Method for predicting operation pain model program result based on Internet hospital

Abstract

The invention relates to the technical field of medical information, and discloses a method for predicting a surgical pain model program result based on an internet hospital, and the method comprises the steps: 1, collecting the preoperative data of a patient, carrying out the postoperative pain risk prediction based on an artificial intelligence model, and generating a postoperative pain prediction result; 2, according to the postoperative pain prediction result in the step 1, an intraoperative analgesia scheme is formulated, and adjustment is carried out based on real-time physiological parameters of a patient in the operation process; 3, according to the intraoperative data in the step 2, a patient postoperative pain management file is established, a postoperative patient wears intelligent wearable equipment, physiological data are monitored in real time, and a personalized multi-mode analgesia scheme is formulated. According to the method, the multi-layer neural network is combined with the Bayesian reasoning mixed technology, accurate prediction of postoperative acute and chronic pain is achieved, compared with an existing single model, the data fusion and prediction accuracy is greatly improved, and the problem that in a traditional scheme, individual difference capture is insufficient is solved.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, specifically to a method for predicting surgical pain model program results based on an internet hospital. Background Technology

[0002] Postoperative pain management is a major challenge in clinical treatment. Traditional pain management methods rely on the experience of medical staff and fixed treatment plans, lacking the accurate identification and real-time adjustment of individual differences. Especially in the prediction and control of postoperative acute pain and long-term chronic pain, current technologies cannot provide personalized and dynamic pain intervention plans.

[0003] Currently, many hospitals rely on manual assessment and simple pain scoring to monitor postoperative pain in patients. These methods typically do not consider individual patient differences and cannot adjust analgesia regimens in real time. Furthermore, due to the lack of prospective analysis of pain development, many patients suffer from postoperative pain, impacting their recovery quality.

[0004] In existing technologies, postoperative pain prediction models mainly rely on traditional statistical methods or simple neural networks. These models often neglect the fusion of multidimensional features from preoperative data, leading to an inability to accurately predict the trend of acute or chronic postoperative pain. Many traditional methods depend on a single pain score, which cannot effectively identify a patient's potential pain risk.

[0005] Currently, most analgesia protocols are fixed and lack the ability to be personalized. In practice, a patient's pain perception, psychological state, and physiological changes during surgery all affect the effectiveness of pain control. The traditional "step-by-step" approach fails to adequately consider real-time patient feedback, leading to significant differences in analgesic efficacy.

[0006] In-hospital pain monitoring systems cannot be effectively integrated with patients' daily physiological feedback data. Current pain management technologies are limited to postoperative in-hospital care and lack continuous, real-time pain monitoring methods. This means that pain management cannot be followed up and adjusted in a timely manner after patients leave the hospital, leading to poor analgesic effects when changes in the patient's pain are not detected in time.

[0007] To address these issues, this invention proposes a method based on the results of a model program for predicting surgical pain from an internet hospital. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method for predicting surgical pain based on an internet hospital model program, thereby resolving the problems mentioned in the background section.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for predicting surgical pain based on an internet hospital model program, comprising: Step 1: Collect the patient's preoperative data, predict postoperative pain risk based on an artificial intelligence model, and generate postoperative pain prediction results; Step 2: Based on the postoperative pain prediction results from Step 1, formulate an intraoperative analgesia plan and adjust it during the operation based on the patient's real-time physiological parameters. Step 3: Based on the intraoperative data from Step 2, establish a postoperative pain management file for the patient. After the operation, the patient wears a smart wearable device to monitor physiological data in real time, and the AI ​​algorithm analyzes the patient's pain status to automatically generate remote pain data. Step 4: Based on the postoperative pain analysis results in Step 3, the analgesia plan is adjusted in combination with the patient's daily physiological data and pain score. The plan is then sent to the Internet hospital platform for doctor review. After remote confirmation by the doctor, the patient's medication plan and analgesia mode are adjusted. Step 5: Based on the patient's pain status and adjustment records in Step 4, predict the patient's risk of chronic pain. After analysis by the artificial intelligence model, output the probability of chronic pain and mark it in the high-risk patient file. Step 6: Based on the chronic pain prediction results from Step 5, develop a postoperative rehabilitation plan to enable patients to receive remote rehabilitation guidance on the Internet hospital platform.

[0010] Preferably, the preoperative data includes basic information, surgical type, anesthesia method, physiological parameters, psychological state, and genetic information; Step 1, which involves predicting postoperative pain based on multimodal data, further includes: Step 1.1: Normalize the patient data. Let the i-th physiological parameter of the patient be... The normalized data is represented as follows: , in, For normalized data, and These are the minimum and maximum values ​​of the parameter; Step 1.2: Construct feature vectors based on the normalized data. : The input is a deep neural network, and the output of the neural network is a postoperative pain score. : , in, This is the weight matrix of the neural network. For bias terms, For feature vectors, It is the Sigmoid activation function. Postoperative pain prediction score for patients; Step 1.3, based on the results obtained in Step 1.2 Combined with the patient's preoperative psychological state index and individual analgesic sensitivity Calculate the personalized analgesia demand index : , in, , , For empirical weight parameters, This refers to the postoperative personalized analgesia needs index, which is the patient's preoperative psychological state index. Used for optimizing the intraoperative analgesia protocol in step 2.

[0011] Preferably, in step 2, an intraoperative analgesia plan is formulated based on the postoperative pain prediction results of step 1. The analgesia plan combines preoperative data and pain prediction level to determine the initial anesthetic dose, and is adjusted during the operation based on the patient's real-time physiological parameters, including heart rate variability, bispectral index of electroencephalography and electromyography signal. The adjustment results serve as the basic data for postoperative pain management. In step 2, an intraoperative analgesia plan is further developed based on the personalized analgesia demand index in step 1.3, including: Step 2.1, based on the personalized analgesia demand index calculated in Step 1.3 Calculate the initial intraoperative anesthetic dose for the patient. The dose is determined by the following linear function: , in, This is the initial anesthetic dose. This is a personalized analgesia demand index for postoperative patients. To estimate the duration of the surgery, Assess the patient's anesthesia tolerance score. , , Weighted parameters for adjusting anesthetic dosage; Step 2.2: During the surgery, collect the patient's physiological parameters in real time, including heart rate variability. Bifrequency index of brainwave Electromyographic signal intensity Analgesia adjustment factor was calculated after standardization. : , in, For analgesia adjustment factor, This is the reference value for normal heart rate variability. The target is the bispectral index of brainwaves. The maximum electromyographic signal intensity, , , To adjust the weighting coefficients; Step 2.3, based on the initial anesthetic dose in Step 2.1 Compared with the analgesia adjustment coefficient in step 2.2 Calculate the current anesthesia adjustment dose : , in, This is the current adjusted anesthetic dose.

[0012] Preferably, in step 3, based on the intraoperative data from step 2, a postoperative pain management file is established for the patient. The file includes preoperative prediction, intraoperative physiological feedback, and analgesia adjustment information, and is connected to a remote pain monitoring system. The patient wears a smart wearable device to continuously upload data on heart rate variability, skin conductance, steps, and sleep quality. The data is input into an artificial intelligence model for postoperative pain analysis to generate a personalized analgesia plan. In step 3, the adjusted anesthetic dose is based on that in step 2.3. Further establish postoperative pain management records, including: Step 3.1: Adjust the anesthesia dosage according to the instructions in Step 2.3. Calculate the initial postoperative pain index The index is based on anesthetic dosage, intraoperative pain feedback parameters, and the patient's preoperative pain sensitivity. calculate: , in, This represents the initial postoperative pain index. This is the current adjusted anesthetic dose. For intraoperative pain feedback parameters, For preoperative pain sensitivity, , , Weighting parameters for calculating the pain index; Step 3.2: Collect postoperative physiological data of the patient in real time, including heart rate variability. Skin conductance Steps and sleep quality Calculate the postoperative analgesia demand index : , in, This represents the postoperative pain relief demand index. For the Sigmoid function, For patients' postoperative heart rate variability, For skin conductivity, The patient's average daily steps, Assess sleep quality. , , , Weighting parameters calculated for analgesia requirements; Step 3.3, based on the initial pain index calculated in Step 3.1 and the postoperative analgesia demand index calculated in step 3.2 Generate personalized pain relief plans : , in, Scoring of personalized pain relief regimens This represents the initial postoperative pain index. This represents the postoperative pain relief demand index. , Adjust the weighting parameters for personalized analgesia regimens.

[0013] Preferably, in step 4, the analgesia plan is adjusted based on the postoperative pain analysis results of step 3, combined with the patient's daily physiological data and pain score. The adjustment plan includes a recommended analgesia mode, analgesic drug, or non-drug analgesia method, and is pushed to the Internet hospital platform for doctor review. After the doctor confirms remotely, the patient's medication plan and analgesia mode are adjusted, and the adjustment record is used as input data for chronic pain prediction. In step 4, the personalized analgesia plan score from step 3.3 is used as a basis. Further optimize the analgesia regimen, including: Step 4.1: Based on the personalized analgesia plan score calculated in Step 3.3. Combined with the patient's postoperative analgesia compliance Calculate the analgesia regimen adjustment factor : , in, Adjustment factors for the analgesia regimen, Scoring of personalized pain relief regimens To improve postoperative pain management compliance, , Adjust the weighting parameters for the analgesia regimen; Step 4.2, based on the analgesia regimen adjustment factor calculated in Step 4.1 Combined with the patient's postoperative pain score Calculate the dosage of analgesics : , in, To recommend the dosage of analgesics, Adjustment factors for the analgesia regimen, The patient's latest pain score, , Calculate weighting parameters for analgesic drug dosage; Step 4.3, the dosage of analgesic drug calculated in step 4.2. Based on the patient's postoperative mobility and the implementation status of pain relief protocols Calculate the analgesia optimization index : , in, To optimize the analgesia index, To recommend the dosage of analgesics, To improve the patient's postoperative mobility, For the implementation status of the pain relief plan, , , Calculate weight parameters for the analgesia optimization index; The analgesia optimization index Used for predicting chronic pain in step 5.

[0014] Preferably, in step 5, the analgesia optimization index calculated in step 4 is used. Further prediction of a patient's risk of chronic pain includes: Step 5.1, based on the analgesia optimization index calculated in step 4.3 Based on the patient's latest postoperative pain score Calculating the initial risk index for chronic pain : , in, As an initial risk index for chronic pain, To optimize the analgesia index, The patient's latest postoperative pain score, To optimize the contribution weight of analgesia Contribute weight to pain score; Step 5.2: Real-time acquisition of postoperative physiological parameter trends, including heart rate variability rate. Skin conductivity change rate and rate of change in activity capacity Then, the comprehensive change index of physiological parameters was calculated. : , in, It is a comprehensive index of physiological parameter changes. The rate of change of heart rate variability. The rate of change of skin conductivity, The rate of change in activity capacity, Weights for changes in heart rate variability. Weighting for changes in skin conductance. Weighting of changes in activity capacity; Step 5.3, combined with the initial chronic pain risk index calculated in Step 5.1. Combined with the physiological parameter change index calculated in step 5.2 The probability of chronic pain is output using the Sigmoid activation function. : , in, To predict the probability of chronic pain Assign weights to the initial risk. The weights contribute to physiological changes, where e is the base of the natural logarithm.

[0015] Preferably, the preoperative pain prediction model uses a method that combines multilayer neural networks and Bayesian inference to predict postoperative acute pain scores, calculate the evolution trend of postoperative chronic pain, and optimize personalized analgesia plans. The preoperative pain prediction model further includes: a. Multimodal input modeling: Based on the patient's preoperative data X, a high-order feature mapping function is used. Perform multimodal feature fusion: , in, For the input feature matrix, This is the first layer weight matrix. For bias terms, It is a non-linear activation function; b. Postoperative pain score prediction: Calculate the pain score at 24 hours postoperatively. and average pain score over 7 days : , , in, For deep neural networks, nonlinear mapping functions, , This is the weight matrix of the neural network. For bias terms; c. Chronic pain evolution prediction: Preoperative data X influences the probability of long-term postoperative pain. Estimated through Bayesian inference: , in, Given the distribution of preoperative data for chronic pain, This represents the prior probability of chronic pain. This represents the overall distribution of preoperative data.

[0016] Ultimately, the chronic pain development trend index: , in, As an index of the trend of chronic pain development; d. Intraoperative analgesia optimization: Based on preoperative pain prediction results, reinforcement learning was used to optimize the intraoperative analgesia strategy. , in, The optimal pain relief strategy is... As a discount factor, Rewards for feedback on pain relief effects. These are the preoperative and intraoperative states. For the intraoperative decision time step, Expected value calculation.

[0017] Preferably, the postoperative pain management, combined with outpatient physiological feedback, uses dynamic data to adjust personalized pain management plans, achieving continuous pain monitoring and personalized intervention, further including: a. Physiological feedback data fusion: Patients wear wearable devices after surgery to record physiological data in real time. : , in, For heart rate variability, For skin conductivity, Assess sleep quality. For activity steps, , , , As weight, For bias terms; b. Dynamic prediction of pain scores: Predicting future pain scores using an LSTM model: , in, Rate the pain at the next moment. For the current pain score, For physiological feedback input; c. Personalized medication adjustment outside the hospital: Optimizing analgesia protocols using deep reinforcement learning. , in, For pain relief regimen In state The value function under, As a reward for pain relief feedback, As a discount factor, For pain relief regimen In state The maximum value function under; d. Remote monitoring and doctor intervention: Patient pain scores, analgesia compliance, and physiological feedback data are uploaded to the Internet hospital system, allowing doctors to monitor remotely and provide personalized intervention.

[0018] Based on a terminal device, the terminal device includes a hardware and software platform for implementing pain prediction and management methods. The hardware includes wearable devices, sensors, and communication modules, and the software includes intelligent applications for data acquisition, AI analysis, pain assessment, medication recommendation, and physician intervention.

[0019] Based on the storage medium, the storage medium contains software programs for implementing pain prediction and management methods, the software programs including computer instructions for data acquisition, AI analysis, pain assessment, medication recommendation, and physician intervention.

[0020] This invention provides a method for predicting surgical pain using a model program based on an internet hospital. It has the following beneficial effects: 1. This invention employs a hybrid technique combining multi-layer neural networks and Bayesian inference to achieve accurate prediction of postoperative acute and chronic pain. Compared with existing single models, this invention significantly improves data fusion and prediction accuracy, and solves the problem of insufficient capture of individual differences in traditional solutions.

[0021] 2. This invention introduces reinforcement learning technology to optimize intraoperative analgesia strategies in real time, enabling personalized analgesia modes and medication adjustments. Unlike traditional fixed analgesia protocols, this invention significantly improves upon the shortcomings of previous methods that could not dynamically adapt to changes in the patient's condition.

[0022] 3. This invention integrates physiological feedback data from inside and outside the hospital and enables remote continuous monitoring through wearable devices to improve postoperative pain management. Compared with traditional technologies that are limited to in-hospital monitoring, this invention breaks through the limitations of time and space to intervene and overcomes the shortcomings of existing technologies in remote management. Attached Figure Description

[0023] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0024] To enable those skilled in the art to understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0025] The present invention will now be described in detail with reference to the accompanying drawings: Example: Please see the appendix Figure 1 This invention provides a method for predicting surgical pain based on an internet hospital model program, comprising: Step 1: Collect the patient's preoperative data, predict postoperative pain risk based on an artificial intelligence model, and generate postoperative pain prediction results; Step 1.1: Normalize the patient data. Let the i-th physiological parameter of the patient be... The normalized data is represented as follows: , in, For normalized data, and These are the minimum and maximum values ​​of the parameter; Step 1.2: Construct feature vectors based on the normalized data. : Furthermore, the input is a deep neural network, and the output of the neural network is a postoperative pain score. : , in, This is the weight matrix of the neural network. For bias terms, For feature vectors, It is the Sigmoid activation function. Postoperative pain prediction score for patients; Step 1.3, based on the results obtained in Step 1.2 Combined with the patient's preoperative psychological state index and individual analgesic sensitivity Calculate the personalized analgesia demand index : , in, , , For empirical weight parameters, The postoperative personalized analgesia demand index and the patient's preoperative psychological state index. For optimization of intraoperative analgesia protocols in step 2; Step 2: Based on the postoperative pain prediction results from Step 1, formulate an intraoperative analgesia plan and adjust it during the operation based on the patient's real-time physiological parameters. Step 2.1, based on the personalized analgesia demand index calculated in Step 1.3 Calculate the initial intraoperative anesthetic dose for the patient. The dosage is determined by the following linear function: , in, This is the initial anesthetic dose. This is a personalized analgesia demand index for postoperative patients. To estimate the duration of the surgery, Assess the patient's anesthesia tolerance score. , , Weighted parameters for adjusting anesthetic dosage; Step 2.2: During the surgery, collect the patient's physiological parameters in real time, including heart rate variability. Bifrequency index of brainwave Electromyographic signal intensity Analgesia adjustment factor was calculated after standardization. : , in, For analgesia adjustment factor, This is the reference value for normal heart rate variability. The target is the bispectral index of brainwaves. The maximum electromyographic signal intensity, , , To adjust the weighting coefficients; Step 2.3, based on the initial anesthetic dose in Step 2.1 Compared with the analgesia adjustment coefficient in step 2.2 Calculate the current anesthesia adjustment dose : , in, This is the current adjusted anesthetic dose; Step 3: Based on the intraoperative data from Step 2, establish a postoperative pain management file for the patient. After the operation, the patient wears a smart wearable device to monitor physiological data in real time, and the AI ​​algorithm analyzes the patient's pain status to automatically generate a remote pain score. Step 3.1: Adjust the anesthesia dosage according to the instructions in Step 2.3. Calculate the initial postoperative pain index The index is based on anesthetic dosage, intraoperative pain feedback parameters, and the patient's preoperative pain sensitivity. calculate: , in, This represents the initial postoperative pain index. This is the current adjusted anesthetic dose. For intraoperative pain feedback parameters, For preoperative pain sensitivity, , , Weighting parameters for calculating the pain index; Step 3.2: Collect postoperative physiological data of the patient in real time, including heart rate variability. Skin conductance Steps and sleep quality Calculate the postoperative analgesia demand index : , in, This represents the postoperative pain relief demand index. For the Sigmoid function, For patients' postoperative heart rate variability, For skin conductivity, The patient's average daily steps, Assess sleep quality. , , , Weighting parameters calculated for analgesia requirements; Step 3.3, based on the initial pain index calculated in Step 3.1 and the postoperative analgesia demand index calculated in step 3.2 Generate personalized pain relief plans : , in, Scoring of personalized pain relief regimens This represents the initial postoperative pain index. This represents the postoperative pain relief demand index. , Adjusting weighting parameters for personalized analgesia regimens; Step 4: Based on the postoperative pain analysis results in Step 3, the analgesia plan is adjusted in combination with the patient's daily physiological data and pain score. The plan is then sent to the Internet hospital platform for doctor review. After remote confirmation by the doctor, the patient's medication plan and analgesia mode are adjusted. Step 4.1: Based on the personalized analgesia plan score calculated in Step 3.3. Combined with the patient's postoperative analgesia compliance Calculate the analgesia regimen adjustment factor : , in, Adjustment factors for the analgesia regimen, Scoring of personalized pain relief regimens To improve postoperative pain management compliance, , Adjust the weighting parameters for the analgesia regimen; Step 4.2, based on the analgesia regimen adjustment factor calculated in Step 4.1 Combined with the patient's postoperative pain score Calculate the dosage of analgesics : , in, To recommend the dosage of analgesics, Adjustment factors for the analgesia regimen, The patient's latest pain score, , Calculate weighting parameters for analgesic drug dosage; Step 4.3, the dosage of analgesic drug calculated in step 4.2. Based on the patient's postoperative mobility and the implementation status of pain relief protocols Calculate the analgesia optimization index : , in, To optimize the analgesia index, To recommend the dosage of analgesics, To improve the patient's postoperative mobility, For the implementation status of the pain relief plan, , , Calculate weight parameters for the analgesia optimization index; Analgesia optimization index Chronic pain prediction for step 5; Step 5: Based on the patient's pain status and adjustment records in Step 4, predict the patient's risk of chronic pain. After analysis by the artificial intelligence model, output the probability of chronic pain and mark it in the high-risk patient file. Step 5.1, based on the analgesia optimization index calculated in step 4.3 Based on the patient's latest postoperative pain score Calculating the initial risk index for chronic pain : , in, As an initial risk index for chronic pain, To optimize the analgesia index, The patient's latest postoperative pain score, To optimize the contribution weight of analgesia Contribute weight to pain score; Step 5.2: Real-time acquisition of postoperative physiological parameter trends, including heart rate variability rate. Skin conductivity change rate and rate of change in activity capacity Then, the comprehensive change index of physiological parameters was calculated. : , in, It is a comprehensive index of physiological parameter changes. The rate of change of heart rate variability. The rate of change of skin conductivity, The rate of change in activity capacity, Weights for changes in heart rate variability. Weighting for changes in skin conductance. Weighting of changes in activity capacity; Step 5.3, combined with the initial chronic pain risk index calculated in Step 5.1. Combined with the physiological parameter change index calculated in step 5.2 The probability of chronic pain is output using the Sigmoid activation function. : , in, To predict the probability of chronic pain Assign weights to the initial risk. The weights contribute to physiological changes, where e is the base of the natural logarithm; Step 6: Based on the chronic pain prediction results from Step 5, develop a postoperative rehabilitation plan to enable patients to receive remote rehabilitation guidance on the Internet hospital platform.

[0026] Step 1 involves comprehensively collecting multidimensional preoperative data from patients. Normalization and feature extraction processes are used to ensure high-quality, reasonably distributed input data. Deep neural networks are employed to extract key features, effectively capturing individual differences. Compared to traditional data processing methods, this significantly improves the accuracy and robustness of subsequent predictions.

[0027] Step 2 calculates the initial anesthetic dose using a linear model and dynamically adjusts the analgesia strategy based on the patient's real-time physiological parameters. This overcomes the rigidity of traditional fixed analgesia protocols and enables personalized medication. It is simple, direct, and responsive, ensuring precise and flexible analgesia during surgery. Step 3 utilizes smart wearable devices and AI algorithms to establish a postoperative pain management file for the patient, collecting and analyzing physiological data in real time. Continuous monitoring allows for timely feedback and intervention regarding changes in pain status. Compared to single in-hospital monitoring sessions, this approach comprehensively records the patient's recovery process, improving management efficiency. Step 4 calculates the analgesia regimen adjustment factors and drug dosages, and sends them to the doctor for review to ensure the scientific validity of the intervention and timely response to changes in the patient's condition. Compared to traditional methods that rely on subjective judgment, this approach is objective and effective, ensuring the precise implementation of the analgesia regimen. Step 5 utilizes the analgesia optimization index and the comprehensive change index of physiological parameters to calculate the predictive probability of chronic pain using the Sigmoid function. This prospective risk assessment provides early warning for clinicians, aiding in timely intervention. Compared to traditional methods that cannot predict long-term pain trends, it has high practical value and accuracy. Step 6 involves developing a personalized rehabilitation plan for the patient and providing remote rehabilitation guidance through an internet platform, breaking the traditional time and space limitations of in-hospital rehabilitation. Patients can continuously receive professional intervention, resulting in a consistent and efficient overall rehabilitation outcome.

[0028] In summary, the steps of this invention form a complete pain prediction and management process, ultimately achieving personalized, real-time, and scientific pain management, and significantly improving patients' analgesia and rehabilitation experience.

[0029] The preoperative pain prediction model uses a fusion of multilayer neural networks and Bayesian inference to predict postoperative acute pain scores, calculate the evolution trend of postoperative chronic pain, and optimize personalized analgesia plans. The preoperative pain prediction model further includes: a. Multimodal input modeling: Based on the patient's preoperative data X, a high-order feature mapping function is used. Perform multimodal feature fusion: , in, For the input feature matrix, This is the first layer weight matrix. For bias terms, It is a non-linear activation function; b. Postoperative pain score prediction: Calculate the pain score at 24 hours postoperatively. and average pain score over 7 days : , , in, For deep neural networks, nonlinear mapping functions, , This is the weight matrix of the neural network. For bias terms; c. Chronic pain evolution prediction: Preoperative data X influences the probability of long-term postoperative pain. Estimated through Bayesian inference: , in, Given the distribution of preoperative data for chronic pain, This represents the prior probability of chronic pain. This represents the overall distribution of preoperative data.

[0030] Ultimately, the chronic pain development trend index: , in, As an index of the trend of chronic pain development; d. Intraoperative analgesia optimization: Based on preoperative pain prediction results, reinforcement learning was used to optimize the intraoperative analgesia strategy. , in, The optimal pain relief strategy is... As a discount factor, Rewards for feedback on pain relief effects. These are the preoperative and intraoperative states. For the intraoperative decision time step, Expected value calculation.

[0031] This step uses a high-order feature mapping function to achieve deep fusion of multidimensional patient data, which can comprehensively capture individual differences. Compared with traditional single data input, it avoids missing key information, solves the problem of information silos, and makes predictions more accurate. By assessing the impact of preoperative data on long-term pain development through Bayesian inference, a chronic pain evolution trend index is formed. This approach, combining prior knowledge with data distribution, is more scientific and rigorous than traditional empirical estimation, effectively providing early warning of chronic pain risks. Reinforcement learning is introduced to optimize analgesia strategies, making real-time decisions on analgesia plans based on the patient's perioperative condition. Compared to fixed procedures, this approach flexibly adapts to changes in the patient's condition, reducing the occurrence of over- or under-analgesia and improving medication accuracy.

[0032] Postoperative pain management, combined with out-of-hospital physiological feedback, uses dynamic data to adjust personalized pain management plans, achieving continuous pain monitoring and personalized intervention, and further includes: a. Physiological feedback data fusion: Patients wear wearable devices after surgery to record physiological data in real time. : , in, For heart rate variability, For skin conductivity, Assess sleep quality. For activity steps, , , , As weight, For bias terms; b. Dynamic prediction of pain scores: Predicting future pain scores using an LSTM model: , in, Rate the pain at the next moment. For the current pain score, For physiological feedback input; c. Personalized medication adjustment outside the hospital: Optimizing analgesia protocols using deep reinforcement learning. , in, For pain relief regimen In state The value function under, As a reward for pain relief feedback, As a discount factor, For pain relief regimen In state The maximum value function under; d. Remote monitoring and doctor intervention: Patient pain scores, analgesia compliance, and physiological feedback data are uploaded to the Internet hospital system, allowing doctors to monitor remotely and provide personalized intervention.

[0033] This step utilizes wearable devices to collect real-time data on the patient's heart rate variability, skin conductance, sleep score, and steps, performing multi-dimensional fusion to comprehensively reflect the patient's postoperative physiological state, providing greater accuracy than single-data monitoring. It effectively overcomes the shortcomings of traditional methods, such as providing incomplete data and poor real-time performance.

[0034] An LSTM model is used to predict future pain scores, providing short, direct, and rapid feedback. The model focuses on current scores and can capture trend changes. Compared to traditional static prediction methods, it is better able to handle pain fluctuations and adjust intervention plans accordingly.

[0035] By uploading patients' pain scores, analgesia compliance, and physiological data to the internet hospital system, doctors can remotely monitor and intervene in real time. Compared to traditional in-hospital monitoring, this breaks down time and space limitations, ensuring timely adjustments to treatment plans.

[0036] Based on terminal devices, the terminal devices include hardware and software platforms for implementing pain prediction and management methods. The hardware includes wearable devices, sensors, and communication modules, while the software includes intelligent applications for data acquisition, AI analysis, pain assessment, medication recommendation, and physician intervention.

[0037] Based on the storage medium, the storage medium contains software programs for implementing pain prediction and management methods, including computer instructions for data acquisition, AI analysis, pain assessment, medication recommendation, and physician intervention.

[0038] The terminal device organically integrates wearable devices, sensors, and communication modules, working in conjunction with a smart application to automate the entire process from data collection, AI analysis, pain assessment to medication recommendation and physician intervention. The hardware provides real-time, multi-dimensional physiological data monitoring, ensuring accurate data and immediate feedback; the software platform, through artificial intelligence algorithms and intelligent decision support, enables personalized pain management and intervention. Overall, it overcomes the limitations of traditional medical monitoring methods, such as single-point monitoring and delayed feedback, providing patients with a comprehensive, accurate, and dynamic pain management solution.

[0039] The software program instructions contained in the storage medium constitute the core operating logic of the pain prediction and management system. Through computer instructions, data acquisition, AI analysis, pain assessment, medication recommendation, and physician intervention are achieved, ensuring the system's automation and efficiency. The storage medium can stably and efficiently run complex algorithms, while guaranteeing the flexibility and scalability of the management scheme in practical applications. This enables the medical system to respond quickly to changes in patient status, further improving the scientific rigor and accuracy of clinical decision-making.

[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for predicting surgical pain based on a model program from an internet hospital, characterized in that, include: Step 1: Collect the patient's preoperative data, predict postoperative pain risk based on an artificial intelligence model, and generate postoperative pain prediction results; Step 2: Based on the postoperative pain prediction results from Step 1, formulate an intraoperative analgesia plan and adjust it during the operation based on the patient's real-time physiological parameters. Step 3: Based on the intraoperative data from Step 2, establish a postoperative pain management file for the patient. After the operation, the patient wears a smart wearable device to monitor physiological data in real time, and the AI ​​algorithm analyzes the patient's pain status to automatically generate remote pain data. Step 4: Based on the postoperative pain analysis results in Step 3, the analgesia plan is adjusted in combination with the patient's daily physiological data and pain score. The plan is then sent to the Internet hospital platform for doctor review. After remote confirmation by the doctor, the patient's medication plan and analgesia mode are adjusted. Step 5: Based on the patient's pain status and adjustment records in Step 4, predict the risk of the patient developing chronic pain. After analysis by the artificial intelligence model, output the probability of chronic pain and mark it in the high-risk patient file. Step 6: Based on the chronic pain prediction results from Step 5, develop a postoperative rehabilitation plan to enable patients to receive remote rehabilitation guidance on the Internet hospital platform.

2. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, The preoperative data includes basic information, type of surgery, anesthesia method, physiological parameters, psychological state, and genetic information; Step 1, which involves predicting postoperative pain based on multimodal data, further includes: Step 1.1: Normalize the patient data. Let the i-th physiological parameter of the patient be... The normalized data is represented as follows: ， in, For normalized data, and These are the minimum and maximum values ​​of the parameter; Step 1.2: Construct feature vectors based on the normalized data. : The input is a deep neural network, and the output of the neural network is a postoperative pain score. : ， in, This is the weight matrix of the neural network. For bias terms, For feature vectors, It is the Sigmoid activation function. Postoperative pain prediction score for patients; Step 1.3, based on the results obtained in Step 1.2 Combined with the patient's preoperative psychological state index and individual analgesic sensitivity Calculate the personalized analgesia demand index : ， in, , , For empirical weight parameters, This refers to the postoperative personalized analgesia needs index, which is the patient's preoperative psychological state index. Used for optimizing the intraoperative analgesia protocol in step 2.

3. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, In step 2, an intraoperative analgesia plan is formulated based on the postoperative pain prediction results of step 1. The analgesia plan combines preoperative data and pain prediction level to determine the initial anesthetic dose, and is adjusted during the operation based on the patient's real-time physiological parameters, including heart rate variability, bispectral index of electroencephalography and electromyography signal. The adjustment results serve as the basic data for postoperative pain management. In step 2, an intraoperative analgesia plan is further developed based on the personalized analgesia demand index in step 1.3, including: Step 2.1, based on the personalized analgesia demand index calculated in Step 1.3 Calculate the initial intraoperative anesthetic dose for the patient. The dose is determined by the following linear function: ， in, This is the initial anesthetic dose. This is a personalized analgesia demand index for postoperative patients. To estimate the duration of the surgery, Assess the patient's anesthesia tolerance score. , , Weighted parameters for adjusting anesthetic dosage; Step 2.2: During the surgery, collect the patient's physiological parameters in real time, including heart rate variability. Bifrequency index of brainwave Electromyographic signal intensity Analgesia adjustment factor was calculated after standardization. : ， in, For analgesia adjustment factor, This is the reference value for normal heart rate variability. The target is the bispectral index of brainwaves. The maximum electromyographic signal intensity, , , To adjust the weighting coefficients; Step 2.3, based on the initial anesthetic dose in Step 2.1 Compared with the analgesia adjustment coefficient in step 2.2 Calculate the current anesthesia adjustment dose : ， in, This is the current adjusted anesthetic dose.

4. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, In step 3, based on the intraoperative data from step 2, a postoperative pain management file is established for the patient. The file includes preoperative prediction, intraoperative physiological feedback and analgesia adjustment information, and is connected to a remote pain monitoring system. The patient wears a smart wearable device to continuously upload data on heart rate variability, skin conductance, steps and sleep quality. The data is input into an artificial intelligence model for postoperative pain analysis to generate a personalized analgesia plan. In step 3, the adjusted anesthetic dose is based on that in step 2.

3. Further establish postoperative pain management records, including: Step 3.1: Adjust the anesthesia dosage according to the instructions in Step 2.

3. Calculate the initial postoperative pain index The index is based on anesthetic dosage, intraoperative pain feedback parameters, and the patient's preoperative pain sensitivity. calculate: ， in, This represents the initial postoperative pain index. This is the current adjusted anesthetic dose. For intraoperative pain feedback parameters, For preoperative pain sensitivity, , , Weighting parameters for calculating the pain index; Step 3.2: Collect postoperative physiological data of the patient in real time, including heart rate variability. Skin conductance Steps and sleep quality Calculate the postoperative analgesia demand index : ， in, This represents the postoperative pain relief demand index. For the Sigmoid function, For patients' postoperative heart rate variability, For skin conductivity, The patient's average daily steps, Assess sleep quality. , , , Weighting parameters calculated for analgesia requirements; Step 3.3, based on the initial pain index calculated in Step 3.1 and the postoperative analgesia demand index calculated in step 3.2 Generate personalized pain relief plans : ， in, Scoring of personalized pain relief regimens This represents the initial postoperative pain index. This represents the postoperative pain relief demand index. , Adjust the weighting parameters for personalized analgesia regimens.

5. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, In step 4, based on the postoperative pain analysis results of step 3, the analgesia plan is adjusted in combination with the patient's daily physiological data and pain score. The adjustment plan includes recommending analgesia mode or analgesic drug selection or non-pharmacological analgesia method, and is pushed to the Internet hospital platform for doctor review. After the doctor confirms remotely, the patient's analgesia plan is adjusted, and the adjustment record is used as input data for chronic pain prediction. In step 4, the personalized analgesia plan score from step 3.3 is used as a basis. Further optimize the analgesia regimen, including: Step 4.1: Based on the personalized analgesia plan score calculated in Step 3.

3. Combined with the patient's postoperative analgesia compliance Calculate the analgesia regimen adjustment factor : ， in, Adjustment factors for the analgesia regimen, Scoring of personalized pain relief regimens To improve postoperative pain management compliance, , Adjust the weighting parameters for the analgesia regimen; Step 4.2, based on the analgesia regimen adjustment factor calculated in Step 4.1 Combined with the patient's postoperative pain score Calculate the dosage of analgesics : ， in, To recommend the dosage of analgesics, Adjustment factors for the analgesia regimen, The patient's latest pain score, , Calculate weighting parameters for analgesic drug dosage; Step 4.3, the dosage of analgesic drug calculated in step 4.

2. Based on the patient's postoperative mobility and the implementation status of pain relief protocols Calculate the analgesia optimization index : ， in, To optimize the analgesia index, To recommend the dosage of analgesics, To improve the patient's postoperative mobility, For the implementation status of the pain relief plan, , , Calculate weight parameters for the analgesia optimization index; The analgesia optimization index Used for predicting chronic pain in step 5.

6. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, In step 5, the analgesia optimization index calculated in step 4 is used. Further prediction of a patient's risk of chronic pain includes: Step 5.1, based on the analgesia optimization index calculated in step 4.3 Based on the patient's latest postoperative pain score Calculating the initial risk index for chronic pain : ， in, As an initial risk index for chronic pain, To optimize the analgesia index, The patient's latest postoperative pain score, To optimize the contribution weight of analgesia Contribute weight to pain score; Step 5.2: Real-time acquisition of postoperative physiological parameter trends, including heart rate variability rate. Skin conductivity change rate and rate of change in activity capacity Then, the comprehensive change index of physiological parameters was calculated. : ， in, It is a comprehensive index of physiological parameter changes. The rate of change of heart rate variability. The rate of change of skin conductivity, The rate of change in activity capacity, Weights for changes in heart rate variability. Weighting for changes in skin conductance. Weighting of changes in activity capacity; Step 5.3, combined with the initial chronic pain risk index calculated in Step 5.

1. Combined with the physiological parameter change index calculated in step 5.2 The probability of chronic pain is output using the Sigmoid activation function. : ， in, To predict the probability of chronic pain Assign weights to the initial risk. The weights contribute to physiological changes, where e is the base of the natural logarithm.

7. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, The preoperative pain prediction model uses a fusion of multilayer neural networks and Bayesian inference to predict postoperative acute pain scores, calculate the evolution trend of postoperative chronic pain, and optimize personalized analgesia plans. The preoperative pain prediction model further includes: a. Multimodal input modeling: Based on the patient's preoperative data X, a high-order feature mapping function is used. Perform multimodal feature fusion: ， in, For the input feature matrix, This is the first layer weight matrix. For bias terms, It is a non-linear activation function; b. Postoperative pain score prediction: Calculate the pain score at 24 hours postoperatively. and average pain score over 7 days : ， ， in, For deep neural networks, nonlinear mapping functions, , This is the weight matrix of the neural network. For bias terms; c. Chronic pain evolution prediction: Preoperative data X influences the probability of long-term postoperative pain. Estimated through Bayesian inference: ， in, Given the distribution of preoperative data for chronic pain, This represents the prior probability of chronic pain. This represents the overall distribution of preoperative data. Finally, the chronic pain progression trend index: ， in, As an index of the trend of chronic pain development; d. Intraoperative analgesia optimization: Based on preoperative pain prediction results, reinforcement learning was used to optimize the intraoperative analgesia strategy. ， in, The optimal pain relief strategy is... As a discount factor, Rewards for feedback on pain relief effects. These are the preoperative and intraoperative states. For the intraoperative decision time step, Expected value calculation.

8. The method for predicting surgical pain based on an internet hospital model program according to claim 1, characterized in that, The postoperative pain management, combined with out-of-hospital physiological feedback, uses dynamic data to adjust personalized pain management plans, achieving continuous pain monitoring and personalized intervention, and further includes: a. Physiological feedback data fusion: Patients wear wearable devices after surgery to record physiological data in real time. : ， in, For heart rate variability, For skin conductivity, Assess sleep quality. For activity steps, , , , As weight, For bias terms; b. Dynamic prediction of pain scores: Predicting future pain scores using an LSTM model: ， in, Rate the pain at the next moment. For the current pain score, For physiological feedback input; c. Personalized medication adjustment outside the hospital: Optimizing analgesia protocols using deep reinforcement learning. ， in, For pain relief regimen In state The value function under, As a reward for pain relief feedback, As a discount factor, For pain relief regimen In state The maximum value function under; d. Remote monitoring and doctor intervention: Patient pain data, analgesia compliance, and physiological feedback data are uploaded to the Internet hospital system, allowing doctors to monitor remotely and provide personalized intervention.

9. Based on a terminal device, characterized in that, The terminal device includes a hardware and software platform for implementing the pain prediction and management method of claim 1. The hardware includes wearable devices, sensors, and communication modules, and the software includes intelligent applications for data acquisition, AI analysis, pain assessment, drug recommendation, and physician intervention.

10. Based on a storage medium, characterized in that, The storage medium contains a software program for implementing the pain prediction and management method of claim 1, the software program including computer instructions for data acquisition, AI analysis, pain assessment, drug recommendation, and physician intervention.

Images