Neurosurgery risk prediction method and system based on big data analysis

Abstract

The invention relates to the technical field of health risk assessment, in particular to a neurosurgery risk prediction method and system based on big data analysis, and the method comprises the steps: obtaining the neural signal intensity of a brain region in preoperative functional image data of a patient, analyzing the balance degree of neural signal transmission, and analyzing the balance of oxygen metabolism distribution according to the nerve conduction rate. And preoperative nerve state information is formed. According to the method, by accurately extracting the preoperative neural activity signals and analyzing the signal transmission delay, the fluctuation amplitude and the oxygen metabolism balance, the accuracy of individual neural function state evaluation is improved. In the postoperative recovery stage, the abnormal region is dynamically recognized in combination with the neural signal change trend, and the metabolism matching capacity is evaluated, so that the recovery mode is more refined. Neural network connection adjustment capability analysis enhances signal transmission stability evaluation, and abnormal region identification accuracy is improved by combining time deviation and oxygen metabolism recovery consistency.

Description

Technical Field

[0001] The present invention relates to the technical field of health risk assessment, and particularly to a neurosurgical risk prediction method and system based on big data analysis. Background Art

[0002] The technical field of health risk assessment includes technical means for analyzing, predicting, and managing the health status of individuals or groups. The core content is to quantitatively evaluate the health risks of individuals or groups based on multiple data sources, including clinical medical data, biometric data, lifestyle data, etc., through methods such as data modeling, feature extraction, and risk factor analysis. Health risk assessment is widely applied in disease prevention, personalized medicine, health management, etc. Common technical means include risk modeling based on statistical methods, health data mining using machine learning algorithms, and risk prediction combined with epidemiological models. In addition, health risk assessment also involves fields such as medical image analysis and gene data interpretation to improve the accuracy of early disease screening.

[0003] Among them, the neurosurgical risk prediction method based on big data analysis refers to a method for evaluating potential risks during neurosurgical operations using large-scale data analysis technology. It covers links such as patient medical history data extraction, preoperative physiological index analysis, and intraoperative monitoring data collection. By constructing a risk prediction model specific to neurosurgery, it evaluates possible postoperative complications, nerve injuries, etc. Specifically, statistical modeling is performed using historical case data, combined with intraoperative physiological signal analysis to identify key factors affecting postoperative recovery. At the same time, key variables are screened based on feature engineering techniques, and the risk probabilities under different surgical plans are calculated by training a risk prediction model. In addition, this method also uses pattern recognition technology to analyze the trends of intraoperative data and optimizes the risk prediction results by combining neural network calculation models, so as to provide data support for preoperative decision-making in neurosurgical operations.

[0004] The existing technology relies on statistical modeling in neurosurgical risk prediction, focusing on the trend analysis of historical case data and making it difficult to accurately capture the microscopic characteristics of individual preoperative neural activities. The preoperative risk assessment lacks the comprehensive quantification of the balance of nerve signal transmission and the oxygen metabolism state, resulting in a relatively rough setting of the postoperative recovery benchmark and making it difficult to accurately adapt to the individual recovery ability. The postoperative risk assessment mainly relies on intraoperative physiological signal monitoring, but lacks the tracking of the dynamic changes in the neural network connections after surgery, making the construction of the postoperative recovery model incomplete and affecting the accurate judgment of the recovery trend. The identification of abnormal regions relies on the determination of a single signal threshold and fails to comprehensively evaluate by combining the deviation of nerve conduction time and the consistency of oxygen metabolism recovery, reducing the accuracy of abnormal region screening. The assessment of individual nerve tolerance is only based on the preoperative fixed parameter setting, ignoring the nerve signal adjustment ability during the postoperative recovery process, resulting in a deviation in the prediction of the nerve load-bearing capacity after surgery. The risk level classification adopts a static standard, lacking real-time analysis of the nerve signal fluctuations during the postoperative recovery process, reducing the flexibility of risk assessment. The tolerance and risk level do not form a matching assessment system, restricting the accuracy of high-risk region screening and postoperative early warning and affecting the formulation of personalized intervention strategies. Summary of the Invention

[0005] To solve the technical problems existing in the prior art, an embodiment of the present invention provides a neurosurgical risk prediction method and system based on big data analysis. The technical solution is as follows:

[0006] A neurosurgical risk prediction method based on big data analysis includes the following steps:

[0007] S1: Obtain the nerve signal intensity of the brain region in the preoperative functional imaging data of the patient, analyze the balance of nerve signal transmission, and analyze the balance of oxygen metabolism distribution according to the nerve conduction rate to form preoperative nerve state information;

[0008] [[ID=!4]]S2: Call the nerve signal intensity in the preoperative nerve function state information as the benchmark state for postoperative recovery, detect the change trend of postoperative nerve signals, and analyze the oxygen metabolism adaptation ability of the brain region according to the change trend to obtain the characteristics of the postoperative recovery mode;

[0009] S3: Extract the neural network dynamic adjustment ability in the characteristics of the postoperative recovery mode, calculate the difference in the oxygen metabolism recovery speed of different brain regions, evaluate the tolerance range of the nerve load change during the postoperative recovery process, and obtain individual nerve tolerance information;

[0010] S4: According to the individual nerve tolerance information, calculate the deviation degree of the nerve recovery trend, and adjust the risk classification standard of the brain region according to the deviation degree to obtain the postoperative risk level distribution result;

[0011] S5: Combine the postoperative risk level distribution results and individual nerve tolerance information, and based on the matching degree between the tolerance and the risk level, screen the brain regions where the risk level exceeds the risk classification standard to obtain the postoperative risk prediction results.

[0012] As a further solution of the present invention, the preoperative nerve function state information includes signal transmission balance, oxygen metabolism utilization balance, and nerve activity intensity distribution; the postoperative recovery mode characteristics include signal change amplitude, metabolic adaptation ability, and recovery ability assessment; the individual nerve tolerance information includes signal transmission stability, time deviation distribution, oxygen metabolism recovery consistency, oxygen metabolism recovery speed, and nerve load bearing range; the postoperative risk level distribution results include recovery level division, risk level classification, recovery stability assessment, and risk classification standard adjustment; the postoperative risk prediction results include tolerance matching analysis, recovery deviation assessment, high-risk brain region screening, and risk warning range adjustment.

[0013] As a further solution of the present invention, the steps for obtaining the preoperative nerve function state information are as follows:

[0014] S101: Obtain the preoperative functional imaging data of the patient, extract the signal intensity of nerve activity in the brain region, calculate the signal delay and fluctuation amplitude in the signal transmission path according to the nerve conduction rate, calculate the signal fluctuation amplitude using the mean and standard deviation of the nerve activity signal, and calculate the signal delay based on the mean value of the signal delay and its maximum deviation value to obtain the signal fluctuation amplitude and delay value of the brain region.

[0015] S102: Based on the signal fluctuation amplitude and delay value of the brain region, calculate the signal transmission balance between different brain regions, calculate the signal balance using the variance of the signal fluctuation amplitude of each region, and combine the root mean square value of the signal delay to judge the deviation degree of signal transmission to obtain the signal transmission balance of the brain region.

[0016] S103: Call the oxygen metabolism rate data of the brain region, detect the regional distribution of the oxygen metabolism rate in different brain regions, calculate the balance of oxygen metabolism utilization based on the signal transmission balance of the brain region, calculate the oxygen metabolism utilization balance using the mean and standard deviation of the signal transmission balance and the oxygen metabolism rate, and use the formula:

[0017]

[0018] where E OM represents the oxygen metabolism utilization balance index, S i represents the signal transmission balance of the i-th brain region, O i represents the oxygen metabolism rate of the i-th brain region, n represents the total number of brain regions, represents the mean value of the oxygen metabolism rate of all brain regions. Represents the summation operation over all brain regions;

[0019] The operation obtains the oxygen metabolism utilization balance index. By combining the signal transmission balance degree of the brain region and the oxygen metabolism utilization balance index, the characteristics of neural activity signals are integrated to obtain the preoperative neural function state information.

[0020] As a further solution of the present invention, the steps for obtaining the characteristics of the postoperative recovery mode are as follows:

[0021] S201: Based on the preoperative neural function state information, the neural activity signal intensity is called as the benchmark state for postoperative recovery, the change trend of the postoperative neural signal is detected, the neural signal intensity at different postoperative time points is obtained, and the change amount of the neural signal intensity relative to the preoperative benchmark state is calculated. The signal change trend of the brain region is calculated using the neural signal change rate to obtain the postoperative neural signal change trend;

[0022] S202: Based on the postoperative neural signal change trend, the brain regions that exceed the preset neural signal change threshold during the postoperative recovery process are identified, the signal intensity change value at different postoperative time points is calculated, and it is judged whether the change value exceeds the set threshold. The formula is used:

[0023]

[0024] where R NS,i represents the relative change rate of the neural signal in the i-th brain region, A NS,i,t represents the neural signal intensity in the i-th brain region at time t, A NS,i,0 represents the preoperative benchmark neural signal intensity in the i-th brain region, |A NS,i,t -A NS,i,0 | represents the absolute change value of the neural signal intensity;

[0025] The operation obtains the relative change rate of each brain region, identifies the brain regions that exceed the preset neural signal change threshold, and obtains the postoperative abnormal neural signal regions;

[0026] S203: Based on the postoperative abnormal neural signal regions, combined with the preoperative oxygen metabolism balance, the metabolic adaptation ability of the brain regions that exceed the preset balance threshold within the specified postoperative period is analyzed, the deviation value of the oxygen metabolism rate at each postoperative time point is calculated, the metabolic fluctuation degree of each brain region within the postoperative period is judged, and the recovery ability of the brain region is evaluated based on the metabolic adaptation ability to obtain the characteristics of the postoperative recovery mode.

[0027] As a further solution of the present invention, the steps for obtaining the individual neural tolerance information are as follows:

[0028] S301: Based on the characteristics of the postoperative recovery pattern, extract the dynamic adjustment ability of neural network connections, call the neural signal transmission paths at different postoperative time points, calculate the degree of change between neural connection paths, calculate the path stability index using the time series of neural signal intensities, and obtain the neural network connection dynamic adjustment index;

[0029] S302: Based on the neural network connection dynamic adjustment index, compare the signal intensity range of preoperative neural activity signals with that during postoperative recovery, measure the deviation of the signal transmission time after surgery, calculate the change amplitude of the signal transmission time, calibrate the brain regions exceeding the set deviation threshold as abnormal regions, using the formula:

[0030]

[0031] where D ST,i represents the relative deviation value of the signal transmission time of the i-th brain region, T ST,i,t represents the signal transmission time of the i-th brain region at time t, T ST,i,0 represents the reference signal transmission time of the i-th brain region before surgery, |T ST,i,t -T ST,i,0 | represents the absolute change value of the signal transmission time;

[0032] Perform calculations to obtain the relative deviation values of the signal transmission times of each brain region, identify the abnormal regions exceeding the set deviation threshold, and obtain the postoperative signal transmission abnormal regions;

[0033] S303: Based on the postoperative signal transmission abnormal regions, measure the consistency of the postoperative oxygen metabolism recovery in the abnormal brain regions, analyze the differences in the oxygen metabolism recovery speeds of different brain regions, evaluate the tolerance range of the patient to neural load changes during the postoperative recovery process, calculate the postoperative neural function fitness, and obtain individual neural tolerance information.

[0034] As a further solution of the present invention, the steps for obtaining the postoperative risk level distribution result are:

[0035] S401: Based on the individual neural tolerance information, analyze the deviation degree of the postoperative neural activity recovery trend, call the reference value of the preoperative neural activity signal intensity, compare the neural signal data during the postoperative recovery stage, calculate the change amplitude of the recovery trend of each brain region, and compare the deviation degree between the actual recovery trajectory and the reference trajectory of the neural signal, to obtain the postoperative neural activity recovery deviation index;

[0036] S402: Based on the postoperative neural activity recovery deviation index, determine the recovery levels of different brain regions after surgery, divide the classification range of the postoperative risk levels, calculate the neural signal fluctuation data of each brain region, and evaluate the stability of the neural signals during the postoperative recovery period, using the formula:

[0037]

[0038] Among them, R NS,i represents the average value of the neural signal fluctuations in the i-th brain region, A NS,i,t represents the neural signal intensity in the i-th brain region at time t, A NS,i,0 represents the reference value of the neural signal intensity before surgery in the i-th brain region, |A NS,i,t -A NS,i,0 | represents the deviation value between the neural signal and the reference signal, and T represents the total duration of the postoperative recovery period;

[0039] Calculate the average value of the neural signal fluctuations in each brain region, set the threshold for risk level division, and judge the recovery stability of each brain region based on the average value of the signal fluctuations to obtain the risk level distribution of the postoperative brain regions;

[0040] S403: Based on the risk level distribution of the postoperative brain regions, call the neural signal fluctuation data during the postoperative recovery process, evaluate the stability of the brain regions with different tolerances during the recovery stage, adjust the risk classification criteria according to the changing trend of the recovery stability, calculate the risk fluctuation trend at different time points, adjust the classification range, and comprehensively consider the risk level changes in all time periods to obtain the postoperative risk level distribution result.

[0041] As a further solution of the present invention, the steps for obtaining the postoperative risk prediction result are as follows:

[0042] S501: Based on the postoperative risk level distribution result and the individual neural tolerance information, analyze the matching correlation between the tolerance and the risk level, call the individual neural tolerance data, calculate the brain region distribution under different tolerance levels, set the tolerance classification criteria, and compare with the postoperative risk level distribution to calculate the distribution ratio of different tolerance groups in each risk level interval to obtain the matching index between the tolerance and the risk level;

[0043] S502: Based on the matching index between the tolerance and the risk level, calculate the recovery deviation of the postoperative brain regions under different tolerance levels, call the postoperative recovery data, evaluate the signal change trend of each brain region during the recovery stage, calculate the deviation degree between the recovery trajectory and the standard recovery mode, statistically analyze the recovery deviation data of each brain region at different time points, and set the deviation threshold according to the postoperative risk level standard to screen the brain regions with recovery deviation exceeding the threshold to obtain the postoperative brain recovery deviation distribution;

[0044] S503: Based on the postoperative brain recovery deviation distribution, screen the brain regions with risk levels exceeding the risk classification criteria, adjust the risk warning range of the postoperative recovery state, calculate the risk warning values at different time points, optimize the risk classification threshold, using the formula:

[0045]

[0046] Among them, R FP,i represents the risk warning value of the i-th brain region, P HR,i,t represents the recovery offset value of the i-th brain region at time t, P TH represents the set recovery offset risk threshold, W HR,i represents the individual tolerance weighting factor of the i-th brain region, V HR,i,t represents the rate of change of the recovery speed of the i-th brain region at time t, |P HR,i,t -P TH | represents the absolute deviation between the recovery offset value and the risk threshold, represents the impact factor of the change in recovery speed on risk, and T represents the total duration of the postoperative recovery period;

[0047] Calculate the risk warning values of each brain region through operations, and optimize the risk assessment strategy by combining data from all time periods to finally obtain the postoperative risk prediction results.

[0048] A neurosurgical risk prediction system based on big data analysis, the system includes:

[0049] The preoperative nerve function assessment module obtains the preoperative functional imaging data of the patient, extracts the nerve activity signal intensity of the brain region, calculates the nerve signal transmission balance degree, detects the regional distribution of the oxygen metabolism rate, calculates the oxygen metabolism utilization balance, quantitatively analyzes the nerve function state of the brain region, and obtains the preoperative nerve function state information;

[0050] The postoperative recovery trend analysis module, based on the preoperative nerve function state information, calls the nerve activity signal intensity in the postoperative stage, analyzes the signal change trend, detects the brain regions exceeding the nerve signal change threshold, screens the brain regions with oxygen metabolism balance exceeding the limit during the postoperative recovery period, calculates the metabolic adaptation ability, and obtains the postoperative recovery mode characteristics;

[0051] The neural network tolerance assessment module, based on the postoperative recovery mode characteristics, analyzes the dynamic adjustment ability of the postoperative neural network connection, detects the stability of the postoperative nerve signal transmission path, calculates the brain regions with signal transmission time deviation exceeding the limit, determines the consistency of oxygen metabolism recovery in the abnormal brain regions, calculates the difference in oxygen metabolism recovery rates of different brain regions, and evaluates the range of the patient's postoperative nerve load tolerance to obtain the individual nerve tolerance information;

[0052] The postoperative risk grading module analyzes the deviation degree of the postoperative nerve activity recovery trend based on the individual nerve tolerance information, detects the recovery levels of different brain regions, calls the postoperative recovery mode parameters, screens the brain regions with abnormal changes in postoperative recovery stability, adjusts the risk level division criteria according to the recovery stability, and obtains the postoperative risk level distribution result;

[0053] The postoperative risk prediction module combines the postoperative risk level distribution result and the individual nerve tolerance information, analyzes the matching correlation between the tolerance and the risk level, calculates the recovery deviation under different tolerance levels, screens the brain regions with risk levels exceeding the threshold, adjusts the risk warning range of the postoperative recovery state, and obtains the postoperative risk prediction result.

[0054] The beneficial effects brought by the technical solutions provided in the embodiments of the present invention at least include:

[0055] In the present invention, by accurately extracting the preoperative nerve activity signals, analyzing the signal transmission delay, fluctuation amplitude, and oxygen metabolism balance, the accuracy of individual nerve function state assessment is improved. In the postoperative recovery stage, combined with the change trend of nerve signals, abnormal regions are dynamically identified, and the metabolic adaptation ability is evaluated, making the recovery mode more refined. The analysis of the neural network connection adjustment ability enhances the assessment of signal transmission stability, and combined with the time deviation and the consistency of oxygen metabolism recovery, the accuracy of abnormal region identification is improved. The analysis of the oxygen metabolism recovery speed of different brain regions makes the nerve load bearing range more personalized. Optimizing the recovery level division based on the deviation degree of the recovery trend makes the risk assessment more accurate. Adjusting the risk grading criteria by combining nerve signal fluctuations, improving the risk level matching degree, and optimizing the screening of high-risk regions make the risk warning more refined, providing a scientific basis for personalized postoperative management. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is the flowchart of the method of the present invention;

[0057] Figure 2 is the flowchart for obtaining the preoperative nerve state information of the present invention;

[0058] Figure 3 is the flowchart for obtaining the characteristics of the postoperative recovery mode of the present invention;

[0059] Figure 4 is the flowchart for obtaining the individual nerve tolerance information of the present invention;

[0060] Figure 5 is the flowchart for obtaining the postoperative risk level distribution result of the present invention;

[0061] Figure 6 is the flowchart for obtaining the postoperative risk prediction result of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] The technical solutions in the present invention will be described below in conjunction with the accompanying drawings.

[0063] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two can be selected.

[0064] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same.

[0065] In the embodiments of the present invention, sometimes subscripts such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0066] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail in conjunction with the accompanying drawings and specific embodiments.

[0067] Please refer to Figure 1 , the present invention provides a technical solution: a neurosurgical risk prediction method based on big data analysis, including the following steps:

[0068] S1: Obtain the preoperative functional imaging data of the patient, extract the signal intensity of the nerve activity in the brain region, analyze the signal delay and fluctuation amplitude of the signal transmission path based on the nerve conduction rate, calculate the signal transmission balance degree of different brain regions, call the oxygen metabolism rate of the brain region, detect the regional distribution of the oxygen metabolism rate of the brain region, calculate the balance of oxygen metabolism utilization, and form the preoperative nerve function state information;

[0069] S2: Call the nerve activity signal intensity in the preoperative nerve function state information as the reference state for postoperative recovery, detect the change trend of the postoperative nerve signal, identify the brain regions that exceed the preset nerve signal change threshold during the postoperative recovery process, and analyze the metabolic adaptation ability of the brain regions that exceed the preset balance threshold within a specified postoperative period based on the preoperative oxygen metabolism balance, evaluate the recovery ability of the brain regions, and obtain the postoperative recovery mode characteristics;

[0070] S3: Extract the dynamic adjustment ability of neural network connections in the postoperative recovery mode characteristics, analyze the stability of the postoperative nerve signal transmission path, calibrate the brain regions with signal transmission time deviation exceeding the set deviation threshold as abnormal regions based on the comparison of the signal intensity range during the postoperative recovery period with the preoperative nerve activity signal intensity, measure the consistency of oxygen metabolism recovery in the abnormal brain regions after surgery, analyze the differences in the oxygen metabolism recovery speed of different brain regions, evaluate the tolerance range of the patient to changes in nerve load during the postoperative recovery process, and obtain individual nerve tolerance information;

[0071] S4: According to the individual nerve tolerance information, analyze the deviation degree of the postoperative nerve activity recovery trend, determine the recovery levels of different brain regions after surgery, and divide the classification range of the postoperative risk levels. Call the nerve signal fluctuation data during the postoperative recovery process, evaluate the stability of different brain regions with different tolerances during the recovery stage, and adjust the risk classification criteria based on the changing trend of the recovery stability to obtain the postoperative risk level distribution result;

[0072] S5: Combine the postoperative risk level distribution result and the individual nerve tolerance information, analyze the matching correlation between the tolerance and the risk level, calculate the recovery deviation of the postoperative brain regions at different tolerance levels, screen the brain regions with risk levels exceeding the risk classification criteria based on the recovery results, and adjust the risk warning range of the postoperative recovery state to obtain the postoperative risk prediction result.

[0073] The preoperative nerve function state information includes signal transmission balance, oxygen metabolism utilization balance, and nerve activity intensity distribution; the postoperative recovery mode characteristics include signal change amplitude, metabolic adaptation ability, and recovery ability assessment; the individual nerve tolerance information includes signal transmission stability, time deviation distribution, oxygen metabolism recovery consistency, oxygen metabolism recovery speed, and nerve load tolerance range; the postoperative risk level distribution result includes recovery level division, risk level classification, recovery stability assessment, and risk classification criteria adjustment; the postoperative risk prediction result includes tolerance matching analysis, recovery deviation assessment, screening of high-risk brain regions, and risk warning range adjustment.

[0074] Please refer to Figure 2 , and the steps for obtaining the preoperative nerve function state information are as follows:

[0075] S101: Obtain the preoperative functional imaging data of the patient, extract the signal intensity of nerve activity in the brain regions, calculate the signal delay and fluctuation amplitude in the signal transmission path according to the nerve conduction rate, calculate the signal fluctuation amplitude using the mean and standard deviation of the nerve activity signals, and calculate the signal delay based on the mean of the signal delay and its maximum deviation value to obtain the signal fluctuation amplitude and delay value of the brain regions;

[0076] First, the functional imaging data of a patient can be collected by means such as functional magnetic resonance imaging (fMRI) or electroencephalogram (EEG). In fMRI, the blood oxygenation level-dependent contrast (BOLD) signal is used to characterize the intensity of neural activity in brain regions, while EEG reflects neural activity by measuring potential changes. Suppose the BOLD signal intensity of a specific brain region in a patient is 100 during fMRI scanning and the potential fluctuation range is from -70 μV to +90 μV during EEG detection. Then these two pieces of data can be used for subsequent signal transmission analysis respectively. Secondly, the calculation of the nerve conduction rate can be done by using the nerve fiber length and the known conduction velocity formula. Suppose the nerve fiber length is 50 mm and the conduction velocity is about 60 m / s. Then the signal transmission delay can be calculated as follows:

[0077]

[0078] In actual operation, the nerve fiber lengths of different brain regions are different, and the conduction velocities also vary due to different degrees of myelin sheath wrapping. Therefore, the signal delay of each brain region needs to be calculated after specific measurement through the nerve connection path. Moreover, the calculation of the signal fluctuation amplitude can use the standard deviation formula. If the BOLD signal sampling values of a specific brain region within a certain period are [90, 95, 100, 110, 105], then the fluctuation amplitude is calculated as follows:

[0079]

[0080] It shows that the fluctuation range of neural signals in this brain region is relatively large, and it is necessary to further analyze its impact on signal transmission. This result indicates that the signal fluctuation amplitude of the current brain region is relatively high, which may affect the stability of signal transmission. After comparison with other brain regions, it can be determined whether there are abnormal signal fluctuations in this brain region. Finally, the signal fluctuation amplitude and delay values of brain regions are obtained.

[0081] S102: Based on the signal fluctuation amplitude and delay values of brain regions, calculate the signal transmission balance degree between different brain regions. Calculate the signal balance degree by using the variance of the signal fluctuation amplitude of each region, and combine the root mean square value of the signal delay to judge the deviation degree of signal transmission, so as to obtain the signal transmission balance degree of brain regions;

[0082] For the signal delay values of all brain regions, the root mean square error (RMSD) can be calculated to measure its stability. Suppose the measured signal delay values of five brain regions are 0.83 ms, 1.2 ms, 0.95 ms, 1.05 ms, and 1.1 ms respectively. Then the calculation is as follows:

[0083]

[0084] 0.13 ms;

[0085] The calculation of the balance degree also needs to be combined with the variance analysis of the signal fluctuation amplitude. Suppose the signal fluctuation amplitudes of five brain regions are 7.07, 6.5, 8.0, 7.8, and 6.9 respectively, then the variance calculation is as follows:

[0086]

[0087] The smaller the variance value, the higher the signal balance degree. This result shows that the signal transmission balance degree in the current brain region is relatively high, which means that the neural signal transmission between different brain regions is relatively stable, ensuring the accuracy of subsequent functional analysis, and finally obtaining the signal transmission balance degree of the brain region.

[0088] S103: Call the oxygen metabolism rate data of the brain region, detect the regional distribution of the oxygen metabolism rate in different brain regions, calculate the balance of oxygen metabolism utilization based on the signal transmission balance degree of the brain region, and calculate the balance of oxygen metabolism utilization using the mean and standard deviation of the signal transmission balance degree and the oxygen metabolism rate. Use the formula:

[0089]

[0090] Among them, E OM represents the oxygen metabolism utilization balance index, S i represents the signal transmission balance degree of the i-th brain region, O i represents the oxygen metabolism rate of the i-th brain region, n represents the total number of brain regions, represents the mean of the oxygen metabolism rates of all brain regions, represents the summation operation for all brain regions;

[0091] Obtain the oxygen metabolism utilization balance index through the operation, combine the signal transmission balance degree of the brain region and the oxygen metabolism utilization balance index, integrate the neural activity signal characteristics, and obtain the preoperative neural function state information;

[0092] The oxygen metabolism rate is measured by PET scan. Taking the oxygen metabolism rates (unit: μmolO2 / g / min) of five brain regions as an example: 4.5, 5.0, 4.8, 5.2, 4.6, calculate the balance of oxygen metabolism utilization in combination with the signal transmission balance degree. Use the formula:

[0093] [[ID=3�]]

[0094] Among them, the values of each parameter are as follows:

[0095] Signal transmission balance degree: S i are 0.9, 0.85, 0.88, 0.92, and 0.87 respectively;

[0096] Oxygen metabolism rate: O iThey are 4.5, 5.0, 4.8, 5.2, 4.6 respectively;

[0097] Mean value of oxygen metabolism:

[0098] Substitute into the calculation:

[0099]

[0100] Finally:

[0101] E OM = 4.26 - 0.255 = 4.005;

[0102] The results show that the oxygen metabolism utilization balance index of the current brain region is 4.005, indicating that the distribution of oxygen metabolism utilization among brain regions is relatively uniform, meaning that neural activities are relatively coordinated. If this value is too low, it indicates that there may be abnormal oxygen metabolism in some brain regions. This value, combined with the regional distribution of the signal transmission balance and oxygen metabolism rate in the brain region, can be further integrated into the complete preoperative nerve function status information, and finally the preoperative nerve function status information is obtained.

[0103] Please refer to Figure 3 , and the steps for obtaining the characteristics of the postoperative recovery mode are as follows:

[0104] S201: Based on the preoperative nerve function status information, call the neural activity signal intensity as the baseline state of postoperative recovery, detect the change trend of postoperative nerve signals, obtain the neural signal intensity at different postoperative time points, and calculate the change amount of the neural signal intensity relative to the preoperative baseline state. Use the neural signal change rate to calculate the signal change trend of the brain region to obtain the postoperative nerve signal change trend;

[0105] The preoperative nerve function status information contains the neural activity signal intensity of different brain regions, and this signal intensity can be measured by functional magnetic resonance imaging (fMRI) or electroencephalogram (EEG). Suppose the BOLD signal intensity of a specific brain region of a patient in the preoperative fMRI scan is 120, and the average potential signal measured by EEG is 85 μV, then these data can be used as the preoperative baseline state. Secondly, obtain the neural signal intensity at different postoperative time points. The acquisition method of postoperative data is the same as that of preoperative data. Suppose on the 1st, 3rd, and 7th days after surgery, the BOLD signal intensities of this brain region are 115, 110, and 125 respectively, and the EEG potential averages are 83 μV, 79 μV, and 90 μV respectively. Then these data are used for subsequent calculation of the change trend. Furthermore, calculate the change amount of the neural signal relative to the preoperative baseline state. Use the difference to calculate the change amount. That is, the relative change amount of the postoperative BOLD signal can be expressed as:

[0106] ΔS t = St -S0;

[0107] Substituting specific values, on the 3rd day after surgery:

[0108] ΔS3 = 110 - 120 = -10;

[0109] This indicates a decrease in neural activity in this brain region. And on the 7th day after surgery:

[0110] ΔS7 = 125 - 120 = 5;

[0111] It represents a recovery of neural activity. Finally, the change trend of neural signals in each brain region is calculated using the change rate, and the calculation formula is as follows:

[0112]

[0113] On the 3rd day after surgery:

[0114]

[0115] On the 7th day after surgery:

[0116]

[0117] The results show that the neural signal in this brain region shows a downward trend on the 3rd day after surgery and rebounds on the 7th day after surgery. This indicates that there are certain phased changes in the neural recovery process, and this trend can be further used to evaluate the recovery status of the entire brain region, and finally the change trend of postoperative neural signals is obtained.

[0118] S202: Based on the change trend of postoperative neural signals, identify the brain regions that exceed the preset neural signal change threshold during the postoperative recovery process, calculate the change value of the signal intensity at different postoperative time points, and determine whether the change value exceeds the set threshold. The formula is:

[0119]

[0120] Where R NS,i represents the relative change rate of the neural signal in the i-th brain region, A NS,i,t represents the neural signal intensity in the i-th brain region at time t, A NS,i,0 represents the preoperative reference neural signal intensity in the i-th brain region, |A NS,i,t -A NS,i,0 | represents the absolute change value of the neural signal intensity;

[0121] Through calculation, obtain the relative change rate of each brain region, identify the brain regions that exceed the preset neural signal change threshold, and obtain the postoperative abnormal neural signal regions;

[0122] First, set the threshold for nerve signal change. Based on the statistical data of nerve signal changes during the normal postoperative recovery process, set a reasonable threshold range. For example, a change rate exceeding ±10% may indicate abnormality. Calculate the signal intensity change values at different postoperative time points, and evaluate whether it exceeds the threshold by calculating the relative change rate. Use the formula:

[0123]

[0124] Suppose the preoperative signal intensity of a certain brain region is A NS,i,0 = 100, and the postoperative signal intensity is A NS,i,3 = 85. Substitute into the calculation:

[0125]

[0126] If the set threshold is 10%, then the relative change rate of this brain region exceeds the set threshold and should be identified as an abnormal region. Similarly, for the 7th day after surgery, if the measured signal intensity is 95, the calculation is as follows:

[0127]

[0128] This value is within the threshold range and is not identified as an abnormal region. This result indicates that on the 3rd day after surgery, the nerve signal change rate of this brain region exceeds the threshold, indicating that there may be abnormal recovery in this brain region. On the 7th day after surgery, its change rate returns to the normal range, indicating that the nerve function recovery may have improved. Further combining the data of other time points, it is possible to comprehensively evaluate whether there are fluctuating abnormalities or persistent abnormal brain regions during the entire recovery process, and finally obtain the postoperative abnormal nerve signal regions.

[0129] S203: Based on the postoperative abnormal nerve signal regions, combined with the preoperative oxygen metabolism balance, analyze the metabolic adaptation ability of the brain regions that exceed the preset balance threshold within the specified postoperative period. Calculate the deviation values of the oxygen metabolism rates at each postoperative time point, judge the metabolic fluctuation degree of each brain region during the postoperative period, and evaluate the recovery ability of the brain regions based on the metabolic adaptation ability to obtain the postoperative recovery mode characteristics;

[0130] Obtain the oxygen metabolism balance data of each brain region before surgery, which is measured by positron emission tomography (PET). Suppose the average oxygen metabolism of a certain patient's brain region before surgery is 4.8 μmolO2 / g / min. Then, obtain the oxygen metabolism rates at different time points after surgery. Suppose the oxygen metabolism rates on the 3rd and 7th days after surgery are 4.2 and 5.1 μmolO2 / g / min respectively. Furthermore, calculate the oxygen metabolism deviation value after surgery to evaluate the metabolic fluctuation situation. If the set balance threshold is 0.2, then the oxygen metabolism deviation value on the 3rd day after surgery exceeds the threshold, indicating that the metabolic adaptation ability of this brain region is low. While the deviation value on the 7th day after surgery is less than the threshold, indicating that the metabolic adaptation ability of this brain region tends to be stable. This result shows that in the early stage after surgery (the 3rd day), abnormal metabolic fluctuations may occur in some brain regions due to insufficient metabolic adaptation ability. While on the 7th day after surgery, the metabolic state tends to be stable, indicating that the recovery ability of this brain region is good. Further combining with the trend of nerve signal changes, the postoperative recovery mode can be comprehensively evaluated, and finally the characteristics of the postoperative recovery mode can be obtained.

[0131] Please refer to Figure 4 , the steps for obtaining individual nerve tolerance information are as follows:

[0132] S301: Based on the characteristics of the postoperative recovery mode, extract the dynamic adjustment ability of neural network connections, call the nerve signal transmission paths at different time points after surgery, calculate the degree of change between nerve connection paths, and use the time series of nerve signal intensity to calculate the path stability index to obtain the dynamic adjustment index of neural network connections;

[0133] Obtain the neural signal transmission paths of the patient at different time points after surgery. This path is obtained by comparing the pre- and post-operative fMRI measurement data. fMRI can be used to measure the functional connection strength between brain regions and record the changes in the neural network at multiple time points before and after surgery in the form of time series data. For example, assume that the main neural connection path of a certain patient before surgery is A→B→C→D, and the paths on the 1st day, 5th day, and 10th day after surgery are A→B→C→D, A→C→D, and A→B→D respectively. It can be observed that there are partial losses or reorganizations of neural connections. Secondly, calculate the degree of change between neural connection paths, and measure the neural network adjustment ability with path stability. The changes in neural connections can be reflected in the number of path accesses, the signal strength between nodes, and the continuity of connections. For example, if the signal strength of A→B before surgery is 0.75, and the signal strength of A→C measured on the 5th day after surgery drops to 0.6, it indicates that the adaptability of this neural path is weak. Furthermore, by statistically analyzing the signal change trends at different post-operative periods, the stability of the dynamic adjustment of neural network connections can be analyzed. Assume that the neural signal strengths of a certain patient at different time points after surgery are 0.75, 0.7, 0.6, 0.65, and 0.7 respectively. It can be calculated that the fluctuation range of neural connections is relatively large, indicating that during the post-operative recovery process, the neural network needs to perform strong self-adjustment. Further combining the data at different time points, the dynamic adjustment ability of neural network connections can be comprehensively evaluated, and finally the dynamic adjustment index of neural network connections can be obtained.

[0134] S302: Based on the dynamic adjustment index of neural network connections, compare the signal strength of pre-operative neural activity with the signal strength change range during post-operative recovery, determine the deviation of the post-operative signal transmission time, calculate the change amplitude of the signal transmission time, and mark the brain regions exceeding the set deviation threshold as abnormal regions. Use the formula:

[0135]

[0136] where D ST,i represents the relative deviation value of the signal transmission time of the i-th brain region, T ST,i,t represents the signal transmission time of the i-th brain region at time t, T ST,i,0 represents the reference signal transmission time of the i-th brain region before surgery, |T ST,i,t -T ST,i,0 | represents the absolute change value of the signal transmission time;

[0137] Calculate the relative deviation values of the signal transmission times of each brain region through operations, identify the abnormal regions exceeding the set deviation threshold, and obtain the post-operative signal transmission abnormal regions;

[0138] "Obtain the nerve signal transmission times in different brain regions before surgery. Suppose a patient's signal transmission time from region A to region B is measured as 8.5 ms before surgery, and from region B to region C is 10.2 ms. Then these values are used as a baseline. Secondly, obtain the signal transmission times at different time points after surgery. Suppose the signal transmission time from region A to region B is measured as 9.1 ms and from region B to region C is 11.0 ms on the 3rd day after surgery. Then calculate the change amplitude of the signal transmission time after surgery, which is expressed by the normalization formula:

[0139]

[0140] Substitute the data:

[0141]

[0142]

[0143] Furthermore, set the deviation threshold of the signal transmission time. For example, set the threshold to 10%. Then the change amplitudes of both A→B and B→C do not exceed the threshold, so they are not judged as abnormal. Suppose the transmission time of A→B is measured as 10.5 ms for a certain patient on the 7th day after surgery. Then calculate:

[0144]

[0145] This value exceeds the threshold range. Therefore, the A→B path is marked as the abnormal region of signal transmission after surgery. This result indicates that during the postoperative recovery process of this patient, the change in the signal transmission time of some brain regions exceeds the normal range, suggesting that there may be postoperative functional abnormalities, and finally the abnormal region of signal transmission after surgery is obtained.

[0146] S303: Based on the abnormal region of signal transmission after surgery, determine the consistency of the postoperative oxygen metabolism recovery in the abnormal brain region, analyze the differences in the oxygen metabolism recovery speeds of different brain regions, evaluate the tolerance range of the patient to the change in nerve load during the postoperative recovery process, calculate the postoperative nerve function adaptability, and obtain the individual nerve tolerance information;

[0147] Obtain the oxygen metabolism balance data of the abnormal area before surgery, which is measured by positron emission tomography (PET), and record the oxygen metabolism rate of the brain area before surgery. For example, if the oxygen metabolism rate of the A→B path before surgery is 4.8 μmolO2 / g / min, this value is used as the benchmark. Secondly, obtain the oxygen metabolism rates at different time points after surgery. Suppose the oxygen metabolism rate measured on the 3rd day after surgery is 4.2 μmolO2 / g / min, and on the 7th day it is 4.9 μmolO2 / g / min. Then, the recovery trend of oxygen metabolism can be compared. If the oxygen metabolism rate in the early postoperative period (such as the 3rd day) is significantly lower than the preoperative benchmark value, it may indicate that the recovery process of this brain area is slower. If on the 7th day after surgery, the oxygen metabolism rate rises to 4.9 μmolO2 / g / min, it indicates that the metabolic function of the brain area tends to recover. Furthermore, analyze the differences in the oxygen metabolism recovery speeds of different brain regions. Suppose the oxygen metabolism rate of area A is 4.2 μmolO2 / g / min and that of area B is 4.6 μmolO2 / g / min on the 3rd day after surgery for the patient. Then, it shows that the metabolic recovery speed of area B is faster than that of area A. By further comparing the metabolic recovery trends of different brain regions at different time points after surgery, the tolerance range of the patient to changes in neural load during the postoperative recovery process can be evaluated. For example, if the oxygen metabolism rate has not recovered to the benchmark value on the 14th day after surgery, it means that the patient's tolerance may be low. If the oxygen metabolism rate remains stable after the 10th day, it can be inferred that the patient has a strong neural adaptation ability. This result shows that there are individual differences in the neural tolerance of different patients, and personalized evaluation can be carried out by combining the data at different time points to finally obtain individual neural tolerance information.

[0148] Please refer to Figure 5 , and the steps for obtaining the postoperative risk level distribution result are as follows:

[0149] S401: Based on the individual neural tolerance information, analyze the deviation degree of the postoperative neural activity recovery trend, call the preoperative neural activity signal intensity benchmark value, compare the neural signal data in the postoperative recovery stage, calculate the change range of the recovery trend of each brain region, and compare the deviation degree between the actual recovery trajectory and the benchmark trajectory of the neural signal to obtain the postoperative neural activity recovery deviation index;

[0150] First, obtain the baseline value of the patient's preoperative neural activity signal intensity, which is usually measured by functional magnetic resonance imaging (fMRI) or electroencephalogram (EEG). Suppose the average EEG signal of a specific brain region of a patient before surgery is 85 μV, and the BOLD signal intensity measured by fMRI is 120. These values serve as the baseline state of neural activity. Secondly, obtain the neural activity data at different time points after surgery. Suppose on the 1st, 5th, and 10th days after surgery, the EEG signals of this brain region are 83 μV, 79 μV, and 90 μV respectively, and the BOLD signals are 115, 110, and 125 respectively. These data are used to calculate the recovery trend. Furthermore, calculate the deviation index of neural signal recovery. Compare the baseline signal with the signal data during the postoperative recovery process, and use the difference to calculate the deviation degree of the postoperative recovery trajectory. For example, if the EEG signal drops to 79 μV on the 5th day after surgery, then relative to the baseline value of 85 μV, the deviation degree is: ΔS = 79 - 85 = -6. If the set normal recovery range threshold is ±5 μV, then this value exceeds the normal range, indicating that the recovery progress of this brain region is deviated. Finally, combine the data of multiple time points, compare the deviation between the postoperative recovery trend and the baseline trend, and obtain the deviation index of postoperative neural activity recovery.

[0151] S402: Based on the deviation index of postoperative neural activity recovery, determine the recovery levels of different brain regions after surgery and divide the classification range of postoperative risk levels. Calculate the neural signal fluctuation data of each brain region and evaluate the stability of neural signals during the postoperative recovery period. Use the formula:

[0152]

[0153] where, R NS,i represents the average neural signal fluctuation of the i-th brain region, A NS,i,t represents the neural signal intensity of the i-th brain region at time t, A NS,i,0 represents the baseline value of the neural signal intensity of the i-th brain region before surgery, |A NS,i,t -A NS,i,0 | represents the deviation value between the neural signal and the baseline signal, and T represents the total duration of the postoperative recovery period;

[0154] Calculate the average neural signal fluctuation of each brain region through the operation, and set the threshold for dividing the risk level. Judge the recovery stability of each brain region based on the average signal fluctuation to obtain the postoperative brain region risk level distribution;

[0155] Obtain the neural signal fluctuation data during the postoperative recovery process, which is obtained by continuously monitoring the EEG or fMRI signals. Suppose the signal fluctuation amplitudes at different time points after surgery are 3 μV, 7 μV, and 2 μV respectively. Then calculate the average fluctuation and set the threshold for dividing the risk level. Use the following formula to calculate the average neural signal fluctuation:

[0156]

[0157] Suppose the preoperative signal intensity reference value of a certain patient is 100, and the signal intensities on the 1st, 5th, and 10th days after surgery are 95, 110, and 105 respectively. Then calculate its fluctuation mean value:

[0158]

[0159] Furthermore, set the risk level classification standard. Suppose the fluctuation mean value is in the range of 0 - 5 is low risk, 5 - 10 is medium risk, and above 10 is high risk. Then the risk level of this patient is medium risk. This result indicates that the fluctuation amplitude of the postoperative nerve signal can be used to evaluate the recovery level of different brain regions and divide the risk level range, and finally obtain the postoperative brain region risk level distribution.

[0160] S403: Based on the postoperative brain region risk level distribution, call the nerve signal fluctuation data during the postoperative recovery process, evaluate the stability of brain regions with different tolerances during the recovery stage, adjust the risk classification standard according to the changing trend of the recovery stability, calculate the risk fluctuation trend at different time points, adjust the classification range, and comprehensively consider the risk level changes in all time periods to obtain the postoperative risk level distribution result;

[0161] Obtain the nerve signal fluctuation conditions at different postoperative time points. Suppose the signal fluctuations of a specific brain region measured on the 3rd and 7th days after surgery for a certain patient are 8 μV and 3 μV respectively. Then it indicates that the recovery stability of this brain region is gradually improving. Secondly, analyze the changing trend of the recovery stability and calculate the risk fluctuation trend at different time points. Suppose the fluctuation index on the 3rd day is 0.15 and it drops to 0.08 on the 7th day. Then it indicates that the recovery is tending to be stable. Furthermore, adjust the risk classification standard, and optimize the risk classification range according to the risk level changes in different time periods. For example, if the originally set low - risk range is 0 - 5 μV, and it is observed that the fluctuations of most patients are lower than 3 μV in the later stage of recovery, then the low - risk standard can be adjusted to 0 - 3 μV to improve the classification accuracy. This result indicates that by analyzing the trend changes of the recovery stability, the postoperative risk level classification standard can be optimized, and finally the postoperative risk level distribution result can be obtained.

[0162] Please refer to Figure 6 , the steps to obtain the postoperative risk prediction result are:

[0163] S501: Based on the postoperative risk level distribution results and individual nerve tolerance information, analyze the matching correlation between tolerance and risk level, call the individual nerve tolerance data, calculate the brain region distribution under different tolerance levels, set the tolerance grading standard, compare with the postoperative risk level distribution, calculate the distribution ratio of different tolerance groups in each risk level interval, and obtain the tolerance and risk level matching index;

[0164] Obtain the recovery status of different brain regions of the patient after surgery, and extract the preoperative individual nerve tolerance information. Tolerance is defined as the adaptability of different brain regions of the patient to nerve load changes in the preoperative state, and this value is calculated through the preoperative nerve signal fluctuation amplitude and preoperative nerve activity balance. Then, call the postoperative recovery data, obtain the change in nerve signal intensity in each brain region, calculate the recovery trend of the postoperative nerve signal, and evaluate the nerve recovery ability of each region according to the tolerance information. For brain regions with higher tolerance, if the postoperative recovery trend is stable, it can be determined that the recovery is good; if the postoperative recovery trend deviates from the benchmark, it indicates that the region is more severely damaged. For brain regions with lower tolerance, a more strict assessment of their postoperative recovery status is required. Suppose a patient's left frontal lobe tolerance is 85 and the right temporal lobe tolerance is 60. On the 30th day after surgery, the nerve signal of the left frontal lobe has recovered to 92% of the benchmark value, while the right temporal lobe has only recovered to 75% of the benchmark value. It can be determined that the recovery ability of the right temporal lobe is lower. By calculating the recovery deviation of different brain regions at different tolerance levels, a tolerance and risk level matching index is formed.

[0165] S502: Based on the tolerance and risk level matching index, calculate the recovery deviation of the postoperative brain regions at different tolerance levels, call the postoperative recovery data, evaluate the signal change trend of each brain region during the recovery stage, calculate the deviation degree between the recovery trajectory and the standard recovery mode, count the recovery deviation data of each brain region at different time points, and set the deviation threshold according to the postoperative risk level standard. Screen the brain regions with recovery deviation exceeding the threshold to obtain the postoperative brain recovery deviation distribution;

[0166] Obtain the changes in the neural signal intensity of each brain region, calculate its deviation from the preoperative reference signal, compare with the law of neural signal changes in the normal recovery mode, extract the neural signal trajectory in the postoperative recovery stage of the patient, analyze the degree of deviation of the trajectory from the standard recovery mode, calculate the signal recovery deviation in the three stages of the initial, middle, and late recovery periods respectively. For example, in the initial recovery period (days 0 - 30), if the signal intensity of a certain brain region decreases by 20% compared to the preoperative reference, while in the normal recovery mode, this brain region only decreases by 10%, then the recovery offset value is 10%. Similarly, in the late recovery period (after 60 days), if the signal of a certain brain region recovers to 90% of the reference value, while in the normal mode it should recover to 95%, then the recovery offset value is 5%. Through inductive analysis of the data in each stage, screen the brain regions with recovery offset exceeding the set threshold, and count their distribution at each risk level, and finally obtain the postoperative brain recovery offset distribution.

[0167] S503: Based on the postoperative brain recovery offset distribution, screen the brain regions with risk levels exceeding the risk classification standard, adjust the risk warning range of the postoperative recovery status, calculate the risk warning values at different time points, optimize the risk classification threshold, using the formula:

[0168]

[0169] where, R FP,i represents the risk warning value of the i-th brain region, P HR,i,t represents the recovery offset value of the i-th brain region at time t, P TH represents the set recovery offset risk threshold, W HR,i represents the individual tolerance weighting factor of the i-th brain region, V HR,i,t represents the recovery speed change rate of the i-th brain region at time t, |P HR,i,t -P TH | represents the absolute deviation between the recovery offset value and the risk threshold, represents the influence factor of the recovery speed change on the risk, T represents the total duration of the postoperative recovery period;

[0170] Perform operations to obtain the risk warning values of each brain region, and optimize the risk assessment strategy by combining the data of all time periods, and finally obtain the postoperative risk prediction results;

[0171] Suppose the postoperative recovery offset value of a certain brain region is 15% on the 10th day, 18% on the 20th day, and 12% on the 30th day. According to the preoperative neural tolerance assessment, the calculation method is the normalized result of the tolerance score. That is, if the tolerance score is between 0 - 100 and the brain region tolerance score is 80 (full score 100), then the weighting factor W HR,i is calculated as follows: In the normal recovery mode, the statistical data of postoperative recovery deviation in different brain regions are based on the preoperative nerve tolerance and postoperative recovery speed. The normal deviation should be controlled within ±10%. Therefore, the recovery deviation risk threshold P TH is set at 10%. The change rates of the recovery speed are 2.5 on the 10th day, 2.0 on the 20th day, and 1.5 on the 30th day. Substitute these values into the formula for calculation:

[0172] For the 10th day:

[0173]

[0174] For the 20th day:

[0175]

[0176] For the 30th day:

[0177]

[0178] Accumulate the risk values at all time points:

[0179] R FP,i = 2.003 + 3.356 + 0.884 = 6.243;

[0180] Associate the calculation result with the risk prediction. The obtained risk warning value R FP,i = 6.243. If the postoperative risk warning threshold is set at 6, this value is obtained from the analysis of the recovery data of high-risk brain regions. Exceeding this value means that the postoperative risk may increase, and special attention should be paid to the recovery of this brain region. Therefore, the risk level of this brain region is close to the high-risk area, and the recovery strategy needs to be further adjusted. Otherwise, the risk of postoperative complications may increase. This result indicates that the recovery status of this brain region deviates from the normal recovery trajectory, and its tolerance level is not sufficient to support the current recovery process, thus leading to an increase in postoperative recovery risk. This numerical result is directly related to the determination of postoperative risk prediction. If this value is significantly higher than the threshold, it indicates that the patient needs to receive more strict postoperative monitoring and rehabilitation intervention. Finally, this risk warning value is used to adjust the postoperative risk prediction range to obtain the postoperative risk prediction result.

[0181] A neurosurgical risk prediction system based on big data analysis, the system includes:

[0182] The preoperative nerve function evaluation module obtains the preoperative functional imaging data of the patient, extracts the nerve activity signal intensity of the brain region, calculates the nerve signal transmission balance degree, detects the regional distribution of the oxygen metabolism rate, calculates the oxygen metabolism utilization balance, and quantitatively analyzes the nerve function state of the brain region to obtain the preoperative nerve function state information;

[0183] The postoperative recovery trend analysis module, based on the preoperative neurological function status information, calls the neural activity signal intensity in the postoperative stage, analyzes the signal change trend, detects the brain regions exceeding the neural signal change threshold, screens the brain regions with abnormal oxygen metabolism balance during the postoperative recovery period, calculates the metabolic adaptability, and obtains the characteristics of the postoperative recovery pattern;

[0184] The neural network tolerance evaluation module, based on the characteristics of the postoperative recovery pattern, analyzes the dynamic adjustment ability of the postoperative neural network connection, detects the stability of the postoperative neural signal transmission path, calculates the brain regions with excessive signal transmission time deviation, determines the consistency of oxygen metabolism recovery in the abnormal brain regions, calculates the difference in oxygen metabolism recovery rate in different brain regions, evaluates the range of the patient's postoperative neural load tolerance, and obtains the individual neural tolerance information;

[0185] The postoperative risk grading module, based on the individual neural tolerance information, analyzes the deviation degree of the postoperative neural activity recovery trend, detects the recovery levels of different brain regions, calls the postoperative recovery pattern parameters, screens the brain regions with abnormal changes in postoperative recovery stability, adjusts the risk level classification standard according to the recovery stability, and obtains the postoperative risk level distribution result;

[0186] The postoperative risk prediction module combines the postoperative risk level distribution result and the individual neural tolerance information, analyzes the matching correlation between the tolerance and the risk level, calculates the recovery deviation under different tolerance levels, screens the brain regions with risk levels exceeding the threshold, adjusts the risk warning range of the postoperative recovery state, and obtains the postoperative risk prediction result.

[0187] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A neurosurgical risk prediction method based on big data analysis, characterized in that, It includes the following steps: S1: Obtain the neural signal intensity of the brain region in the preoperative functional imaging data of the patient, analyze the balance of neural signal transmission, analyze the balance of oxygen metabolism distribution based on the neural conduction rate, and form the preoperative neural state information; S2: Invoke the neural signal intensity in the preoperative neural function state information as the reference state for postoperative recovery, detect the changing trend of postoperative neural signals, analyze the oxygen metabolism adaptation ability of the brain region based on the changing trend, and obtain the characteristics of the postoperative recovery mode; S3: Extract the dynamic adjustment ability of the neural network in the characteristics of the postoperative recovery mode, calculate the difference in the oxygen metabolism recovery speed of different brain regions, evaluate the tolerance range of neural load changes during the postoperative recovery process, and obtain the individual neural tolerance information; S4: According to the individual neural tolerance information, calculate the deviation degree of the neural recovery trend, adjust the risk classification standard of the brain region based on the deviation degree, and obtain the postoperative risk level distribution result; S5: Combine the postoperative risk level distribution result and the individual neural tolerance information, and screen the brain regions with risk levels exceeding the risk classification standard based on the matching degree between the tolerance and the risk level, to obtain the postoperative risk prediction result.

2. The method for predicting neurosurgical risks based on big data analysis according to claim 1, wherein: The preoperative neural function state information includes signal transmission balance, oxygen metabolism utilization balance, and neural activity intensity distribution; the characteristics of the postoperative recovery mode include signal change amplitude, metabolic adaptation ability, and recovery ability evaluation; the individual neural tolerance information includes signal transmission stability, time deviation distribution, oxygen metabolism recovery consistency, oxygen metabolism recovery speed, and neural load tolerance range; the postoperative risk level distribution result includes recovery level division, risk level classification, recovery stability evaluation, and risk classification standard adjustment; the postoperative risk prediction result includes tolerance matching analysis, recovery deviation evaluation, high-risk brain region screening, and risk warning range adjustment.

3. The neurosurgical risk prediction method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the preoperative neural function state information are as follows: S101: Obtain the preoperative functional imaging data of the patient, extract the signal intensity of neural activity in the brain region, calculate the signal delay and fluctuation amplitude in the signal transmission path according to the neural conduction rate, calculate the signal fluctuation amplitude using the mean and standard deviation of the neural activity signal, and calculate the signal delay based on the mean value of the signal delay and its maximum deviation value, to obtain the signal fluctuation amplitude and delay value of the brain region; S102: Based on the signal fluctuation amplitude and delay value of the brain region, calculate the signal transmission balance between different brain regions, calculate the signal balance using the variance of the signal fluctuation amplitude of each region, and combine the root mean square value of the signal delay to judge the deviation degree of signal transmission, to obtain the signal transmission balance of the brain region; S103: Invoke the oxygen metabolism rate data of the brain region, detect the regional distribution of the oxygen metabolism rate in different brain regions, calculate the balance of oxygen metabolism utilization based on the signal transmission balance of the brain region, calculate the oxygen metabolism utilization balance using the mean and standard deviation of the signal transmission balance and the oxygen metabolism rate, using the formula: Among them, E OM represents the oxygen metabolism utilization balance index, S i represents the signal transmission balance degree of the i-th brain region, O i represents the oxygen metabolism rate of the i-th brain region, n represents the total number of brain regions, represents the average value of the oxygen metabolism rates of all brain regions, represents the summation operation for all brain regions; The operation obtains the oxygen metabolism utilization balance index. By combining the signal transmission balance degree of the brain region with the oxygen metabolism utilization balance index, the characteristics of neural activity signals are integrated to obtain the preoperative neural function state information.

4. The neurosurgical risk prediction method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the characteristics of the postoperative recovery mode are as follows: S201: Based on the preoperative neural function state information, the neural activity signal intensity is called as the reference state for postoperative recovery. The change trend of the postoperative neural signal is detected, the neural signal intensity at different postoperative time points is obtained, and the change amount of the neural signal intensity relative to the preoperative reference state is calculated. The change trend of the signal in the brain region is calculated using the neural signal change rate to obtain the postoperative neural signal change trend; S202: Based on the postoperative neural signal change trend, the brain regions exceeding the preset neural signal change threshold during the postoperative recovery process are identified. The signal intensity change value at different postoperative time points is calculated, and it is judged whether the change value exceeds the set threshold. The formula is used: Among them, R NS,i represents the relative change rate of the neural signal of the i-th brain region, A NS,i,t represents the neural signal intensity of the i-th brain region at time t, A NS,i,0 represents the preoperative reference neural signal intensity of the i-th brain region, |A NS,i,t -A NS,i,0 | represents the absolute change value of the neural signal intensity; The operation obtains the relative change rate of each brain region, identifies the brain regions exceeding the preset neural signal change threshold, and obtains the postoperative abnormal neural signal regions; S203: Based on the postoperative abnormal neural signal regions, combined with the preoperative oxygen metabolism balance, the metabolic adaptation ability of the brain regions exceeding the preset balance threshold within the specified postoperative period is analyzed. The deviation value of the oxygen metabolism rate at each postoperative time point is calculated, and the metabolic fluctuation degree of each brain region within the postoperative period is judged. The recovery ability of the brain region is evaluated based on the metabolic adaptation ability to obtain the characteristics of the postoperative recovery mode.

5. The neurosurgical risk prediction method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the individual neural tolerance information are as follows: S301: Based on the characteristics of the postoperative recovery mode, the dynamic adjustment ability of the neural network connection is extracted. The neural signal transmission paths at different postoperative time points are called, and the change degree between the neural connection paths is calculated. The path stability index is calculated using the time series of the neural signal intensity to obtain the neural network connection dynamic adjustment index; S302: Based on the neural network connection dynamic adjustment index, the change range of the signal intensity between the preoperative neural activity signal intensity and the postoperative recovery period is compared. The deviation of the postoperative signal transmission time is measured, the change amplitude of the signal transmission time is calculated, and the brain regions exceeding the set deviation threshold are marked as abnormal regions. The formula is used: Among them, D ST,i represents the relative deviation value of the signal transmission time of the i-th brain region, T ST,i,t represents the signal transmission time of the i-th brain region at time t, T ST,i,0 represents the reference signal transmission time of the i-th brain region before surgery, |T ST,i,t -T ST,i,0 | represents the absolute change value of the signal transmission time; The operation obtains the relative deviation value of the signal transmission time of each brain region, identifies the abnormal regions exceeding the set deviation threshold, and obtains the postoperative signal transmission abnormal regions; S303: Based on the postoperative signal transmission abnormal regions, the consistency of the postoperative oxygen metabolism recovery in the brain abnormal regions is measured. The differences in the oxygen metabolism recovery speed of different brain regions are analyzed. The tolerance range of the patient to the change in neural load during the postoperative recovery process is evaluated. The postoperative neural function fitness is calculated to obtain the individual neural tolerance information.

6. The neurosurgical risk prediction method based on big data analysis according to claim 1, characterized in that: The steps for obtaining the postoperative risk level distribution result are as follows: S401: Based on the individual nerve tolerance information, analyze the deviation degree of the postoperative nerve activity recovery trend, call the preoperative nerve activity signal intensity reference value, compare the nerve signal data in the postoperative recovery stage, calculate the change amplitude of the recovery trend in each brain region, and compare the deviation degree between the actual recovery trajectory and the reference trajectory of the nerve signal to obtain the postoperative nerve activity recovery deviation index; S402: Based on the postoperative nerve activity recovery deviation index, determine the recovery levels of different brain regions after surgery, divide the classification range of the postoperative risk levels, calculate the nerve signal fluctuation data in each brain region, evaluate the stability of the nerve signals during the postoperative recovery period, and use the formula: Among them, R NS,i represents the mean value of the neural signal fluctuations in the i-th brain region, A NS,i,t represents the neural signal intensity in the i-th brain region at time t, A NS,i,0 represents the reference value of the neural signal intensity before surgery in the i-th brain region, |A NS,i,t -A NS,i,0 | represents the deviation value between the neural signal and the reference signal, and T represents the total duration of the postoperative recovery period; Operate to obtain the mean value of the nerve signal fluctuations in each brain region, set the risk level division threshold, and judge the recovery stability of each brain region based on the mean value of the signal fluctuations to obtain the postoperative brain region risk level distribution; S403: Based on the postoperative brain region risk level distribution, call the nerve signal fluctuation data during the postoperative recovery process, evaluate the stability of brain regions with different tolerances during the recovery stage, adjust the risk classification standard according to the changing trend of the recovery stability, calculate the risk fluctuation trend at different time points, adjust the classification range, and comprehensively consider the risk level changes in all time periods to obtain the postoperative risk level distribution result.

7. The method for predicting the risk of neurosurgery based on big data analysis according to claim 1, wherein: The steps for obtaining the postoperative risk prediction result are as follows: S501: Based on the postoperative risk level distribution result and the individual nerve tolerance information, analyze the matching correlation between the tolerance and the risk level, call the individual nerve tolerance data, calculate the brain region distribution under different tolerance levels, set the tolerance classification standard, and compare the postoperative risk level distribution, calculate the distribution ratio of different tolerance groups in each risk level interval to obtain the tolerance and risk level matching index; S502: Based on the tolerance and risk level matching index, calculate the recovery deviation of the postoperative brain regions under different tolerance levels, call the postoperative recovery data, evaluate the signal change trend of each brain region during the recovery stage, calculate the deviation degree between the recovery trajectory and the standard recovery mode, count the recovery deviation data of each brain region at different time points, and set the deviation threshold according to the postoperative risk level standard, and screen the brain regions with recovery deviation exceeding the threshold to obtain the postoperative brain recovery deviation distribution; S503: Based on the postoperative brain recovery deviation distribution, screen the brain regions with risk levels exceeding the risk classification standard, adjust the risk warning range of the postoperative recovery state, calculate the risk warning values at different time points, optimize the risk classification threshold, and use the formula: Among them, R FP,i represents the risk warning value of the i-th brain region, P HR,i,t represents the recovery offset value of the i-th brain region at time t, P TH represents the set risk threshold for recovery offset, W HR,i represents the individual tolerance weighting factor of the i-th brain region, V HR,i,t represents the rate of change of the recovery speed of the i-th brain region at time t, |P HR,i,t -P TH | represents the absolute deviation between the recovery offset value and the risk threshold, represents the impact factor of the change in recovery speed on risk, and T represents the total duration of the postoperative recovery period; Operate to obtain the risk warning values of each brain region, and optimize the risk assessment strategy by combining the data of all time periods to finally obtain the postoperative risk prediction result.

8. A neurosurgical risk prediction system based on big data analysis, characterized in that, Execute according to the neurosurgical risk prediction method based on big data analysis described in any one of claims 1-7. The system includes: The preoperative neurological function assessment module obtains the preoperative functional imaging data of the patient, extracts the neural activity signal intensity in the brain region, calculates the neural signal transmission balance degree, detects the regional distribution of the oxygen metabolism rate, calculates the oxygen metabolism utilization balance, quantitatively analyzes the neurological function state of the brain region, and obtains the preoperative neurological function state information; The postoperative recovery trend analysis module, based on the preoperative neurological function state information, calls the neural activity signal intensity in the postoperative stage, analyzes the signal change trend, detects the brain regions exceeding the neural signal change threshold, screens the brain regions with abnormal oxygen metabolism balance during the postoperative recovery period, calculates the metabolic adaptation ability, and obtains the postoperative recovery mode characteristics; The neural network tolerance assessment module, based on the postoperative recovery mode characteristics, analyzes the dynamic adjustment ability of the postoperative neural network connection, detects the stability of the postoperative neural signal transmission path, calculates the brain regions with excessive signal transmission time deviation, determines the consistency of oxygen metabolism recovery in the abnormal brain regions, calculates the difference in oxygen metabolism recovery rate in different brain regions, evaluates the range of postoperative neural load tolerance of the patient, and obtains the individual neural tolerance information; The postoperative risk grading module, based on the individual neural tolerance information, analyzes the deviation degree of the postoperative neural activity recovery trend, detects the recovery levels of different brain regions, calls the postoperative recovery mode parameters, screens the brain regions with abnormal changes in postoperative recovery stability, adjusts the risk level classification standard according to the recovery stability, and obtains the postoperative risk level distribution result; The postoperative risk prediction module combines the postoperative risk level distribution result and the individual neural tolerance information, analyzes the matching correlation between the tolerance and the risk level, calculates the recovery deviation under different tolerance levels, screens the brain regions with risk levels exceeding the threshold, adjusts the risk warning range of the postoperative recovery state, and obtains the postoperative risk prediction result.

Images