Patient tracking method based on multi-dimensional data integration and machine learning algorithm

Abstract

The invention relates to the technical field of patient management, and discloses a patient tracking method based on multi-dimensional data integration and a machine learning algorithm. The method comprises the steps of constructing a perioperative anesthesia management model and executing preoperative anesthesia scheme recommendation, intraoperative anesthesia effect supervision, postoperative analgesia scheme recommendation and postoperative analgesia effect supervision by using the perioperative anesthesia management model. According to the method, the historical scheme and effect are matched through the preoperative data, the preoperative stability index updating scheme is calculated, preoperative anesthesia scheme recommendation is provided for the patient, the time and energy of medical staff for collecting and carding data are shortened, and the accuracy and individuation of the anesthesia scheme are improved; an anesthesia deviation value is calculated and monitored in real time during operation, so that anesthesia effect abnormity is found in time and an alarm is given out; the operation stability index is calculated based on the preoperative and intraoperative data after operation, the analgesia scheme is updated and supervised, the analgesia requirement of the patient is met, the accuracy, individuation and automation level of medical service are integrally improved, and the safety and rehabilitation quality of the patient are guaranteed.

Description

Technical Field

[0001] The present application relates to the field of patient management technology, and specifically to a patient tracking method based on multidimensional data integration and machine learning algorithms. Background Art

[0002] In the medical field, traditional patient management methods expose many problems when faced with massive and complex patient data, specifically:

[0003] 1. Medical data comes from a wide range of sources, including medical history, laboratory test results, imaging data in the electronic medical record system, and personalized information provided by the patient himself, such as drug allergy history, previous surgical experience, etc. These data are often stored in a scattered manner and lack an effective integration mechanism, which causes medical staff to spend a lot of time and energy to collect and sort out when making decisions, increasing the workload and making mistakes easily due to human negligence.

[0004] 2. In the process of patient tracking, whether it is preoperative evaluation, intraoperative monitoring or postoperative management, the efficiency of data analysis and utilization is low, especially in anesthesia management. Traditional methods are difficult to accurately predict anesthesia risks and timely adjust analgesia plans, and cannot fully meet the individual needs of patients, seriously affecting the formulation of medical decisions and the quality of diagnosis and treatment effects. Chinese patent number CN118609801B discloses a surgical anesthesia information early warning method and system based on clinical decision-making assistance, but the anesthesia risk identification model and wound healing level identification model in the invention have poor universality, and in medical scenarios, transparent and explainable models are crucial to building trust and assisting clinical judgment, but the model output of the invention is only a risk level probability value vector, which cannot clearly explain the specific impact of each factor on the result, and the model lacks interpretability.

[0005] In summary, there is an urgent need for a new technical solution for patient tracking based on multidimensional data integration and machine learning algorithms to improve the accuracy, personalization, and automation of medical services and ensure the safety and rehabilitation quality of patients. Summary of the invention

[0006] The purpose of this application is to provide a patient tracking method based on multidimensional data integration and machine learning algorithms to solve the technical problems raised in the above background technology.

[0007] To achieve the above objectives, the present application discloses the following technical solutions: A patient tracking method based on multidimensional data integration and machine learning algorithm, comprising:

[0008] Constructing a perioperative anesthesia management model, and using the perioperative anesthesia management model to perform preoperative anesthesia plan recommendation, intraoperative anesthesia effect supervision, postoperative analgesia plan recommendation, and postoperative analgesia effect supervision;

[0009] The recommended preoperative anesthesia plan includes: collecting preoperative data of the patient; matching the anesthesia history plan of patients of the same type and the anesthesia history effect generated by the anesthesia history plan based on the preoperative data; calculating a preoperative stability index based on the preoperative data, and updating the anesthesia history plan based on the preoperative stability index to obtain an anesthesia real-time plan and the anesthesia prediction effect generated by the anesthesia real-time plan; packaging and recommending the anesthesia history plan and the anesthesia history effect generated by the anesthesia history plan, the anesthesia real-time plan and the anesthesia prediction effect generated by the anesthesia real-time plan, and the process of updating the anesthesia history plan based on the preoperative stability index; wherein, the preoperative stability index is used to characterize the fluctuation of the patient's physiological state before surgery, and the updating of the anesthesia history plan based on the preoperative stability index is to update the anesthesia operation in the anesthesia history plan.

[0010] The supervision of intraoperative anesthesia effect includes: collecting intraoperative data generated by implementing the anesthesia real-time plan for the patient; calculating the anesthesia deviation value of the intraoperative data compared with the anesthesia history effect and the anesthesia prediction effect within a preset warning period. When the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generating an intraoperative real-time warning, and displaying the intraoperative real-time warning and the process of calculating the anesthesia deviation value of the intraoperative data compared with the anesthesia history effect and the anesthesia prediction effect within a preset warning period in real time; wherein, the anesthesia deviation value is used to characterize the deviation degree of the intraoperative anesthesia real-time effect of the patient compared with the anesthesia history effect and the anesthesia prediction effect.

[0011] The recommended postoperative analgesia plan includes: collecting postoperative data of the patient; matching the analgesia history plan of patients of the same type and the analgesia history effect generated by the analgesia history plan based on the postoperative data; calculating a surgical stability index based on the preoperative data and the intraoperative data, and updating the analgesia history plan based on the surgical stability index to obtain an analgesia real-time plan and the analgesia prediction effect generated by the analgesia real-time plan; packaging and recommending the analgesia history plan and the analgesia history effect generated by the analgesia history plan, the analgesia real-time plan and the analgesia prediction effect generated by the analgesia real-time plan, and the process of updating the analgesia history plan based on the surgical stability index; wherein, the surgical stability index is used to characterize the fluctuation of the patient's physiological state before and during surgery, and the updating of the analgesia history plan based on the surgical stability index is to update the analgesia operation in the analgesia history plan.

[0012] The postoperative analgesia effect supervision includes: collecting analgesia data generated by implementing the real-time analgesia plan for the patient; calculating the analgesia deviation value of the analgesia data compared with the historical analgesia effect and the predicted analgesia effect within the warning period. When the analgesia deviation value does not meet the preset analgesia deviation threshold, generating a real-time analgesia warning, and displaying the real-time analgesia warning and the process of calculating the analgesia deviation value of the analgesia data compared with the historical analgesia effect and the predicted analgesia effect within the warning period in real time; wherein, the analgesia deviation value is used to characterize the deviation degree of the real-time analgesia effect of the patient after surgery compared with the historical analgesia effect and the predicted analgesia effect.

[0013] Preferably, the preoperative data is specifically:

[0014] Collecting preoperative data by integrating the hospital's electronic case system and the patient's preoperative physiological monitoring system. The preoperative data at least includes the patient's basic information, preoperative laboratory test results, preoperative imaging examination results, and preoperative vital sign data.

[0015] Preferably, calculating the preoperative stability index based on the preoperative data, and updating the historical anesthesia plan based on the preoperative stability index to obtain the real-time anesthesia plan and the predicted anesthesia effect generated by the real-time anesthesia plan, specifically:

[0016] Calculating the preoperative stability index using the preoperative stability index calculation formula. The preoperative stability index calculation formula is:

[0017]

[0018] Wherein, ∫|Δp v | is the integral of the change rate of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data numbers, and PSI is the calculated preoperative stability index;

[0019] Updating the historical anesthesia plan using the anesthesia historical plan update formula. The anesthesia historical plan update formula is:

[0020] A j_new =max[b*Δ(PSI norm_1 *A j_old )+c*PSI norm_2 *E j_old

[0021] Wherein, A j_old is the anesthesia historical plan j of the same type of patient obtained by matching, PSI norm_1 is the correction parameter for correcting the anesthesia historical plan based on the normalized preoperative stability index, Δ(PSI norm_1 *A​j_old ) is the change rate of the revised anesthesia history plan j based on the correction parameter, E j_old is the anesthesia history effect of the anesthesia history plan j, PSI norm_2 is the correction parameter for correcting the anesthesia history effect based on the preoperative stability index after normalization, b and c are preset weights, max[] represents taking the maximum value, A j_new is the calculated and screened real-time anesthesia plan, and the corresponding anesthesia prediction effect at this time is PSI norm_2 *E j_old .

[0022] Preferably, the intraoperative data is specifically:

[0023] Collect intraoperative data through the operating room system of the integrated hospital and the intraoperative physiological monitoring system of the patient. The intraoperative data at least includes the intraoperative vital sign data, anesthesia drug usage data, and surgical operation data of the patient.

[0024] Preferably, calculate the anesthesia deviation value of the intraoperative data compared with the anesthesia history effect and the anesthesia prediction effect within a preset warning period. When the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generate a real-time intraoperative warning, specifically:

[0025] Calculate the anesthesia deviation value using the anesthesia deviation value calculation formula. The anesthesia deviation value formula is:

[0026] AneD = max[Δ(E j_now , E j_old ), Δ(E j_now , PSI norm_2 *E j_old )]

[0027] Wherein, E j_now is the real-time anesthesia effect obtained based on the intraoperative data, Δ(E j_now , E j_old ) is the change rate of the real-time anesthesia effect compared with the anesthesia history effect, Δ(E j_now , PSI norm_2 *E j_old ) is the change rate of the real-time anesthesia effect compared with the anesthesia prediction effect, max[] represents taking the maximum value, AneD is the calculated anesthesia deviation value, and when a real-time intraoperative warning is generated, AneD τ is the anesthesia deviation threshold.

[0028] Preferably, the postoperative data is specifically:

[0029] The electronic medical record system of the integrated hospital and the postoperative physiological monitoring system of the patient collect postoperative data, which at least includes the patient's basic information, postoperative laboratory test results, postoperative imaging examination results, and postoperative vital sign data.

[0030] Preferably, calculate the surgical stability index based on the preoperative data and the intraoperative data, and update the analgesia historical plan based on the surgical stability index to obtain the real-time analgesia plan and the predicted analgesic effect generated by the real-time analgesia plan. Specifically:

[0031] Calculate the surgical stability index using the surgical stability index calculation formula. The surgical stability index calculation formula is:

[0032]

[0033] Among them, ∫|Δp v | is the integral of the change rate of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data numbers, ∫|Δs n | is the integral of the change rate of the postoperative data numbered n, d n is the weight of the postoperative data numbered n obtained by regression analysis, m is the total number of postoperative data numbers, e is a preset distribution parameter for preoperative and intraoperative, and SSI is the calculated preoperative stability index;

[0034] Update the analgesia historical plan using the analgesia historical plan update formula. The analgesia historical plan update formula is:

[0035] Ana k_new =max[f*Δ(SSI norm_1 *Ana k_old )+g*SSI norm_2 *E_Ana k_old

[0036] Among them, Ana k_old is the analgesia historical plan k of the same type of patient obtained by matching, SSI norm_1 is the correction parameter for correcting the analgesia historical plan based on the normalized surgical stability index, Δ(SSI norm_1 *Ana k_old ) is the change rate of the analgesia historical plan k corrected based on the correction parameter, E_Ana k_old is the analgesia historical effect of the analgesia historical plan k, SSI norm_2 is the correction parameter for correcting the analgesia historical effect based on the normalized surgical stability index, f and g are preset weights, max[] represents taking the maximum value, Ana k_new ​For calculating the selected real-time analgesia plan, the corresponding predicted analgesia effect at this time is SSI norm_2 *E_Ana k_old .

[0037] Preferably, the analgesia data is specifically:

[0038] Collect analgesia data using the patient's postoperative physiological monitoring system, and the analgesia data at least includes the quantified value of wound healing, the quantified value of the recovery of physical activity ability, and the pain score.

[0039] Preferably, calculating the analgesia deviation value of the analgesia data compared with the analgesia historical effect and the analgesia predicted effect within the warning period, and when the analgesia deviation value does not meet the preset analgesia deviation threshold, generating a real-time analgesia warning, specifically:

[0040] Calculate the analgesia deviation value using the analgesia deviation value calculation formula, and the analgesia deviation value formula is:

[0041] AnaD = max[Δ(E_Ana k_now , E_Ana k_old ), Δ(E_Ana k_now , SSI norm_2 *E_Ana k_old )]

[0042] Wherein, E_Ana k_now is the real-time analgesia effect obtained based on the analgesia data, Δ(E_Ana k_now , E_Ana k_old ) is the change rate of the real-time analgesia effect compared with the analgesia historical effect, Δ(E_Ana k_now , SSI norm_2 *E_Ana k_old ) is the change rate of the real-time analgesia effect compared with the analgesia predicted effect, max[] represents taking the maximum value, AnaD is the calculated analgesia deviation value, and when a real-time analgesia warning is generated, and AneD τ is the analgesia deviation threshold.

[0043] Preferably, matching the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by the anesthesia historical plan based on the preoperative data, specifically:

[0044] Perform clustering of the same type of patients based on the preoperative data;

[0045] Obtain the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by the anesthesia historical plan;

[0046] Calculate the preoperative data feature distance between the patient and patients of the same type, and screen and output the anesthesia history plans of patients of the same type that meet the preset preoperative data feature distance threshold and the anesthesia history effects generated by the anesthesia history plans; wherein, the preoperative data feature distance is used to characterize the similarity between the patient and patients of the same type.

[0047] The matching of the analgesia history plan of patients of the same type and the analgesia history effect generated by the analgesia history plan based on the postoperative data is specifically as follows:

[0048] Cluster patients of the same type based on the postoperative data.

[0049] Obtain the analgesia history plans of patients of the same type and the analgesia history effects generated by the analgesia history plans.

[0050] Calculate the postoperative data feature distance between the patient and patients of the same type, and screen and output the analgesia history plans of patients of the same type that meet the preset postoperative data feature distance threshold and the analgesia history effects generated by the analgesia history plans; wherein, the postoperative data feature distance is used to characterize the similarity between the patient and patients of the same type.

[0051] Beneficial effects: The patient tracking method based on multi-dimensional data integration and machine learning algorithms in this application realizes the comprehensive tracking management of patients from preoperative to postoperative; by collecting preoperative data, matching historical plans and effects, and calculating the preoperative stability index to update the plan, it can provide a preoperative anesthesia plan recommendation that is more suitable for the patient's own condition, reduce the time and energy of medical staff to collect and sort out data, reduce human negligence errors, and improve the accuracy and personalization of the anesthesia plan; during the operation, by collecting data to calculate the anesthesia deviation value and real-time supervision, it can timely detect abnormal anesthesia effects and issue warnings to ensure the safety of surgical anesthesia; after the operation, based on the preoperative and intraoperative data, calculate the surgical stability index to update the analgesia plan and supervise the analgesia effect, effectively meet the patient's analgesia needs, and overall improve the accuracy, personalization and automation level of medical services, and ensure the patient's safety and rehabilitation quality. Description of the Drawings

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a flow block diagram of the patient tracking method based on multi-dimensional data integration and machine learning algorithms provided by the embodiments of the present application. Detailed Embodiments

[0054] The technical solutions in the embodiments of the present application will be described clearly and completely below. Apparently, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0055] In this article, the term "including" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent in such a process, method, article or device. Without further limitation, the elements defined by the statement "including..." do not exclude the existence of additional identical elements in the process, method, article or device including the said elements.

[0056] This embodiment discloses a Figure 1 patient tracking method based on multi-dimensional data integration and machine learning algorithms as shown in

[0057] S1: Construct a perioperative anesthesia management model; it should be noted that the perioperative anesthesia management model in this embodiment is constructed based on existing multi-dimensional data integration technologies and machine learning algorithms;

[0058] S2: Use the perioperative anesthesia management model to execute preoperative anesthesia plan recommendation;

[0059] S3: Use the perioperative anesthesia management model to execute intraoperative anesthesia effect supervision;

[0060] S4: Use the perioperative anesthesia management model to execute postoperative analgesia plan recommendation;

[0061] S5: Use the perioperative anesthesia management model to execute postoperative analgesia effect supervision;

[0062] The preoperative anesthesia plan recommendation includes: collecting the preoperative data of the patient; matching the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by the anesthesia historical plan based on the preoperative data; calculating the preoperative stability index based on the preoperative data, and updating the anesthesia historical plan based on the preoperative stability index to obtain the anesthesia real-time plan and the anesthesia prediction effect generated by the anesthesia real-time plan; packaging and recommending the anesthesia historical plan and the anesthesia historical effect generated by the anesthesia historical plan, the anesthesia real-time plan and the anesthesia prediction effect generated by the anesthesia real-time plan, and the process of updating the anesthesia historical plan based on the preoperative stability index; wherein, the preoperative stability index is used to characterize the physiological state fluctuation of the patient before surgery, and updating the anesthesia historical plan based on the preoperative stability index is to update the anesthesia operation in the anesthesia historical plan.

[0063] Intraoperative anesthesia effect monitoring includes: collecting intraoperative data generated by implementing the real-time anesthesia plan for the patient; calculating the anesthesia deviation value of the intraoperative data compared with the historical anesthesia effect and the predicted anesthesia effect within a preset warning period. When the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generating a real-time intraoperative warning and displaying the real-time intraoperative warning and the process of calculating the anesthesia deviation value of the intraoperative data compared with the historical anesthesia effect and the predicted anesthesia effect in real time; wherein, the anesthesia deviation value is used to characterize the deviation degree of the patient's real-time anesthesia effect during the operation compared with the historical anesthesia effect and the predicted anesthesia effect.

[0064] Postoperative analgesia plan recommendation includes: collecting the patient's postoperative data; matching the analgesia historical plan of patients of the same type and the analgesia historical effect generated by the analgesia historical plan based on the postoperative data; calculating the surgical stability index based on the preoperative data and the intraoperative data, and updating the analgesia historical plan based on the surgical stability index to obtain the real-time analgesia plan and the predicted analgesia effect generated by the real-time analgesia plan; packaging and recommending the analgesia historical plan and the analgesia historical effect generated by the analgesia historical plan, the real-time analgesia plan and the predicted analgesia effect generated by the real-time analgesia plan, and the process of updating the analgesia historical plan based on the surgical stability index; wherein, the surgical stability index is used to characterize the physiological state fluctuation of the patient before and during the operation, and updating the analgesia historical plan based on the surgical stability index is the analgesia operation in updating the analgesia historical plan based on the surgical stability index.

[0065] Postoperative analgesia effect monitoring includes: collecting analgesia data generated by implementing the real-time analgesia plan for the patient; calculating the analgesia deviation value of the analgesia data compared with the analgesia historical effect and the predicted analgesia effect within the warning period. When the analgesia deviation value does not meet the preset analgesia deviation threshold, generating a real-time analgesia warning and displaying the real-time analgesia warning and the process of calculating the analgesia deviation value of the analgesia data compared with the analgesia historical effect and the predicted analgesia effect in real time; wherein, the analgesia deviation value is used to characterize the deviation degree of the patient's real-time analgesia effect after the operation compared with the analgesia historical effect and the predicted analgesia effect.

[0066] Through the above, this embodiment realizes the comprehensive tracking management of the patient from pre-operation to post-operation; by collecting preoperative data and matching historical plans and effects, calculating the preoperative stability index to update the plan, it can provide a preoperative anesthesia plan recommendation that is more suitable for the patient's own situation, reduce the time and energy of medical staff to collect and sort out data, reduce human negligence and errors, and improve the accuracy and personalization of the anesthesia plan; during the operation, by collecting data to calculate the anesthesia deviation value and monitoring in real time, it can timely detect abnormal anesthesia effects and issue warnings to ensure the safety of surgical anesthesia; after the operation, based on the preoperative and intraoperative data, calculate the surgical stability index to update the analgesia plan, and monitor the analgesia effect, effectively meeting the patient's analgesia needs, and overall improving the accuracy, personalization and automation level of medical services, and ensuring the patient's safety and rehabilitation quality.

[0067] Specifically, the preoperative data is as follows:

[0068] Integrate the electronic case system of the hospital and the patient's preoperative physiological monitoring system to collect preoperative data, which at least includes the patient's basic information, preoperative laboratory test results, preoperative imaging examination results, and preoperative vital sign data.

[0069] By the above, this embodiment uses the integrated existing hospital electronic case system and preoperative physiological monitoring system to collect preoperative data, realizing the comprehensive and accurate acquisition of key data such as the patient's preoperative basic information, examination results, and vital signs, ensuring a reliable data basis for subsequent operations such as matching of the same type of patients based on data and calculation of the preoperative stability index, thereby improving the accuracy and effectiveness of preoperative anesthesia plan recommendation, providing strong support for the entire perioperative anesthesia management, reducing decision-making errors caused by data missing or inaccurate, and further improving the quality of medical services and patient safety.

[0070] Specifically, calculate the preoperative stability index based on the preoperative data, and update the anesthesia historical plan based on the preoperative stability index to obtain the anesthesia real-time plan and the anesthesia prediction effect generated by the anesthesia real-time plan, specifically as follows:

[0071] Use the preoperative stability index calculation formula to calculate the preoperative stability index. The preoperative stability index calculation formula is:

[0072]

[0073] Among them, ∫|Δp v | is the integral of the change rate of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data numbers, and PSI is the calculated preoperative stability index;

[0074] Use the anesthesia historical plan update formula to update the anesthesia historical plan. The anesthesia historical plan update formula is:

[0075] A j_new =max[b*Δ(PSI norm_1 *A j_old )+c*PSI norm_2 *E j_old

[0076] Among them, A j_old is the anesthesia historical plan j of the matched same type of patients, PSI norm_1 is the correction parameter for correcting the anesthesia historical plan based on the preoperative stability index after normalization processing, Δ(PSI norm_1 *A j_old ​) is the change rate of the anesthesia history plan j after correction based on the correction parameters, E j_old is the anesthesia history effect of anesthesia history scheme j, PSI norm_2 is the correction parameter of the preoperative stability index after normalization for correcting the anesthesia history effect, b and c are the preset weights, max[] means to find the maximum value, A j_new To calculate the selected real-time anesthesia plan, the corresponding anesthesia prediction effect is PSI norm_2 *E j_old .

[0077] It should be noted that the normalization process of the preoperative stability index in this embodiment is based on any existing normalization technology, and is intended to convert the preoperative stability index into a quantitative value that can be used to adjust the anesthesia operation based on the common sense known to those skilled in the art. In a specific example, the anesthesia operation in this embodiment may be, but is not limited to, the amount of anesthesia medication, the duration of anesthesia, etc. Similarly, it should be noted that the normalization process of the surgical stability index in the following text adopts the same principle as the normalization process of the preoperative stability index, and this text will not be repeated in the following text.

[0078] Based on the above, this embodiment utilizes the preoperative stability index calculation formula and the anesthesia history plan update formula to realize the reasonable optimization of the anesthesia history plan based on the dynamic changes of the patient's preoperative data. By accurately calculating the preoperative stability index, the patient's preoperative physiological state fluctuations are accurately evaluated, and the anesthesia history plan is updated in combination with the correction parameters. The generated real-time anesthesia plan is more in line with the individual characteristics of the patient, improving the pertinence and adaptability of the anesthesia plan, helping to reduce the anesthesia risk, improve the anesthesia effect, and ensure the smooth progress of the operation. At the same time, it also provides a scientific decision-making basis for medical staff and enhances the reliability of anesthesia management.

[0079] Specifically, the intraoperative data are as follows:

[0080] The hospital's operating room system and the patient's intraoperative physiological monitoring system are integrated to collect intraoperative data, which at least includes the patient's intraoperative vital signs data, anesthetic drug usage data and surgical operation data.

[0081] Based on the above, this embodiment utilizes the integration of existing hospital operating room systems and intraoperative physiological monitoring systems to collect intraoperative data, thereby achieving real-time acquisition of key information such as the patient's intraoperative vital signs, anesthetic drug use, and surgical operations, providing a basis for calculating anesthesia deviation values, thereby being able to promptly detect deviations between intraoperative anesthesia effects and expectations, and promptly issue warnings and display the calculation process, making it easier for medical staff to quickly take measures to adjust the anesthesia plan, ensure the patient's life safety and anesthesia stability during surgery, effectively avoid surgical risks caused by anesthesia problems, and improve surgical success rates and patient safety.

[0082] Specifically, calculate the anesthesia deviation value of the intraoperative data compared with the anesthesia historical effect and the anesthesia predicted effect within a preset warning period. When the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generate a real-time intraoperative warning, specifically:

[0083] Calculate the anesthesia deviation value using the anesthesia deviation value calculation formula. The anesthesia deviation value formula is:

[0084] AneD = max[Δ(E j_now , E j_old ), Δ(E j_now , PSI norm_2 * E j_old )]

[0085] Wherein, E j_now is the real-time anesthesia effect obtained based on the intraoperative data. Δ(E j_now , E j_old ) is the change rate of the real-time anesthesia effect compared with the anesthesia historical effect. Δ(E j_now , PSI norm_2 * E j_old ) is the change rate of the real-time anesthesia effect compared with the anesthesia predicted effect. max[] represents taking the maximum value. AneD is the calculated anesthesia deviation value. When , generate a real-time intraoperative warning. AneD τ is the anesthesia deviation threshold.

[0086] It should be noted that the evaluation of the real-time anesthesia effect in this embodiment adopts the same evaluation method as the evaluation of the anesthesia historical effect. This evaluation method can be any existing anesthesia effect evaluation method, and only needs to keep the evaluation methods of the real-time anesthesia effect and the anesthesia historical effect the same. Similarly, it should be noted that the evaluation of the real-time analgesia effect in this embodiment adopts the same evaluation method as the evaluation of the analgesia historical effect. This evaluation method can be any existing analgesia effect evaluation method, and only needs to keep the evaluation methods of the real-time analgesia effect and the analgesia historical effect the same.

[0087] Through the above, this embodiment uses the anesthesia deviation value calculation formula to achieve precise quantitative evaluation and real-time monitoring of the intraoperative anesthesia effect. By comparing the differences between the intraoperative data and the anesthesia historical effect and the predicted effect, calculate the anesthesia deviation value, and judge whether to issue a warning according to the preset threshold, which can promptly detect anesthesia abnormalities, provide clear decision-making prompts for medical staff, enable them to quickly respond, accurately adjust anesthesia operations, ensure that the patient's anesthesia state is always within the safe and effective range, reduce the occurrence of intraoperative complications, and improve the surgical quality and the patient's rehabilitation effect.

[0088] Specifically, the postoperative data, specifically:

[0089] The hospital's electronic medical record system and the patient's postoperative physiological monitoring system are integrated to collect postoperative data, which at least includes the patient's basic information, postoperative laboratory test results, postoperative imaging test results and postoperative vital signs data.

[0090] Based on the above, this embodiment utilizes the integration of the existing hospital electronic medical record system and postoperative physiological monitoring system to collect postoperative data, thereby realizing the comprehensive collection of the patient's basic information, examination results, vital signs and other data after surgery, which provides the necessary basis for subsequent operations such as matching patients of the same type and calculating the surgical stability index based on postoperative data, and helps to generate more appropriate postoperative analgesia plans, promote postoperative recovery of patients, improve the integrity and effectiveness of medical services, and ensure that patients smoothly pass through the postoperative recovery period.

[0091] Specifically, the surgical stability index is calculated based on the preoperative data and intraoperative data, and the analgesia history plan is updated based on the surgical stability index to obtain the real-time analgesia plan and the analgesia prediction effect generated by the real-time analgesia plan, specifically:

[0092] The surgical stability index was calculated using the surgical stability index calculation formula. The surgical stability index calculation formula is:

[0093]

[0094] Among them, ∫Δp v | is the integral of the rate of change of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data, ∫Δs n | is the integral of the rate of change of postoperative data numbered n, d n is the weight of the postoperative data numbered n obtained by regression analysis, m is the total number of postoperative data, e is the preset preoperative and intraoperative allocation parameter, and SSI is the calculated preoperative stability index;

[0095] The analgesia history plan update formula is used to update the analgesia history plan. The analgesia history plan update formula is:

[0096] Ana k_new =max[f*Δ(SSI norm_1 *Ana k_old )+g*SSI norm_2 *E_Ana k_old ]

[0097] Among them, Ana k_old To match the analgesia history of patients of the same type, SSI norm_1 The surgical stability index is a modified parameter used to modify the analgesic history plan based on normalization. norm_1*Ana k_old ) is the change rate of the analgesia historical plan k after being corrected based on the correction parameter, E_Ana k_old is the analgesia historical effect of the analgesia historical plan k, SSI norm_2 is the correction parameter for correcting the analgesia historical effect based on the normalized surgical stability index, f and g are preset weights, max[] represents taking the maximum value, Ana k_new is the real-time analgesia plan calculated and screened, and the corresponding analgesia prediction effect at this time is SSI norm_2 *E_Ana k_old .

[0098] Through the above, this embodiment uses the surgical stability index calculation formula and the analgesia historical plan update formula to realize the optimization of the analgesia historical plan by comprehensively considering the physiological state fluctuations of the patient before and during the operation. By accurately calculating the surgical stability index, the overall condition of the patient before and after the operation can be comprehensively evaluated, and the analgesia historical plan is updated in combination with the correction parameter, so that the generated real-time analgesia plan better meets the actual needs of the patient, improves the accuracy and personalization of the analgesia effect, reduces the postoperative pain of the patient, promotes the physical recovery, and improves the postoperative quality of life and rehabilitation satisfaction of the patient.

[0099] Specifically, the analgesia data is specifically as follows:

[0100] The analgesia data is collected by using the patient's postoperative physiological monitoring system. The analgesia data at least includes the quantified value of wound healing, the quantified value of the recovery of physical activity ability, and the pain score.

[0101] Through the above, this embodiment uses the existing patient's postoperative physiological monitoring system to collect analgesia data, realizes the effective collection of key information such as postoperative wound healing, physical activity ability, and pain degree of the patient, provides a direct basis for calculating the analgesia deviation value and supervising the analgesia effect, can timely discover the deficiencies of the analgesia plan and make adjustments, ensure that the patient obtains a good analgesia effect after the operation, reduce the adverse effects of pain on the patient's rehabilitation, improve the comfort and rehabilitation confidence of the patient, and promote the recovery of the patient's physical function.

[0102] Specifically, calculate the analgesia deviation value of the analgesia data compared with the analgesia historical effect and the analgesia prediction effect within the warning period. When the analgesia deviation value does not meet the preset analgesia deviation threshold, generate a real-time analgesia warning, specifically as follows:

[0103] Use the analgesia deviation value calculation formula to calculate the analgesia deviation value. The analgesia deviation value formula is:

[0104] AnaD = max[Δ(E_Ana k_now , E_Ana k_old ), Δ(E_Ana k_now , SSInorm_2 *E_Ana k_old )]

[0105] Among them, E_Ana k_now is the real-time analgesic effect obtained based on analgesic data, Δ(E_Ana k_now , E_Ana k_old ) is the change rate of the real-time analgesic effect compared with the historical analgesic effect, Δ(E_Ana k_now , SSI norm_2 *E_Ana k_old ) is the change rate of the real-time analgesic effect compared with the predicted analgesic effect, max[] represents finding the maximum value, AnaD is the calculated analgesic deviation value, when it generates a real-time analgesic warning, and AneD τ is the analgesic deviation threshold.

[0106] Through the above, this embodiment uses the analgesic deviation value calculation formula to achieve accurate evaluation and dynamic management of the postoperative analgesic effect. By comparing the differences between the analgesic data and the historical analgesic effect and the predicted effect, the analgesic deviation value is calculated, and whether to issue a warning is judged according to the preset threshold, which can timely detect the fluctuations of the analgesic effect, provide strong support for medical staff to adjust the analgesic plan, ensure the stability and effectiveness of the patient's postoperative analgesia, is conducive to the patient's physical recovery and mental health, and improves the overall medical service quality.

[0107] Specifically, based on the preoperative data, match the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by this anesthesia historical plan. Specifically:

[0108] Cluster the same type of patients based on the preoperative data;

[0109] Obtain the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by this anesthesia historical plan;

[0110] Calculate the preoperative data feature distance between the patient and the same type of patients, and screen out the anesthesia historical plan of the same type of patients and the anesthesia historical effect generated by this anesthesia historical plan that meet the preset preoperative data feature distance threshold for output; among them, the preoperative data feature distance is used to represent the similarity degree between the patient and the same type of patients;

[0111] Based on the postoperative data, match the analgesic historical plan of the same type of patients and the analgesic historical effect generated by this analgesic historical plan. Specifically:

[0112] Cluster the same type of patients based on the postoperative data;

[0113] Obtain the analgesic historical plan of the same type of patients and the analgesic historical effect generated by this analgesic historical plan;

[0114] Calculate the postoperative data feature distance between the patient and patients of the same type, and screen and output the analgesic historical plans of patients of the same type that meet the preset postoperative data feature distance threshold and the analgesic historical effects generated by the analgesic historical plans; wherein, the postoperative data feature distance is used to characterize the similarity between the patient and patients of the same type.

[0115] It should be noted that the preoperative data feature distance and the postoperative data feature distance in this embodiment can, but are not limited to, calculating the Euclidean distance between the preoperative data vector composed of preoperative data and the postoperative data vector composed of postoperative data.

[0116] Through the above, this embodiment uses clustering and feature distance calculation based on preoperative and postoperative data to accurately screen out the anesthesia and analgesic historical plans and effects of patients of the same type similar to the patient. By calculating the feature distance and screening according to the threshold, the most valuable historical experience data can be found, providing reliable reference for updating the anesthesia and analgesic plans of the current patient, improving the accuracy and effectiveness of plan recommendation, reducing decision-making biases caused by individual differences, and enhancing the personalized level of medical services and patient satisfaction.

[0117] As a preferred implementation manner of this embodiment, this embodiment uses existing visualization technologies to display the analysis and output process of this embodiment, thereby enhancing the interpretability of the perioperative anesthesia management model and providing a basis for doctors to establish trust with the technology.

[0118] In summary, the patient tracking method based on multi-dimensional data integration and machine learning algorithms in this embodiment realizes the comprehensive tracking management of patients from preoperative to postoperative; by collecting preoperative data and matching historical plans and effects, calculating the preoperative stability index to update the plan, it can provide a preoperative anesthesia plan recommendation more suitable for the patient's own situation, reduce the time and energy of medical staff in collecting and sorting data, reduce human negligence errors, and improve the accuracy and personalization of the anesthesia plan; during the operation, by collecting data to calculate the anesthesia deviation value and real-time supervision, it can timely detect abnormal anesthesia effects and issue warnings to ensure the safety of surgical anesthesia; after the operation, based on preoperative and intraoperative data, calculate the surgical stability index to update the analgesic plan and supervise the analgesic effect, effectively meeting the patient's analgesic needs, and overall enhancing the accuracy, personalization and automation level of medical services, and ensuring the safety and rehabilitation quality of patients.

[0119] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, or other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. The computer-readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transmission of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM, or other optical disc storage, magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0120] Finally, it should be noted that the above are only the preferred embodiments of the present application and are not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A patient tracking method based on multidimensional data integration and machine learning algorithm, characterized in that: include: Constructing a perioperative anesthesia management model, and using the perioperative anesthesia management model to perform preoperative anesthesia plan recommendation, intraoperative anesthesia effect supervision, postoperative analgesia plan recommendation, and postoperative analgesia effect supervision; The preoperative anesthesia plan recommendation includes: collecting the patient's preoperative data; matching the anesthesia history plan of the same type of patients and the anesthesia history effect generated by the anesthesia history plan based on the preoperative data; calculating the preoperative stability index based on the preoperative data, and updating the anesthesia history plan based on the preoperative stability index to obtain the anesthesia real-time plan and the anesthesia predicted effect generated by the anesthesia real-time plan; packaging the anesthesia history plan and the anesthesia history effect generated by the anesthesia history plan, the anesthesia real-time plan and the anesthesia predicted effect generated by the anesthesia real-time plan, and the process of updating the anesthesia history plan based on the preoperative stability index for recommendation; wherein the preoperative stability index is used to characterize the fluctuation of the patient's physiological state before the operation, and the updating of the anesthesia history plan based on the preoperative stability index is to update the anesthesia operation in the anesthesia history plan based on the preoperative stability index; The intraoperative anesthesia effect supervision includes: collecting intraoperative data generated by executing the real-time anesthesia program on the patient; calculating the anesthesia deviation value of the intraoperative data compared with the historical anesthesia effect and the predicted anesthesia effect within a preset early warning period, and when the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generating an intraoperative real-time warning, and displaying the intraoperative real-time warning and the process of calculating the anesthesia deviation value of the intraoperative data compared with the historical anesthesia effect and the predicted anesthesia effect within the preset early warning period in real time; wherein, the anesthesia deviation value is used to characterize the degree of deviation of the patient's real-time anesthesia effect during the operation compared with the historical anesthesia effect and the predicted anesthesia effect; The postoperative analgesia plan recommendation includes: collecting the patient's postoperative data; matching the analgesia history plan of the same type of patients and the analgesia history effect generated by the analgesia history plan based on the postoperative data; calculating the surgical stability index based on the preoperative data and the intraoperative data, and updating the analgesia history plan based on the surgical stability index to obtain the analgesia real-time plan and the analgesia prediction effect generated by the analgesia real-time plan; packaging the analgesia history plan and the analgesia history effect generated by the analgesia history plan, the analgesia real-time plan and the analgesia prediction effect generated by the analgesia real-time plan, and the process of updating the analgesia history plan based on the surgical stability index for recommendation; wherein, the surgical stability index is used to characterize the fluctuation of the patient's physiological state before and during surgery, and the updating of the analgesia history plan based on the surgical stability index is to update the analgesia operation in the analgesia history plan based on the surgical stability index; The postoperative analgesia effect supervision includes: collecting analgesia data generated by executing the real-time analgesia plan on the patient; calculating the analgesia deviation value of the analgesia data compared with the historical analgesia effect and the predicted analgesia effect within the early warning period, and when the analgesia deviation value does not meet the preset analgesia deviation threshold, generating an analgesia real-time warning, and displaying the analgesia real-time warning and the process of calculating the analgesia deviation value of the analgesia data compared with the historical analgesia effect and the predicted analgesia effect within the early warning period in real time; wherein, the analgesia deviation value is used to characterize the degree of deviation of the patient's postoperative real-time analgesia effect compared with the historical analgesia effect and the predicted analgesia effect.

2. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The preoperative data are specifically: The electronic medical record system of the hospital and the preoperative physiological monitoring system of the patient are integrated to collect preoperative data, which at least includes the patient's basic information, preoperative laboratory test results, preoperative imaging test results and preoperative vital signs data.

3. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The calculation of the preoperative stability index based on the preoperative data, and the updating of the anesthesia history plan based on the preoperative stability index to obtain the real-time anesthesia plan and the anesthesia prediction effect generated by the real-time anesthesia plan are specifically as follows: The preoperative stability index was calculated using the preoperative stability index calculation formula, which is: Among them, ∫Δp v | is the integral of the rate of change of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data, and PSI is the calculated preoperative stability index; The anesthesia history plan is updated using an anesthesia history plan update formula, and the anesthesia history plan update formula is: A j_new =max[b*Δ(PSI norm_1 *A j_old )+c*PSI norm_2 *E j_old ] Among them, A j_old To match the anesthesia history of patients of the same type, PSI norm_1 PSI is the correction parameter for correcting the anesthesia history plan based on the normalized preoperative stability index. norm_1 *A j_old ) is the change rate of the anesthesia history plan j after correction based on the correction parameters, E j_old is the anesthesia history effect of anesthesia history scheme j, PSI norm_2 is the correction parameter for correcting the anesthesia history effect based on the normalized preoperative stability index, b and c are the preset weights, max[] represents the maximum value, A j_new To calculate the selected real-time anesthesia plan, the corresponding anesthesia prediction effect is PSI norm_2 *E j_old .

4. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The intraoperative data are specifically: The hospital's operating room system and the patient's intraoperative physiological monitoring system are integrated to collect intraoperative data, which at least includes the patient's intraoperative vital signs data, anesthetic drug usage data and surgical operation data.

5. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 3, characterized in that: The calculation of the anesthesia deviation value of the intraoperative data compared with the anesthesia historical effect and the anesthesia predicted effect within the preset warning period, when the anesthesia deviation value does not meet the preset anesthesia deviation threshold, generates an intraoperative real-time warning, specifically: The anesthesia deviation value is calculated using the anesthesia deviation value calculation formula, and the anesthesia deviation value formula is: AneD=max[Δ(E j_now ,AND j_old ),Δ(E j_now ,PSI norm_2 *AND j_old )] Among them, E j_now is the real-time anesthesia effect based on intraoperative data, Δ(E j_now ,E j_old ) is the change rate of the real-time anesthesia effect compared with the historical anesthesia effect, Δ(E j_now ,PSI norm_2 *E j_old ) is the change rate of the real-time anesthesia effect compared to the predicted anesthesia effect, max[] represents the maximum value, and Aned is the calculated anesthesia deviation value. Generate real-time intraoperative warnings, AneD τ The threshold for anesthesia deviation.

6. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The postoperative data are specifically: The hospital's electronic medical record system and the patient's postoperative physiological monitoring system are integrated to collect postoperative data, which at least includes the patient's basic information, postoperative laboratory test results, postoperative imaging test results and postoperative vital signs data.

7. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The calculation of the surgical stability index based on the preoperative data and the intraoperative data, and the updating of the analgesia history scheme based on the surgical stability index to obtain the real-time analgesia scheme and the analgesia prediction effect generated by the real-time analgesia scheme are specifically as follows: The surgical stability index was calculated using a surgical stability index calculation formula, which is: Among them, ∫Δp v | is the integral of the rate of change of the preoperative data numbered v, a v is the weight of the preoperative data numbered v obtained by regression analysis, u is the total number of preoperative data, ∫Δs n | is the integral of the rate of change of postoperative data numbered n, d n is the weight of the postoperative data numbered n obtained by regression analysis, m is the total number of postoperative data, e is the preset preoperative and intraoperative allocation parameter, and SSI is the calculated preoperative stability index; The analgesia history plan is updated using the analgesia history plan update formula, and the analgesia history plan update formula is: Ana k_new =max[f*Δ(SSI norm_1 *Anna k_old )+g*SSI norm_2 *E_Ana k_old ] Among them, Ana k_old To match the analgesic history of patients of the same type, SSI norm_1 The surgical stability index is a modified parameter used to modify the analgesic history plan based on normalization. norm_1 *Ana k_old ) is the rate of change of the analgesic history regimen k after correction based on the correction parameters, E_Ana k_old is the analgesic history effect of the analgesic regimen k, SSI norm_2 is the correction parameter of the surgical stability index after normalization for correcting the analgesic history effect, f and g are the preset weights, max[] means to find the maximum value, Ana k_new To calculate the selected real-time analgesia solution, the corresponding analgesia prediction effect is SSI norm_2 *E_Ana k_old .

8. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 1, characterized in that: The analgesic data are specifically: The analgesia data is collected using the patient's postoperative physiological monitoring system, and the analgesia data at least includes the quantitative value of wound healing, the quantitative value of physical activity recovery and the pain score.

9. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 7, characterized in that: The analgesia deviation value of the analgesia data compared with the analgesia historical effect and the analgesia predicted effect in the warning period is calculated, and when the analgesia deviation value does not meet the preset analgesia deviation threshold, an analgesia real-time warning is generated, specifically: The analgesia deviation value is calculated using the analgesia deviation value calculation formula, and the analgesia deviation value formula is: AnaD=max[Δ(E_Ana k_now ,E_Ana k_old ),Δ(E_Anna k_now ,SSI norm_2 *And_Ana k_old )] Among them, E_Ana k_now is the real-time analgesic effect based on analgesic data, Δ(E_Ana k_now ,E_Ana k_old ) is the change rate of the real-time analgesic effect compared with the historical analgesic effect, Δ(E_Ana k_now ,SSI norm_2 *E_Ana k_old ) is the change rate of the real-time analgesic effect compared to the predicted analgesic effect, max[] represents the maximum value, AnaD is the calculated analgesic deviation value, when Generates real-time analgesia warnings, Aned τ Deviation from the analgesic threshold.

10. The patient tracking method based on multidimensional data integration and machine learning algorithm according to claim 7, characterized in that: The anesthesia history schemes for patients of the same type matched based on the preoperative data and the anesthesia history effects produced by the anesthesia history schemes are specifically: Clustering patients of the same type based on the preoperative data; Obtain the anesthesia history plans of patients of the same type and the anesthesia history effects produced by the anesthesia history plans; Calculate the preoperative data feature distance between the patient and patients of the same type, screen the anesthesia history plans of patients of the same type that meet the preset preoperative data feature distance threshold and the anesthesia history effects produced by the anesthesia history plans for output; wherein the preoperative data feature distance is used to characterize the similarity between the patient and patients of the same type; The analgesia history scheme for patients of the same type and the analgesia history effect produced by the analgesia history scheme based on the postoperative data are specifically: Clustering patients of the same type based on the postoperative data; Obtain the historical analgesic regimens of patients of the same type and the historical analgesic effects produced by the analgesic regimens; The postoperative data feature distance between the patient and patients of the same type is calculated, and the analgesia history plans of patients of the same type that meet the preset postoperative data feature distance threshold and the analgesia history effects produced by the analgesia history plans are screened for output; wherein the postoperative data feature distance is used to characterize the degree of similarity between the patient and patients of the same type.

Images