Drug regulation and control method and system for sepsis patient

Abstract

The invention relates to the technical field of sepsis, in particular to a medicine regulation and control method and system for sepsis patients. Comprises: constructing a drug simulation model based on a historical diagnosis and treatment library; acquiring a feature feedback packet of the user to be treated, and setting a primary regulation strategy according to the feature feedback packet and the drug regulation model; acquiring a monitoring data packet according to a preset feedback time node, and judging whether to correct the first-level regulation and control strategy or not according to the monitoring data packet; rapid drug intervention for a user to be treated is achieved by setting drug sub-strategies of various disease types, meanwhile, a digital twinborn model for the user to be treated is constructed by collecting a real-time feedback data packet of the user to be treated and carrying out optimization fitting on an initial simulation sub-model, refined and real-time regulation and control of drug dosage are achieved, and the treatment effect of the user to be treated is improved. Therefore, the treatment efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sepsis, in particular to a drug regulation method and system for sepsis patients. BACKGROUND

[0002] Sepsis and its more dangerous stage, septic shock, are life-threatening organ dysfunction caused by the body's response to infection, with high morbidity, rapid disease progression, and extremely high mortality. It is one of the main causes of death in intensive care units worldwide.

[0003] Currently, the clinical treatment of sepsis mainly relies on the bundle scheme recommended by the "Sepsis Rescue Movement Guidelines", which includes early broad-spectrum antibiotic use, adequate fluid resuscitation, and necessary vasoactive drug support. However, this standardized strategy faces severe challenges in clinical practice, leading to poor patient outcomes. Specifically, the existing technology has the following significant defects: Lack of precision in drug treatment. Sepsis patients have significant individual heterogeneity, with different pathogens, immune states, and degrees of organ function impairment. However, antibiotic use and fluid resuscitation in existing schemes are mostly based on group guidelines, lacking dynamic regulation basis for individual patient pathophysiological state in real time, easily leading to insufficient or excessive treatment; The use of vasoactive drugs is passive and lagging. Clinically, vasoconstrictor drugs (such as norepinephrine) are usually used after the patient develops refractory hypotension, and dose adjustment is mostly based on intermittent measurement of macroscopic vital signs. This approach cannot prospectively predict the worsening trend of hemodynamics, and it is difficult to achieve fine and real-time regulation of drug dosage, which may miss the critical intervention window. SUMMARY

[0004] The purpose of the present application is to solve the above technical problems, and the present application provides a drug regulation method and system for sepsis patients, aiming to improve the drug regulation accuracy and treatment efficiency of sepsis patients.

[0005] In some embodiments of the present application, by analyzing historical diagnosis and treatment data, multiple disease types and patient types are generated, and by constructing a dual diagnosis and treatment model, drug adaptability for different patient individuals is achieved, improving the drug regulation accuracy and treatment efficiency of sepsis patients.

[0006] In some embodiments of the present application, by setting drug sub-strategies for each disease type, rapid drug intervention for the user to be treated is achieved, and by collecting real-time feedback data packets of the user to be treated, the initial simulation sub-model is optimized and fitted, a digital twin model for the user to be treated is constructed, fine and real-time regulation of drug dosage is achieved, thereby improving treatment efficiency.

[0007] In some embodiments of this application, a method for drug regulation in patients with sepsis is provided, including: A drug simulation model was constructed based on a historical medical records database; Obtain the feature feedback package of the user to be treated, and set the primary regulation strategy based on the feature feedback package and the drug regulation model; The monitoring data packets are obtained according to the preset feedback time nodes, and the primary control strategy is adjusted based on the monitoring data packets. Among them, drug simulation models include: drug regulation models and functional simulation models.

[0008] In some embodiments of this application, the construction of a drug simulation model includes: Multiple disease types are generated based on the historical medical records; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model based on all drug sub-strategies; Multiple patient types are generated based on the historical medical records; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct functional sub-models for the target patient type; Functional sub-models for each patient are generated sequentially, and a functional simulation model is generated based on all functional sub-models.

[0009] In some embodiments of this application, constructing a functional sub-model for the target patient type includes: Generate associated data packages for the target patient type based on the historical medical records; Generate a simulation sub-model and a fluctuation evaluation value f for the target patient type based on the associated data package; f=[ β i *s i ]; Where θ1 represents the number of volatility evaluation indicators; β i s is the influencing factor of the i-th fluctuation evaluation index; i This is the reference value for the i-th fluctuation evaluation index; According to the fluctuation evaluation value f, a fitting period of the target patient type is set; According to the simulation sub-model and the fitting period, a function sub-model of the target patient type is set.

[0010] In some embodiments of the present application, a first regulation strategy is set, including: According to the feature feedback package, a disease feature package of the user to be treated and a patient feature package are generated; The disease feature package and a matching value of each disease type are generated; A drug sub-strategy of the disease type corresponding to the maximum value in all matching values is set as the strategy to be executed; The patient feature package and a similarity value of each patient type are generated; A function sub-model of the patient type corresponding to the maximum value in all similarity values is set as the target function model; According to the target function model and the strategy to be executed, the first regulation strategy is set.

[0011] In some embodiments of the present application, according to the target function model and the strategy to be executed, the first regulation strategy is set, including: According to the target function model, a target fitting period and a first simulation model are generated; A first execution instruction of the strategy to be executed is generated, and a plurality of fitting time nodes are set in the target fitting period according to the first execution instruction; A feedback data package of the user to be treated at the current fitting time node is obtained; According to the feedback data package, a simulation deviation value c of the first simulation model is generated; A simulation deviation threshold C1 is preset; If c>C1, a fitting correction instruction is generated; If c<C1, the current fitting time node does not generate a fitting correction instruction; According to all fitting correction instructions in the target fitting period and the first simulation model, a second simulation model is generated; According to the second simulation model, the first regulation strategy is set.

[0012] In some embodiments of the present application, the first regulation strategy includes: A plurality of regulation time nodes are set; A disease feature package of the user to be treated at the current regulation time node is obtained; According to the disease feature package, a first drug strategy of the current regulation time node is set; According to the second simulation model, a simulation result of the first drug strategy is generated; According to the simulation result, a compensation instruction of the first drug strategy is generated, and a first execution strategy of the current regulation time node is generated according to the compensation instruction.

[0013] In some embodiments of the present application, determining whether to modify the first-level control strategy comprises: obtaining a monitoring data packet of a current feedback time node; generating an expected deviation value d of the current feedback time node according to the monitoring data packet; d=r*[ η i *Y(i)*(k i -k' i )] r=U1*[ η i *(k i -k 1i ) 2 ]; wherein r is a deviation correction coefficient; θ2 is the number of monitoring indicators; η i is an influence factor of the i-th monitoring indicator; k i is a real-time reference value of the i-th monitoring indicator generated based on the monitoring data packet; k' i is a safety threshold of the i-th monitoring indicator; k 1i is an expected reference value of the i-th monitoring indicator generated based on the second-level simulation model; U1 is a preset first conversion coefficient; Y(i) is a selection coefficient; if (k i -k' i )>0, Y(i)=1; if (k i -k' i )<0, Y(i)=0; a preset operating deviation value threshold D1; if d>D1, the current feedback time node generates a modification instruction of the first-level control strategy.

[0014] In some embodiments of the present application, a drug control system for sepsis patients is provided, comprising: a central control unit configured to construct a drug simulation model according to a historical diagnosis and treatment database; the drug simulation model comprises a drug control model and a function simulation model; a monitoring unit comprising a plurality of monitoring sub-modules, the monitoring unit being configured to obtain a feature feedback packet and a monitoring data packet of a user to be treated; the central control unit comprises: a first processing module configured to set a first-level control strategy according to the feature feedback packet and the drug control model; a modification module configured to determine whether to modify the first-level control strategy according to the monitoring data packet.

[0015] In some embodiments of the present application, the central control unit further comprises: a second processing module configured to generate a plurality of disease types according to the historical diagnosis and treatment database; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model based on all drug sub-strategies; Multiple patient types are generated based on the historical medical records; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct functional sub-models for the target patient type; Functional sub-models for each patient are generated sequentially, and a functional simulation model is generated based on all functional sub-models.

[0016] In some embodiments of this application, the first processing module is further configured to: Generate symptom feature packages and patient feature packages for the user to be treated based on the feature feedback packages; Generate symptom feature packages and matching values ​​for each symptom type; Set the drug sub-strategy corresponding to the disease type with the highest value among all matching values ​​as the strategy to be executed; Generate patient feature packages and similarity values ​​for each patient type; The functional sub-model corresponding to the patient type with the maximum value among all similar values ​​is set as the target functional model; A primary control strategy is set based on the target functional model and the strategy to be executed.

[0017] Compared with the prior art, the drug regulation method and system for sepsis patients described in this application have the following advantages: By analyzing historical medical data, various disease types and patient types are generated. By constructing a dual diagnosis and treatment model, drug adaptability for different individual patients can be achieved, thereby improving the precision of drug regulation and treatment efficiency for sepsis patients.

[0018] By setting drug sub-strategies for various disease types, rapid drug intervention can be achieved for patients awaiting treatment. At the same time, by collecting real-time feedback data packets from patients awaiting treatment, the initial simulation sub-model is optimized and fitted to construct a digital twin model for patients awaiting treatment, enabling precise and real-time control of drug dosage, thereby improving treatment efficiency. Attached Figure Description

[0019] Figure 1 is a flowchart of a drug regulation method for a sepsis patient in a preferred embodiment of the present application. DETAILED DESCRIPTION

[0020] The specific embodiments of the present application are described in detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present application, but are not used to limit the scope of the present application.

[0021] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the purpose of facilitating the description of the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0022] The terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "multiple" is two or more.

[0023] In the description of the present application, it should be noted that unless otherwise specified and limited, the terms "mounting", "connection", "connection" should be understood in a broad sense, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0024] As shown in Figure 1 A drug regulation method for a sepsis patient in a preferred embodiment of the present application comprises: S101: constructing a drug simulation model based on a historical diagnosis and treatment library; S102: obtaining a feature feedback package of a user to be treated, and setting a first regulation strategy according to the feature feedback package and the drug regulation model; S103: obtaining a monitoring data package according to a preset feedback time node, and determining whether to correct the first regulation strategy according to the monitoring data package; Among them, the drug simulation model includes: a drug regulation model and a function simulation model.

[0025] Specifically, the historical medical records database includes historical medical data on sepsis (i.e., the severity of sepsis, related medication records, different patient conditions and drug responses, treatment progress and other related parameters).

[0026] Specifically, constructing a drug simulation model includes: Multiple disease types are generated based on the historical medical records; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model based on all drug sub-strategies; Multiple patient types are generated based on the historical medical records; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct functional sub-models for the target patient type; Functional sub-models for each patient are generated sequentially, and a functional simulation model is generated based on all functional sub-models.

[0027] Specifically, multiple disease characteristic indicators are set based on relevant data in the historical medical database. These indicators include, but are not limited to, bodily function parameters related to sepsis, such as body temperature, oxygenation index, abnormal white blood cell count, respiratory rate, blood pressure, and blood lactate levels. By quantifying all disease characteristic indicators, the reference values ​​of each indicator are made to fall within the same range. Multiple value intervals for each disease characteristic indicator are generated sequentially, and various disease types are constructed based on random combinations of all value intervals.

[0028] Specifically, by filtering relevant data in the historical medical records database, all medication records for the target disease type are obtained. Through optimization and analysis of all medication records, a drug sub-strategy for the current disease type is generated. The drug sub-strategy includes (drug category, drug dosage, and injection speed). The drug sub-strategy refers to drug parameters that can provide preliminary control for the current disease type without causing overdose (i.e., for the current disease type, this drug sub-strategy may result in insufficient dosage, but will not result in overdose).

[0029] Specifically, multiple physical function indicators are established based on relevant data from the historical medical database. These indicators include, but are not limited to, body age, weight, blood oxygen saturation, history of sepsis, and presence of underlying diseases (hypertension, hyperlipidemia, and other related conditions) that may affect drug efficacy. By quantifying each physical function indicator, the reference values ​​for each indicator are ensured to fall within the same range. Multiple value intervals for each indicator are then generated sequentially, and various patient types are constructed based on random combinations of these intervals.

[0030] Specifically, constructing functional sub-models for the target patient type includes: Generate associated data packages for the target patient type based on the historical medical records; Generate a simulation sub-model and a fluctuation evaluation value f for the target patient type based on the associated data package; f=[ β i *s i ]; Where θ1 represents the number of volatility evaluation indicators; β i s is the influencing factor of the i-th fluctuation evaluation index; i This is the reference value for the i-th fluctuation evaluation index; The fitting period for the target patient type is set based on the fluctuation evaluation value f; Functional sub-models for target patient types are set based on simulation sub-models and fitting periods.

[0031] Specifically, based on the associated data packets, the number of historical patients corresponding to the target patient type is generated. The effectiveness of the same medication strategy among the historical patients corresponding to the target patient type is analyzed, and a difference evaluation value is generated (taking the first historical user as the standard user, the difference between each user and the standard user is generated, and the corresponding difference evaluation value is generated based on the difference. The greater the difference, the greater the corresponding difference evaluation value. The mapping relationship between the two can be set according to historical parameters).

[0032] Specifically, by analyzing all historical patients in the target patient type, anchor patients (i.e., those with the smallest sum of differences between each historical patient and the current patient) are selected. Based on the analysis of the anchor patients' historical medication data and related physical function data, corresponding simulation sub-models are constructed. These simulation sub-models can simulate the changing trends of various physical functions of patients after the drug enters the body.

[0033] Specifically, multiple fluctuation evaluation indicators are generated based on all difference evaluation values. These fluctuation evaluation indicators include, but are not limited to, parameters related to the dispersion of all difference evaluation values, such as the variance, standard deviation, and difference between the maximum and minimum values. The larger the reference value of each fluctuation evaluation indicator, the greater the dispersion of all difference values, which means that there are greater individual differences in the drug among patients in the target patient type.

[0034] Specifically, the influencing factors of each fluctuation evaluation indicator can be set according to their degree of mapping to individual differences. The greater the degree of mapping, the greater the reference value of the corresponding influencing factor.

[0035] Specifically, the larger the fluctuation evaluation value, the greater the individual differences in drug response among patients in the target patient type, and the longer the corresponding fitting period. The mapping relationship between the two can be set based on historical parameters.

[0036] It is understandable that in the above embodiments, by analyzing historical medical data, multiple disease types and patient types are generated. By constructing a dual diagnosis and treatment model, drug adaptability for different individual patients can be achieved, thereby improving the accuracy of drug regulation and treatment efficiency for sepsis patients.

[0037] In a preferred embodiment of this application, a primary control strategy is set, including: Generate symptom feature packages and patient feature packages for the user to be treated based on the feature feedback packages; Generate symptom feature packages and matching values ​​for each symptom type; Set the drug sub-strategy corresponding to the disease type with the highest value among all matching values ​​as the strategy to be executed; Generate patient feature packages and similarity values ​​for each patient type; The functional sub-model corresponding to the patient type with the maximum value among all similar values ​​is set as the target functional model; A primary control strategy is set based on the target functional model and the strategy to be executed.

[0038] Specifically, the feature feedback package includes various bodily function data of the patient. By analyzing all bodily function data, reference values ​​for various symptom characteristic indicators are generated, thus creating the corresponding symptom feature package. Simultaneously, reference values ​​for various bodily function indicators are generated, thereby constructing the corresponding patient feature package.

[0039] Specifically, by generating the reference values ​​of each symptom feature index in the symptom feature package and the reference values ​​of each symptom feature index corresponding to the current symptom type, a corresponding matching value is set. The larger the sum of the differences, the smaller the corresponding matching value. The mapping relationship between the two can be set according to historical parameters.

[0040] Similarly, by generating the difference between the reference values of each physical function index in the patient feature package and the reference values of each physical function index corresponding to the current patient type, and setting a corresponding similarity value, the larger the sum of the differences, the smaller the corresponding similarity value, and the mapping relationship between the two can be set according to historical parameters.

[0041] Specifically, set the primary regulation strategy according to the target function model and the pending execution strategy, including: Generate the target fitting period and the primary simulation model according to the target function model; Generate the primary execution instruction of the pending execution strategy, and set multiple fitting time nodes within the target fitting period according to the primary execution instruction; Obtain the feedback data packet of the user to be treated at the current fitting time node; Generate the simulation deviation value c of the primary simulation model according to the feedback data packet; Preset the simulation deviation value threshold C1; If c > C1, generate a fitting correction instruction; If c < C1, no fitting correction instruction is generated at the current fitting time node; Generate the secondary simulation model according to all the fitting correction instructions within the target fitting period and the primary simulation model; Set the primary regulation strategy according to the secondary simulation model.

[0042] Specifically, set the fitting period in the target function model as the target fitting period, and the simulation sub-model as the primary simulation model.

[0043] Specifically, perform preliminary drug treatment on the user to be treated according to the primary execution instruction (i.e., perform drug treatment on the user to be treated according to the pending execution strategy), and monitor the physical function data of the user to be treated in real time to generate the corresponding feedback data packet.

[0044] Specifically, set the next fitting time node according to the simulation deviation value of the current fitting time node. The larger the simulation deviation value, the shorter the time interval between the current fitting time node and the next fitting time node.

[0045] Specifically, generate the initial simulation result of the pending execution strategy according to the primary simulation model, and generate the corresponding simulation deviation value by comparing the differences between the initial simulation result and the physical function data in the feedback data packet. The larger the difference, the larger the reference value of the corresponding simulation deviation value, and the mapping relationship between the two can be set according to historical parameters.

[0046] Specifically, the simulation deviation threshold can be set based on historical parameters. When the simulation deviation value exceeds the preset threshold, it indicates that the current Level 1 simulation model has poor accuracy in simulating the drug treatment effect on the patient. It is necessary to continue collecting relevant bodily function data to optimize the Level 1 simulation model based on the Level 1 fitting correction instructions, thereby improving the simulation accuracy of the Level 1 simulation model for the patient.

[0047] Specifically, by making multiple fitting corrections within the target fitting period, a secondary simulation model is generated to improve the simulation accuracy for the patients being treated and to provide data support for subsequent drug regulation decisions.

[0048] In a preferred embodiment of this application, the primary control strategy includes: Set multiple control time points; Obtain the symptom profile of the user to be treated at the current regulatory time point; The primary drug strategy for the current regulatory time point is set based on the symptom profile. Simulation results of the primary drug strategy are generated based on the secondary simulation model; Based on the simulation results, a compensation instruction for the primary drug strategy is generated, and based on the compensation instruction, a primary execution strategy for the current control time node is generated.

[0049] Specifically, after the fitting period ends, multiple control time points are set. Based on the symptom feature package collected at the current control time point, the symptom type of the user to be treated at the current control time point is determined, and the corresponding drug sub-strategy for the symptom type is set as the primary drug strategy. A secondary simulation model is then used to simulate the primary drug strategy, predicting its therapeutic effect on the user. Based on the simulation results, the primary drug strategy is adjusted accordingly, and a corresponding primary execution strategy is generated based on the adjustment results to improve the accuracy of drug control.

[0050] It is understood that in the above embodiments, by setting drug sub-strategies for each disease type, rapid drug intervention for the user to be treated can be achieved. At the same time, by collecting real-time feedback data packets from the user to be treated, the initial simulation sub-model is optimized and fitted to construct a digital twin model for the user to be treated, thereby achieving precise and real-time control of drug dosage and improving treatment efficiency.

[0051] In a preferred embodiment of this application, determining whether to modify the primary control strategy includes: Obtain the monitoring data packet for the current feedback time point; The expected deviation value d for the current feedback time point is generated based on the monitoring data packet; d=r*[ η i *Y(i)*(ki -k' i )] r=U1*[ η i *(k i -k 1i ) 2 ]; Where r is the deviation correction coefficient; θ2 is the number of monitoring indicators; η i Let k be the influencing factor of the i-th monitoring indicator; i It is the real-time reference value of the i-th monitoring indicator generated based on the monitoring data packet; k' i k is the safety threshold for the i-th monitoring indicator; 1i To generate the expected reference value of the i-th monitoring indicator based on the secondary simulation model; U1 is the preset first conversion coefficient; Y(i) is the selection coefficient; if (k i -k' i If (k) > 0, Y(i) = 1; if (k) > 0, Y(i) = 1 i -k' i If ) < 0, then Y(i) = 0; Preset operating deviation threshold D1; If d > D1, the current feedback time node generates a correction instruction for the primary control strategy.

[0052] Specifically, the operational deviation threshold can be set based on historical parameters. When the operational deviation value is greater than the preset operational deviation threshold, it indicates that the current primary control strategy is less effective in treating the user and the primary control strategy needs to be optimized and corrected in a timely manner.

[0053] Specifically, by setting a first conversion coefficient, the deviation correction coefficient is kept within a preset value range, and [ η i *(k i -k 1i ) 2 The larger the value of ], the larger the value of the conversion correction factor. The mapping relationship between the two can be set according to historical parameters, and the reference value of the deviation from the correction factor is always greater than 1.

[0054] Specifically, the monitoring indicators include, but are not limited to: body temperature, oxygenation index, abnormal white blood cell count, respiratory rate, blood pressure, blood lactate level, and other bodily functional parameters associated with sepsis. By quantifying each monitoring indicator, the reference values ​​for all indicators are made to fall within the same range.

[0055] Specifically, the safety thresholds for each monitoring indicator can be set based on historical parameters. If the real-time reference value of the current monitoring indicator is greater than the safety threshold, it indicates that the sepsis of the current patient is worsening and requires early warning and expert intervention.

[0056] Specifically, the expected reference value refers to the value set based on the simulation results of the secondary simulation model for the primary execution strategy. The larger the difference between the real-time reference value and the expected reference value, the worse the simulation accuracy of the current secondary simulation model for the user to be treated, and the worse the overall treatment efficiency. Timely optimization and iteration of the secondary simulation model are necessary.

[0057] In another preferred embodiment of a drug regulation method for sepsis patients based on any of the above preferred embodiments, this preferred embodiment provides a drug regulation system for sepsis patients, comprising: The central control unit is used to build a drug simulation model based on the historical medical records. Drug simulation models include: drug regulation models and functional simulation models; The monitoring unit includes multiple monitoring sub-modules. The monitoring unit is used to acquire the characteristic feedback packets and monitoring data packets of the user to be treated. The central control unit includes: The first processing module is used to set the primary regulation strategy based on the feature feedback package and the drug regulation model. The correction module is used to determine whether to correct the primary control strategy based on the monitoring data packets.

[0058] In a preferred embodiment of this application, the central control unit further includes: The second processing module is used to generate multiple disease types based on the historical diagnosis and treatment database; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model based on all drug sub-strategies; Multiple patient types are generated based on the historical medical records; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct functional sub-models for the target patient type; Functional sub-models for each patient are generated sequentially, and a functional simulation model is generated based on all functional sub-models.

[0059] Specifically, the first processing module is also used for: Generate symptom feature packages and patient feature packages for the user to be treated based on the feature feedback packages; Generate symptom feature packages and matching values ​​for each symptom type; Set the drug sub-strategy corresponding to the disease type with the highest value among all matching values ​​as the strategy to be executed; Generate patient feature packages and similarity values ​​for each patient type; The functional sub-model corresponding to the patient type with the maximum value among all similar values ​​is set as the target functional model; A primary control strategy is set based on the target functional model and the strategy to be executed.

[0060] Based on the first concept of this application, by analyzing historical medical data, multiple disease types and patient types are generated. By constructing a dual diagnosis and treatment model, drug adaptability for different individual patients can be achieved, thereby improving the precision of drug regulation and treatment efficiency for sepsis patients.

[0061] According to the second concept of this application, by setting drug sub-strategies for various disease types, rapid drug intervention for the patient can be achieved. At the same time, by collecting real-time feedback data packets from the patient, the initial simulation sub-model is optimized and fitted to construct a digital twin model for the patient, thereby achieving precise and real-time control of drug dosage and improving treatment efficiency.

[0062] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application.

Claims

1. A method for drug regulation in patients with sepsis, characterized in that, It includes: Construct a drug simulation model based on the historical diagnosis and treatment database; Obtain the characteristic feedback package of the user to be treated, and set the primary regulation strategy according to the characteristic feedback package and the drug regulation model; Obtain the monitoring data package according to the preset feedback time node, and judge whether to correct the primary regulation strategy according to the monitoring data package; Among them, the drug simulation model includes: a drug regulation model and a function simulation model.

2. The drug regulation method for sepsis patients as described in claim 1, characterized in that, Constructing a drug simulation model includes: Generate multiple disease types according to the historical diagnosis and treatment database; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model according to all drug sub-strategies; Generate multiple patient types according to the historical diagnosis and treatment database; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct a function sub-model of the target patient type; Generate function sub-models of each patient in sequence, and generate a function simulation model according to all function sub-models.

3. The drug regulation method for sepsis patients as described in claim 2, characterized in that, Constructing a function sub-model of the target patient type includes: Generate the associated data package of the target patient type according to the historical diagnosis and treatment database; Generate a simulation sub-model and the fluctuation evaluation value f of the target patient type according to the associated data package; f=[ b i *s i ]; Where θ1 represents the number of volatility evaluation indicators; β i s is the influencing factor of the i-th fluctuation evaluation index; i This is the reference value for the i-th fluctuation evaluation index; Set the fitting period of the target patient type according to the fluctuation evaluation value f; Set the function sub-model of the target patient type according to the simulation sub-model and the fitting period.

4. The drug regulation method for sepsis patients as described in claim 3, characterized in that, Setting the primary regulation strategy includes: Generate the disease characteristic package and patient characteristic package of the user to be treated according to the characteristic feedback package; Generate the matching values between the disease characteristic package and each disease type; Set the drug sub-strategy of the disease type corresponding to the maximum value among all matching values as the to-be-executed strategy; Generate the similarity values between the patient characteristic package and each patient type; Set the function sub-model of the patient type corresponding to the maximum value among all similarity values as the target function model; Set the primary regulation strategy according to the target function model and the to-be-executed strategy.

5. The drug regulation method for sepsis patients as described in claim 4, characterized in that, Setting the primary regulation strategy according to the target function model and the to-be-executed strategy includes: Generate the target fitting period and the primary simulation model according to the target function model; Generate the primary execution instruction of the to-be-executed strategy, and set multiple fitting time nodes within the target fitting period according to the primary execution instruction; Obtain the feedback data package of the user to be treated at the current fitting time node; Generate the simulation deviation value c of the primary simulation model according to the feedback data package; Preset the simulation deviation value threshold C1; If c > C1, generate a fitting correction instruction; If c < C1, no fitting correction instruction is generated at the current fitting time node; Generate a secondary simulation model according to all fitting correction instructions within the target fitting period and the primary simulation model; Set the primary regulation strategy according to the secondary simulation model.

6. The drug regulation method for sepsis patients as described in claim 5, characterized in that, The primary regulation strategy includes: Set multiple regulation time nodes; Obtain the disease characteristic package of the user to be treated at the current regulation time node; Set the primary drug strategy at the current regulation time node according to the disease characteristic package; Generate the simulation result of the primary drug strategy according to the secondary simulation model; Generate the compensation instruction of the primary drug strategy according to the simulation result, and generate the primary execution strategy at the current regulation time node according to the compensation instruction.

7. The drug regulation method for sepsis patients as described in claim 6, characterized in that, Judging whether to correct the primary regulation strategy includes: Obtain the monitoring data package at the current feedback time node; Generate the expected deviation value d at the current feedback time node according to the monitoring data package; d=r*[ η i *Y(i)*(k i -k' i )] r=U1*[ η i *(k i -k 1i ) 2 ]; Where r is the deviation correction coefficient; θ2 is the number of monitoring indicators; η i Let k be the influencing factor of the i-th monitoring indicator; i It is the real-time reference value of the i-th monitoring indicator generated based on the monitoring data packet; k' i k is the safety threshold for the i-th monitoring indicator; 1i To generate the expected reference value of the i-th monitoring indicator based on the secondary simulation model; U1 is the preset first conversion coefficient; Y(i) is the selection coefficient; if (k i -k' i If (k) > 0, Y(i) = 1; if (k) > 0, Y(i) = 1; i -k' i If ) < 0, then Y(i) = 0; Preset the operation deviation value threshold D1; If d > D1, the current feedback time node generates a correction instruction for the primary control strategy.

8. A drug regulation system for sepsis patients, employing the drug regulation method for sepsis patients according to any one of claims 1-7, characterized in that, include: The central control unit is used to build a drug simulation model based on the historical medical records. The drug simulation model includes: a drug regulation model and a functional simulation model; The monitoring unit includes multiple monitoring sub-modules, and the monitoring unit is used to acquire the feature feedback packets and monitoring data packets of the user to be treated. The central control unit includes: The first processing module is used to set the primary regulation strategy based on the feature feedback package and the drug regulation model. The correction module is used to determine whether to correct the primary control strategy based on the monitoring data packets.

9. The drug regulation system for sepsis patients as described in claim 8, characterized in that, The central control unit also includes: The second processing module is used to generate multiple disease types based on the historical diagnosis and treatment database; Establish a sequence of disease types A, A=(a1,a2…a ... i …a n ), where a i Let be the i-th disease type; n is the number of disease types; Generate drug sub-strategies for each disease type in sequence; Construct a drug regulation model based on all drug sub-strategies; Multiple patient types are generated based on the historical medical records; Establish a patient type sequence B, B = (b1, b2, ..., bb) i …b m ), where b i Let m be the i-th patient type; m is the number of patient types. b is set sequentially according to the patient type sequence B. i For the target patient type; Construct functional sub-models for target patient types; Functional sub-models for each patient are generated sequentially, and a functional simulation model is generated based on all functional sub-models.

10. The drug regulation system for sepsis patients as described in claim 9, characterized in that, The first processing module is also used for: Generate symptom feature packages and patient feature packages for the user to be treated based on the feature feedback packages; Generate symptom feature packages and matching values ​​for each symptom type; Set the drug sub-strategy corresponding to the disease type with the highest value among all matching values ​​as the strategy to be executed; Generate patient feature packages and similarity values ​​for each patient type; The functional sub-model corresponding to the patient type with the maximum value among all similar values ​​is set as the target functional model; A primary control strategy is set based on the target functional model and the strategy to be executed.

Images