence;

[0062] In the preset first coordinate system, if it is determined that at least a partial area of the neurocognitive disorder prediction characteristic curve is above the preset characteristic threshold, the neurocognitive disorder warning information is output.

[0063] Optionally, the processing module is specifically configured to:

[0064] In the preset first coordinate system, a neurocognitive disorder prediction characteristic area is determined based on the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, wherein the neurocognitive disorder prediction characteristic area is the area of the portion of the neurocognitive disorder prediction characteristic curve above the preset characteristic threshold, wherein the abscissa axis of the preset first coordinate system is used to represent time, and the ordinate axis is used to represent the neurocognitive disorder prediction characteristic value;

[0065] If the neurocognitive disorder prediction feature area is greater than a preset feature area threshold, the neurocognitive disorder warning information includes a first warning level;

[0066] If the neurocognitive disorder prediction feature area is less than or equal to the preset feature area threshold, the neurocognitive disorder warning information includes a second warning level, wherein the first warning level is higher than the second warning level.

[0067] Optionally, the processing module is specifically configured to:

[0068] Generating an EEG monitoring curve within the preset period range according to the EEG monitoring signal sequence, and generating a first characteristic curve according to the EEG monitoring curve, wherein the first characteristic curve is used to characterize the first-order derivative of each point on the EEG monitoring curve;

[0069] In a preset second coordinate system, determining a monitoring feature region sequence based on the first characteristic curve, wherein each monitoring feature region in the monitoring feature region sequence is a closed region formed by each portion of the first characteristic curve below the abscissa axis of the preset second coordinate system and the abscissa axis, the abscissa axis of the preset second coordinate system is used to represent time, and the ordinate axis is used to represent the first-order derivative of the EEG monitoring curve;

[0070] Determine, in the monitoring feature region sequence, a monitoring feature region whose area is greater than a preset region area threshold as a target detection feature region, and determine a characteristic time range corresponding to the target detection feature region on the abscissa axis of the preset second coordinate system;

[0071] generating a predicted time range according to the characteristic time range, and determining a neurocognitive disorder prediction characteristic segment from the neurocognitive disorder prediction characteristic curve using the predicted time range, the predicted time range including the characteristic time range;

[0072] generating a second characteristic curve according to the neurocognitive disorder prediction feature segment, wherein the second characteristic curve is used to represent the second-order derivative of each point on the neurocognitive disorder prediction feature segment;

[0073] In a preset third coordinate system, a second characteristic area is determined based on the second characteristic curve, where the second characteristic area is the area of a portion of a second characteristic region enclosed by the second characteristic curve and the abscissa axis of the preset third coordinate system, which is above the abscissa axis of the preset third coordinate system, wherein the boundaries of the second characteristic region at the two endpoints of the second characteristic curve are boundary lines perpendicular to the abscissa axis of the preset third coordinate system, the abscissa axis of the preset third coordinate system is used to represent time, and the ordinate axis is used to represent the second-order derivative of the neurocognitive disorder prediction feature segment;

[0074] If it is determined that the second characteristic area meets the preset characteristic area condition, the neurocognitive disorder warning information includes a third warning level, which is higher than the first warning level.

[0075] Optionally, the processing module is specifically configured to:

[0076] In the preset third coordinate system, a third characteristic area is determined according to the second characteristic curve, where the third characteristic area is the area of a region enclosed by a portion of the second characteristic curve above the abscissa axis of the preset third coordinate system and the abscissa axis;

[0077] It is determined that a ratio of the second characteristic area to the third characteristic area is greater than a preset ratio threshold.

[0078] Optionally, the processing module is specifically configured to:

[0079] Using formula 2, and according to the characteristic time range Determine the forecast timeframe , wherein the formula 2 is:

[0080]

[0081] in, is the starting time of the characteristic time range, is the end time of the characteristic time range, is the starting time of the forecast time range, is the end time of the forecast time range, is the area of the monitoring feature region, is the preset area threshold, Please book in advance.

[0082] In a third aspect, the present application provides an electronic device, comprising:

[0083] processor; and,

[0084] a memory for storing executable instructions of the processor;

[0085] The processor is configured to perform any possible method described in the first aspect by executing the executable instructions.

[0086] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement any possible method described in the first aspect.

[0087] The present application provides a method for processing predictive data of neurocognitive disorders in patients undergoing extracorporeal circulation heart surgery. The method obtains the physiological monitoring data of the target patient during extracorporeal circulation, and then uses a preset neurocognitive disorder prediction model to determine a neurocognitive disorder prediction characteristic value based on the physiological monitoring data. When it is determined that the neurocognitive disorder prediction characteristic value is greater than a preset characteristic threshold, the method outputs neurocognitive disorder warning information, thereby predicting the risk of neurocognitive disorders that may occur in patients undergoing extracorporeal circulation heart surgery, so as to provide an early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0089] Figure 1 This is a flow chart of a method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal cardiac surgery according to an exemplary embodiment of the present application;

[0090] Figure 2 is a flow chart illustrating a method for processing data for predicting neurocognitive disorders in patients undergoing extracorporeal circulation heart surgery according to another exemplary embodiment of the present application;

[0091] Figure 3 1 is a schematic structural diagram of a data processing device for predicting neurocognitive disorders in patients undergoing extracorporeal cardiac surgery according to an exemplary embodiment of the present application;

[0092] Figure 4 It is a structural diagram of an electronic device according to an exemplary embodiment of the present application.

[0093] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0094] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0095] Figure 1 FIG. 1 is a flow chart of a method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal cardiac surgery according to an exemplary embodiment of the present application. Figure 1 As shown, the method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery provided in this embodiment includes:

[0096] S101. Acquire physiological monitoring data of a target patient during extracorporeal circulation.

[0097] In this step, the physiological monitoring data of the target patient during extracorporeal circulation is obtained. The physiological monitoring data includes arterial pressure difference monitoring value, blood glucose monitoring value, body temperature monitoring value and EEG monitoring signal. The target patient is a patient undergoing cardiac surgery.

[0098] S102. Using a preset neurocognitive disorder prediction model, and determining a neurocognitive disorder prediction characteristic value based on physiological monitoring data.

[0099] Optionally, the preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer. The input layer is used to receive physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging and classification prediction on the arterial pressure difference monitoring value, blood glucose monitoring value, body temperature monitoring value and EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

[0100] Furthermore, the fully connected layer can use Formula 1 to generate a neurocognitive disorder prediction feature value based on the arterial pressure difference monitoring value, blood sugar monitoring value, body temperature monitoring value and EEG signal feature vector. , Formula 1 is:

[0101]

[0102] in, is the arterial pressure difference monitoring value, is the blood glucose monitoring value, is the body temperature monitoring value, For EEG monitoring signals, is the configuration weight matrix of the fully connected layer, EEG monitoring signals EEG signal feature vector after processing by convolutional neural network; is the convolution kernel of the convolutional neural network, is the bias term of the fully connected layer, It is a vector concatenation operation.

[0103] Optionally, the above-mentioned preset neurocognitive disorder prediction model can be a deep learning model for predicting the risk of neurocognitive disorder in patients undergoing extracorporeal circulation heart surgery. When training it, a large amount of physiological monitoring data of extracorporeal circulation heart surgery patients can be collected, including arterial pressure difference monitoring values, blood glucose monitoring values, body temperature monitoring values and EEG monitoring signals. Then, these data are preprocessed, such as missing value filling, standardization, etc. Mark whether each data has neurocognitive disorder as a label for the training model to form a training set. Then, define the model structure, including the input layer, convolution layer, fully connected layer and output layer. The convolution layer is responsible for extracting key features from the EEG monitoring signal. The fully connected layer is used to merge all feature vectors and perform classification predictions. The output layer outputs the final neurocognitive disorder prediction feature value.

[0104] During training, the weight matrix of the fully connected layer is now initialized , the bias term of the fully connected layer And the convolution kernel of the convolutional neural network . The convolutional neural network is then trained using the training set, and a loss function (such as binary cross entropy loss) is used to calculate the difference between the predicted value and the actual label. The model parameters are then updated based on the gradient of the loss function, and an optimization algorithm (such as stochastic gradient descent) is used to update the model parameters to minimize the loss function until the model converges or reaches a preset number of training rounds. The model performance is then evaluated on the validation set, and the hyperparameters are adjusted or the model structure is modified based on the performance. Finally, the overall performance of the model is evaluated on an independent test set. The model is then deployed in a real-world application environment for real-time prediction of the risk of neurocognitive disorders in patients.

[0105] S103. Output neurocognitive disorder warning information.

[0106] If it is determined that the neurocognitive disorder prediction characteristic value is greater than the preset characteristic threshold, neurocognitive disorder warning information is output.

[0107] Specifically, when the neurocognitive disorder prediction feature value exceeds a preset feature threshold, a neurocognitive disorder warning message is output. The warning message can be displayed through a graphical interface or an alarm system so that medical staff can take timely measures.

[0108] In this embodiment, by obtaining the physiological monitoring data of the target patient during extracorporeal circulation, a preset neurocognitive disorder prediction model is used to determine the neurocognitive disorder prediction characteristic value based on the physiological monitoring data, and when it is determined that the neurocognitive disorder prediction characteristic value is greater than the preset characteristic threshold, neurocognitive disorder warning information is output, thereby predicting the risk of neurocognitive disorder that may occur in patients undergoing extracorporeal circulation cardiac surgery, so as to provide early warning.

[0109] Figure 2 FIG. 1 is a flow chart of a method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal cardiac surgery according to another exemplary embodiment of the present application. Figure 2 As shown, the method for processing data for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery provided in this embodiment includes:

[0110] S201. Acquire physiological monitoring data of a target patient during extracorporeal circulation.

[0111] In this step, the physiological monitoring data of the target patient during extracorporeal circulation is obtained. The physiological monitoring data includes arterial pressure difference monitoring value, blood glucose monitoring value, body temperature monitoring value and EEG monitoring signal. The target patient is a patient undergoing cardiac surgery.

[0112] S202: Generate a physiological monitoring data sequence according to the physiological monitoring data of the target patient at each time point within a preset period range.

[0113] In this step, a physiological monitoring data sequence can be generated based on the physiological monitoring data of the target patient at each time node within a preset cycle range. The physiological monitoring data sequence includes an arterial pressure difference monitoring value sequence, a blood glucose monitoring value sequence, a body temperature monitoring value sequence, and an EEG monitoring signal sequence.

[0114] S203: Using a preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic value based on physiological monitoring data.

[0115] Optionally, the preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer. The input layer is used to receive physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging and classification prediction on the arterial pressure difference monitoring value, blood glucose monitoring value, body temperature monitoring value and EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

[0116] Furthermore, the fully connected layer can use Formula 1 to generate a neurocognitive disorder prediction feature value based on the arterial pressure difference monitoring value, blood sugar monitoring value, body temperature monitoring value and EEG signal feature vector. , Formula 1 is:

[0117]

[0118] in, is the arterial pressure difference monitoring value, is the blood glucose monitoring value, is the body temperature monitoring value, For EEG monitoring signals, is the configuration weight matrix of the fully connected layer, EEG monitoring signals EEG signal feature vector after processing by convolutional neural network; is the convolution kernel of the convolutional neural network, is the bias term of the fully connected layer, It is a vector concatenation operation.

[0119] Optionally, the above-mentioned preset neurocognitive disorder prediction model can be a deep learning model for predicting the risk of neurocognitive disorder in patients undergoing extracorporeal circulation heart surgery. When training it, a large amount of physiological monitoring data of extracorporeal circulation heart surgery patients can be collected, including arterial pressure difference monitoring values, blood glucose monitoring values, body temperature monitoring values and EEG monitoring signals. Then, these data are preprocessed, such as missing value filling, standardization, etc. Mark whether each data has neurocognitive disorder as a label for the training model to form a training set. Then, define the model structure, including the input layer, convolution layer, fully connected layer and output layer. The convolution layer is responsible for extracting key features from the EEG monitoring signal. The fully connected layer is used to merge all feature vectors and perform classification predictions. The output layer outputs the final neurocognitive disorder prediction feature value.

[0120] During training, the weight matrix of the fully connected layer is now initialized , the bias term of the fully connected layer And the convolution kernel of the convolutional neural network . The convolutional neural network is then trained using the training set, and a loss function (such as binary cross entropy loss) is used to calculate the difference between the predicted value and the actual label. The model parameters are then updated based on the gradient of the loss function, and an optimization algorithm (such as stochastic gradient descent) is used to update the model parameters to minimize the loss function until the model converges or reaches a preset number of training rounds. The model performance is then evaluated on the validation set, and the hyperparameters are adjusted or the model structure is modified based on the performance. Finally, the overall performance of the model is evaluated on an independent test set. The model is then deployed in a real-world application environment for real-time prediction of the risk of neurocognitive disorders in patients.

[0121] S204: Using a preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic curve based on a physiological monitoring data sequence.

[0122] In this step, a preset neurocognitive disorder prediction model can be used to determine a neurocognitive disorder prediction characteristic curve based on a physiological monitoring data sequence.

[0123] S205. Output neurocognitive disorder warning information.

[0124] In the preset first coordinate system, if it is determined that at least a partial area of the neurocognitive disorder prediction characteristic curve is above a preset characteristic threshold, neurocognitive disorder warning information is output.

[0125] Furthermore, in a preset first coordinate system, a neurocognitive disorder prediction characteristic area is determined based on the neurocognitive disorder prediction characteristic curve and a preset characteristic threshold, where the neurocognitive disorder prediction characteristic area is the area of a portion of the neurocognitive disorder prediction characteristic curve above the preset characteristic threshold. The horizontal axis of the preset first coordinate system is used to represent time, and the vertical axis is used to represent the neurocognitive disorder prediction characteristic value.

[0126] If the neurocognitive disorder prediction feature area is greater than the preset feature area threshold, the neurocognitive disorder warning information includes a first warning level;

[0127] If the neurocognitive disorder prediction feature area is less than or equal to the preset feature area threshold, the neurocognitive disorder warning information includes a second warning level, wherein the first warning level is higher than the second warning level.

[0128] Furthermore, after determining the neurocognitive disorder prediction characteristic area according to the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, the method further includes:

[0129] Generating an EEG monitoring curve within a preset period range according to the EEG monitoring signal sequence, and generating a first characteristic curve according to the EEG monitoring curve, wherein the first characteristic curve is used to characterize the first-order derivative of each point on the EEG monitoring curve;

[0130] In a preset second coordinate system, a monitoring feature region sequence is determined based on the first characteristic curve, wherein each monitoring feature region in the monitoring feature region sequence is a closed region formed by each portion of the first characteristic curve below the abscissa axis of the preset second coordinate system and the abscissa axis, wherein the abscissa axis of the preset second coordinate system is used to represent time, and the ordinate axis is used to represent the first-order derivative of the EEG monitoring curve;

[0131] Determine, in the monitoring feature region sequence, a monitoring feature region whose area is greater than a preset region area threshold as a target detection feature region, and determine a characteristic time range corresponding to the target detection feature region on the abscissa axis of a preset second coordinate system;

[0132] generating a predicted time range according to the characteristic time range, and determining a neurocognitive disorder prediction characteristic segment from a neurocognitive disorder prediction characteristic curve using the predicted time range, wherein the predicted time range includes the characteristic time range;

[0133] generating a second characteristic curve according to the neurocognitive disorder prediction feature segment, wherein the second characteristic curve is used to represent the second-order derivative of each point on the neurocognitive disorder prediction feature segment;

[0134] In a preset third coordinate system, a second characteristic area is determined based on the second characteristic curve, where the second characteristic area is the area of a portion of the second characteristic region enclosed by the second characteristic curve and the abscissa axis of the preset third coordinate system, above the abscissa axis of the preset third coordinate system, wherein the boundaries of the second characteristic region at the two endpoints of the second characteristic curve are boundary lines perpendicular to the abscissa axis of the preset third coordinate system, the abscissa axis of the preset third coordinate system is used to represent time, and the ordinate axis is used to represent the second-order derivative of the neurocognitive disorder prediction feature segment;

[0135] If it is determined that the second characteristic area meets the preset characteristic area condition, the neurocognitive disorder warning information includes a third warning level, which is higher than the first warning level.

[0136] Optionally, determining whether the second characteristic area satisfies a preset characteristic area condition includes:

[0137] In a preset third coordinate system, a third characteristic area is determined based on the second characteristic curve, where the third characteristic area is the area of a region enclosed by a portion of the second characteristic curve above the abscissa axis of the preset third coordinate system and the abscissa axis;

[0138] It is determined that the ratio of the second characteristic area to the third characteristic area is greater than a preset ratio threshold.

[0139] In addition, a forecast time range is generated based on the feature time range, including:

[0140] Using formula 2, and according to the characteristic time range Determine the forecast timeframe , where Formula 2 is:

[0141]

[0142] in, is the starting time of the characteristic time range, is the end time of the characteristic time range, is the starting time of the forecast time range, is the end time of the forecast time range, To monitor the characteristic area, is the preset area threshold, Please book in advance.

[0143] Figure 3 FIG. 1 is a schematic diagram of a data processing device for predicting neurocognitive impairment in patients undergoing extracorporeal cardiac surgery according to an exemplary embodiment of the present application. Figure 3 As shown, the prediction data processing device 300 for neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery provided in this embodiment includes:

[0144] An acquisition module 310 is configured to acquire physiological monitoring data of a target patient during extracorporeal circulation, wherein the physiological monitoring data includes an arterial pressure difference monitoring value, a blood glucose monitoring value, a body temperature monitoring value, and an electroencephalogram (EEG) monitoring signal. The target patient is a patient undergoing cardiac surgery.

[0145] a processing module 320 for determining a neurocognitive disorder prediction feature value based on the physiological monitoring data using a preset neurocognitive disorder prediction model;

[0146] The output module 330 is configured to output neurocognitive disorder warning information when it is determined that the neurocognitive disorder prediction characteristic value is greater than a preset characteristic threshold.

[0147] Optionally, the preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer, the input layer is used to receive the physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging and classification prediction on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

[0148] Optionally, the processing module 320 is specifically configured to:

[0149] The fully connected layer uses formula 1 and generates the neurocognitive disorder prediction feature value based on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector , the formula 1 is:

[0150]

[0151] in, is the arterial pressure difference monitoring value, is the blood glucose monitoring value, is the body temperature monitoring value, is the EEG monitoring signal, is the configuration weight matrix of the fully connected layer, The EEG monitoring signal The EEG signal feature vector processed by a convolutional neural network; is the convolution kernel of the convolutional neural network, is the bias term of the fully connected layer, It is a vector concatenation operation.

[0152] Optionally, the processing module 320 is specifically configured to:

[0153] Generate a physiological monitoring data sequence based on the physiological monitoring data of the target patient at each time point within a preset period range, the physiological monitoring data sequence including an arterial pressure difference monitoring value sequence, a blood glucose monitoring value sequence, a body temperature monitoring value sequence, and an electroencephalogram monitoring signal sequence;

[0154] Using the preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic curve based on the physiological monitoring data sequence;

[0155] In the preset first coordinate system, if it is determined that at least a partial area of the neurocognitive disorder prediction characteristic curve is above the preset characteristic threshold, the neurocognitive disorder warning information is output.

[0156] Optionally, the processing module 320 is specifically configured to:

[0157] In the preset first coordinate system, a neurocognitive disorder prediction characteristic area is determined based on the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, wherein the neurocognitive disorder prediction characteristic area is the area of the portion of the neurocognitive disorder prediction characteristic curve above the preset characteristic threshold, wherein the abscissa axis of the preset first coordinate system is used to represent time, and the ordinate axis is used to represent the neurocognitive disorder prediction characteristic value;

[0158] If the neurocognitive disorder prediction feature area is greater than a preset feature area threshold, the neurocognitive disorder warning information includes a first warning level;

[0159] If the neurocognitive disorder prediction feature area is less than or equal to the preset feature area threshold, the neurocognitive disorder warning information includes a second warning level, wherein the first warning level is higher than the second warning level.

[0160] Optionally, the processing module 320 is specifically configured to:

[0161] Generating an EEG monitoring curve within the preset period range according to the EEG monitoring signal sequence, and generating a first characteristic curve according to the EEG monitoring curve, wherein the first characteristic curve is used to characterize the first-order derivative of each point on the EEG monitoring curve;

[0162] In a preset second coordinate system, determining a monitoring feature region sequence based on the first characteristic curve, wherein each monitoring feature region in the monitoring feature region sequence is a closed region formed by each portion of the first characteristic curve below the abscissa axis of the preset second coordinate system and the abscissa axis, the abscissa axis of the preset second coordinate system is used to represent time, and the ordinate axis is used to represent the first-order derivative of the EEG monitoring curve;

[0163] Determine, in the monitoring feature region sequence, a monitoring feature region whose area is greater than a preset region area threshold as a target detection feature region, and determine a characteristic time range corresponding to the target detection feature region on the abscissa axis of the preset second coordinate system;

[0164] generating a predicted time range according to the characteristic time range, and determining a neurocognitive disorder prediction characteristic segment from the neurocognitive disorder prediction characteristic curve using the predicted time range, the predicted time range including the characteristic time range;

[0165] generating a second characteristic curve according to the neurocognitive disorder prediction feature segment, wherein the second characteristic curve is used to represent the second-order derivative of each point on the neurocognitive disorder prediction feature segment;

[0166] In a preset third coordinate system, a second characteristic area is determined based on the second characteristic curve, where the second characteristic area is the area of a portion of a second characteristic region enclosed by the second characteristic curve and the abscissa axis of the preset third coordinate system, which is above the abscissa axis of the preset third coordinate system, wherein the boundaries of the second characteristic region at the two endpoints of the second characteristic curve are boundary lines perpendicular to the abscissa axis of the preset third coordinate system, the abscissa axis of the preset third coordinate system is used to represent time, and the ordinate axis is used to represent the second-order derivative of the neurocognitive disorder prediction feature segment;

[0167] If it is determined that the second characteristic area meets the preset characteristic area condition, the neurocognitive disorder warning information includes a third warning level, which is higher than the first warning level.

[0168] Optionally, the processing module 320 is specifically configured to:

[0169] In the preset third coordinate system, a third characteristic area is determined according to the second characteristic curve, where the third characteristic area is the area of a region enclosed by a portion of the second characteristic curve above the abscissa axis of the preset third coordinate system and the abscissa axis;

[0170] It is determined that a ratio of the second characteristic area to the third characteristic area is greater than a preset ratio threshold.

[0171] Optionally, the processing module 320 is specifically configured to:

[0172] Using formula 2, and according to the characteristic time range Determine the forecast timeframe , wherein the formula 2 is:

[0173]

[0174] in, is the starting time of the characteristic time range, is the end time of the characteristic time range, is the starting time of the forecast time range, is the end time of the forecast time range, is the area of the monitoring feature region, is the preset area threshold, Please book in advance.

[0175] Figure 4 FIG. 1 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 4 As shown, this embodiment provides an electronic device 400 including: a processor 401 and a memory 402; wherein:

[0176] The memory 402 is used to store computer programs. The memory may also be a flash memory.

[0177] The processor 401 is configured to execute the execution instructions stored in the memory to implement each step in the above method. For details, please refer to the relevant description in the above method embodiment.

[0178] Optionally, the memory 402 may be independent or integrated with the processor 401 .

[0179] When the memory 402 is a device independent of the processor 401, the electronic device 400 may further include:

[0180] The bus 403 is used to connect the memory 402 and the processor 401 .

[0181] This embodiment further provides a readable storage medium, in which a computer program is stored. When at least one processor of an electronic device executes the computer program, the electronic device executes the methods provided in the various aforementioned embodiments.

[0182] This embodiment further provides a program product, which includes a computer program stored in a readable storage medium. At least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor can execute the computer program to cause the electronic device to implement the methods provided in the various embodiments described above.

[0183] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered merely as exemplary, and the true scope and spirit of the present application are indicated by the claims.

[0184] It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for processing data to predict neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery, characterized in that: include: Acquiring physiological monitoring data of a target patient during extracorporeal circulation, the physiological monitoring data including an arterial pressure difference monitoring value, a blood glucose monitoring value, a body temperature monitoring value, and an electroencephalogram (EEG) monitoring signal, wherein the target patient is a cardiac surgery patient; Generate a physiological monitoring data sequence based on the physiological monitoring data of the target patient at each time point within a preset period range, the physiological monitoring data sequence including an arterial pressure difference monitoring value sequence, a blood glucose monitoring value sequence, a body temperature monitoring value sequence, and an electroencephalogram monitoring signal sequence; Using a preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction feature value based on the physiological monitoring data; Using the preset neurocognitive disorder prediction model and determining a neurocognitive disorder prediction characteristic curve based on the physiological monitoring data sequence; If it is determined that the neurocognitive disorder prediction characteristic value is greater than a preset characteristic threshold, outputting neurocognitive disorder warning information; including: in a preset first coordinate system, if it is determined that at least a portion of the neurocognitive disorder prediction characteristic curve is above the preset characteristic threshold, outputting the neurocognitive disorder warning information; In the preset first coordinate system, a neurocognitive disorder prediction characteristic area is determined based on the neurocognitive disorder prediction characteristic curve and the preset characteristic threshold, wherein the neurocognitive disorder prediction characteristic area is the area of the portion of the neurocognitive disorder prediction characteristic curve above the preset characteristic threshold, wherein the abscissa axis of the preset first coordinate system is used to represent time, and the ordinate axis is used to represent the neurocognitive disorder prediction characteristic value; If the neurocognitive disorder prediction feature area is greater than a preset feature area threshold, the neurocognitive disorder warning information includes a first warning level; If the neurocognitive disorder prediction feature area is less than or equal to the preset feature area threshold, the neurocognitive disorder warning information includes a second warning level, wherein the first warning level is higher than the second warning level; Generating an EEG monitoring curve within the preset period range according to the EEG monitoring signal sequence, and generating a first characteristic curve according to the EEG monitoring curve, wherein the first characteristic curve is used to characterize the first-order derivative of each point on the EEG monitoring curve; In a preset second coordinate system, determining a monitoring feature region sequence based on the first characteristic curve, wherein each monitoring feature region in the monitoring feature region sequence is a closed region formed by each portion of the first characteristic curve below the abscissa axis of the preset second coordinate system and the abscissa axis, the abscissa axis of the preset second coordinate system is used to represent time, and the ordinate axis is used to represent the first-order derivative of the EEG monitoring curve; Determine, in the monitoring feature region sequence, a monitoring feature region whose area is greater than a preset region area threshold as a target detection feature region, and determine a characteristic time range corresponding to the target detection feature region on the abscissa axis of the preset second coordinate system; Generating a predicted time range based on the characteristic time range, and determining a neurocognitive disorder prediction characteristic segment from the neurocognitive disorder prediction characteristic curve using the predicted time range, wherein the predicted time range includes the characteristic time range; generating the predicted time range based on the characteristic time range includes: Using formula 2, and according to the characteristic time range Determine the forecast timeframe , wherein the formula 2 is: ; Wherein, t1 is the starting time of the characteristic time range, t2 is the ending time of the characteristic time range, t3 is the starting time of the prediction time range, and t4 is the ending time of the prediction time range. is the area of the monitoring feature region, is the preset area threshold, How long in advance should you make a reservation? generating a second characteristic curve according to the neurocognitive disorder prediction feature segment, wherein the second characteristic curve is used to represent the second-order derivative of each point on the neurocognitive disorder prediction feature segment; In a preset third coordinate system, a second characteristic area is determined based on the second characteristic curve, where the second characteristic area is the area of a portion of a second characteristic region enclosed by the second characteristic curve and the abscissa axis of the preset third coordinate system, which is above the abscissa axis of the preset third coordinate system, wherein the boundaries of the second characteristic region at the two endpoints of the second characteristic curve are boundary lines perpendicular to the abscissa axis of the preset third coordinate system, the abscissa axis of the preset third coordinate system is used to represent time, and the ordinate axis is used to represent the second-order derivative of the neurocognitive disorder prediction feature segment; If it is determined that the second characteristic area meets the preset characteristic area condition, the neurocognitive disorder warning information includes a third warning level, which is higher than the first warning level.

2. The method for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery according to claim 1, wherein: The preset neurocognitive disorder prediction model includes an input layer, a convolutional layer, a fully connected layer and an output layer. The input layer is used to receive the physiological monitoring data, the convolutional layer is used to extract the EEG signal feature vector of the EEG monitoring signal, and the fully connected layer is used to perform feature merging, classification and prediction on the arterial pressure difference monitoring value, the blood glucose monitoring value, the body temperature monitoring value and the EEG signal feature vector to output the neurocognitive disorder prediction feature value through the output layer.

3. The method for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery according to claim 1, wherein: The determining that the second characteristic area meets a preset characteristic area condition includes: In the preset third coordinate system, a third characteristic area is determined according to the second characteristic curve, where the third characteristic area is the area of a region enclosed by a portion of the second characteristic curve above the abscissa axis of the preset third coordinate system and the abscissa axis; It is determined that a ratio of the second characteristic area to the third characteristic area is greater than a preset ratio threshold.

4. The method for predicting neurocognitive impairment in patients undergoing extracorporeal circulation heart surgery according to claim 3, wherein: The starting time of the predicted time range is earlier than the starting time of the characteristic time range, and the ending time of the predicted time range is the same as the ending time of the characteristic time range.

5. An electronic device, characterized in that: include: processor; as well as, a memory for storing executable instructions of the processor; The processor is configured to perform the method according to any one of claims 1 to 4 by executing the executable instructions.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 4 when executed by a processor.

Images