Adverse drug reaction early warning method, system and equipment and storage medium

Abstract

The invention discloses an adverse drug reaction early warning method, system and device and a storage medium, the adverse drug reaction early warning method comprises the steps that target user medical data and historical user medical data are acquired, and the target user medical data comprise a doctor's advice table, a drug specification and clinical data; extracting a first multi-source data feature of the medical data of the target user; extracting a second multi-source data feature of the historical user medical data; inputting the first multi-source data features into a trained LightGBM prediction model to obtain a first feature correlation value; inputting the second multi-source data features into a trained LightGBM prediction model to obtain a second feature correlation value; based on the second feature correlation degree value, determining a multi-level early warning threshold value through a clustering algorithm; and performing early warning on the adverse drug reaction of the target user based on the first feature association degree value and the multi-level early warning threshold, thereby realizing timely early warning on the adverse drug reaction of the target user.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of drug adverse reaction early warning, and in particular to a drug adverse reaction early warning method, system, device and storage medium. BACKGROUND

[0002] Investigations show that, in addition to normal disease deaths, drug adverse events rank high in the list of non-normal causes of death worldwide.

[0003] At present, most drug adverse reactions are exposed after a large number of people use the drug for a long time after its marketing, which leads to a long monitoring period required to obtain relatively complete drug safety data, and poses a challenge to the effectiveness and safety research of innovative drugs. Therefore, it is particularly necessary to carry out monitoring of drug adverse events and early warning of drug adverse reactions. SUMMARY

[0004] The present application aims to at least solve the technical problems existing in the prior art. To this end, the present application provides a drug adverse reaction early warning method, system, device and storage medium, which can timely early warn the drug adverse reactions of a target user.

[0005] In a first aspect of the present application, a drug adverse reaction early warning method is provided, comprising the following steps:

[0006] Obtaining target user medical data and historical user medical data, wherein the target user medical data includes a medical order form, a drug instruction manual and clinical data;

[0007] Extracting first multi-source data features of the target user medical data; extracting second multi-source data features of the historical user medical data;

[0008] Inputting the first multi-source data features into a trained LightGBM prediction model to obtain a first feature correlation value; inputting the second multi-source data features into the trained LightGBM prediction model to obtain a second feature correlation value;

[0009] Based on the second feature correlation value, determining a multi-level early warning threshold value through a clustering algorithm;

[0010] Based on the first feature correlation value and the multi-level early warning threshold value, early warning the drug adverse reactions of the target user.

[0011] The drug adverse reaction early warning method according to the embodiments of the present application has at least the following beneficial effects:

[0012] This method acquires medical data from target users and historical users, including medical orders, drug instructions, and clinical data. It extracts first multi-source data features from the target user's medical data and second multi-source data features from the historical user's medical data. The first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value. The second multi-source data features are also input into the trained LightGBM prediction model to obtain a second feature correlation value. Based on the second feature correlation value, a clustering algorithm is used to determine multi-level warning thresholds. Finally, based on the first feature correlation value and the multi-level warning thresholds, timely warnings are issued for adverse drug reactions in target users.

[0013] According to some embodiments of this application, the first multi-source data feature of the target user's medical data is extracted, including:

[0014] Extract drug usage time characteristics, drug usage frequency characteristics, drug dosage reduction characteristics, and other characteristics from the target user's medical data;

[0015] The first multi-source data feature is obtained by integrating the drug usage time feature, the drug usage frequency feature, the drug dosage reduction application feature, and the other features.

[0016] According to some embodiments of this application, the first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value, including:

[0017] The first multi-source data features are input into the trained LightGBM prediction model to obtain the correlation between the target user's drug and health damage probability, time correlation, drug usage frequency correlation, drug reduction application correlation and other correlations, among which the health damage probability correlation includes the kidney damage probability correlation.

[0018] The first feature correlation value is calculated based on the health damage probability correlation, the time correlation, the drug usage frequency correlation, the drug dosage reduction correlation, the other correlations, and the preset weight value.

[0019] According to some embodiments of this application, a first feature correlation value is calculated based on the health injury probability correlation, the time correlation, the drug usage frequency correlation, the drug dosage reduction correlation, the other correlations, and a preset weight value, including:

[0020] Based on the correlation degree of health damage probability, the correlation degree of time, the correlation degree of drug use frequency, the correlation degree of drug dosage reduction, other correlation degrees, and preset weight values, the first feature correlation degree value is calculated by a weighted average method.

[0021] According to some embodiments of this application, based on the second feature correlation value, a multi-level early warning threshold is determined by a clustering algorithm, including:

[0022] Based on the second feature correlation value, clustering is performed using a clustering algorithm to obtain the clustering results;

[0023] Based on the clustering results and preset thresholds, association rule mining is performed to obtain the multi-level early warning thresholds.

[0024] According to some embodiments of this application, historical user medical data does not include medical data related to underlying diseases. Before issuing an early warning for adverse drug reactions of the target user based on the first feature correlation value and the multi-level early warning threshold, the method further includes:

[0025] If the target user has underlying medical conditions, the system matches the target user's underlying medical conditions with a preset drug dictionary for underlying medical conditions to obtain a warning result for adverse drug reactions of the target user. The underlying medical conditions include, but are not limited to, at least one of cardiovascular diseases, respiratory diseases, digestive diseases, endocrine diseases, and nervous system diseases.

[0026] According to some embodiments of this application, based on the first feature correlation value and the multi-level early warning threshold, an early warning is issued for adverse drug reactions of the target user, including:

[0027] The first feature correlation value is matched with the multi-level early warning threshold to obtain the adverse drug reaction warning table for the target user. The adverse drug reaction warning table includes, but is not limited to, drug name, drug warning level and adverse drug reaction description. The drug warning level includes low risk, medium risk and high risk.

[0028] Based on the adverse drug reaction warning table and preset warning rules, early warnings are issued for adverse drug reactions of the target user.

[0029] A second aspect of this application provides a drug adverse reaction early warning system, the drug adverse reaction early warning system comprising:

[0030] The data acquisition module is used to acquire medical data of target users and medical data of historical users, wherein the medical data of target users includes medical order forms, drug instructions and clinical data;

[0031] The feature extraction module is used to extract the first multi-source data features of the target user's medical data and the second multi-source data features of the historical user's medical data.

[0032] The feature correlation value output module is used to input the first multi-source data features into the trained LightGBM prediction model to obtain the first feature correlation value; and to input the second multi-source data features into the trained LightGBM prediction model to obtain the second feature correlation value.

[0033] A multi-level early warning threshold determination module is used to determine multi-level early warning thresholds based on the second feature correlation value using a clustering algorithm.

[0034] The early warning module is used to issue early warnings for adverse drug reactions of the target user based on the first feature correlation value and the multi-level early warning threshold.

[0035] This system acquires medical data from target users and historical users, including medical orders, drug instructions, and clinical data. It extracts first multi-source data features from the target user's medical data and second multi-source data features from the historical user's medical data. The first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value. The second multi-source data features are also input into the trained LightGBM prediction model to obtain a second feature correlation value. Based on the second feature correlation value, a clustering algorithm determines multi-level warning thresholds. Finally, based on the first feature correlation value and the multi-level warning thresholds, the system provides timely warnings about adverse drug reactions for target users.

[0036] A third aspect of this application provides an electronic device for early warning of adverse drug reactions, including at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which are executed by the at least one control processor to enable the at least one control processor to perform the above-described method for early warning of adverse drug reactions.

[0037] In a fourth aspect, this application provides a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the aforementioned adverse drug reaction early warning method.

[0038] It should be noted that the beneficial effects of the second to fourth aspects of this application compared with the prior art are the same as the beneficial effects of the aforementioned adverse drug reaction early warning system compared with the prior art, and will not be described in detail here.

[0039] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0040] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0041] Figure 1 This is a flowchart of a method for early warning of adverse drug reactions according to an embodiment of this application;

[0042] Figure 2 This is a schematic diagram of an embodiment of the adverse drug reaction early warning system provided in this application;

[0043] Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application. Detailed Implementation

[0044] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0045] In the description of this application, the use of terms such as "first," "second," etc., is for the purpose of distinguishing technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or the order of the technical features indicated.

[0046] In the description of this application, it should be understood that the orientation descriptions, such as up, down, etc., are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0047] In the description of this application, it should be noted that, unless otherwise explicitly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.

[0048] Studies show that, apart from deaths from normal illnesses, adverse drug events rank first among the causes of unnatural death worldwide.

[0049] Currently, most adverse drug reactions only emerge after prolonged use by a large population following market launch. This results in long monitoring periods required to obtain comprehensive drug safety data, posing a challenge to efficacy and safety studies of innovative drugs. Therefore, monitoring adverse drug events and providing early warning systems for adverse drug reactions are particularly necessary.

[0050] To address the aforementioned technical deficiencies, embodiments of this application provide a method, system, device, and storage medium for early warning of adverse drug reactions.

[0051] Please see Figure 1 This is a flowchart illustrating a method for early warning of adverse drug reactions provided in an embodiment of this application. The method is applied to an electronic device, which may be a server, etc. Figure 1 As shown, the adverse drug reaction early warning method includes:

[0052] Step S101: Obtain the target user's medical data and historical user medical data, wherein the target user's medical data includes medical order forms, drug instructions and clinical data;

[0053] Step S102: Extract the first multi-source data features of the target user's medical data; extract the second multi-source data features of the historical user's medical data;

[0054] Step S103: Input the first multi-source data features into the trained LightGBM prediction model to obtain the first feature correlation value; input the second multi-source data features into the trained LightGBM prediction model to obtain the second feature correlation value;

[0055] Step S104: Based on the second feature correlation value, determine the multi-level early warning threshold through a clustering algorithm;

[0056] Step S105: Based on the first feature correlation value and the multi-level early warning threshold, issue an early warning for adverse drug reactions of the target user.

[0057] This method acquires medical data from target users and historical users, including medical orders, drug instructions, and clinical data. It extracts first multi-source data features from the target user's medical data and second multi-source data features from the historical user's medical data. The first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value. The second multi-source data features are also input into the trained LightGBM prediction model to obtain a second feature correlation value. Based on the second feature correlation value, a clustering algorithm is used to determine multi-level warning thresholds. Finally, based on the first feature correlation value and the multi-level warning thresholds, timely warnings are issued for adverse drug reactions in target users.

[0058] In some embodiments, extracting first multi-source data features from the target user's medical data includes:

[0059] Step S201: Extract the drug usage time characteristics, drug usage frequency characteristics, drug dosage reduction application characteristics, and other characteristics from the target user's medical data;

[0060] Step S202: Integrate drug usage time characteristics, drug usage frequency characteristics, drug dosage reduction characteristics, and other characteristics to obtain the first multi-source data characteristics.

[0061] This application improves the accuracy of early warning by integrating features from multiple data sources, providing data support for subsequent early warnings.

[0062] In some embodiments, the first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value, including:

[0063] Step S301: Input the first multi-source data features into the trained LightGBM prediction model to obtain the correlation degree between the target user's drug and health damage probability, time correlation degree, drug usage frequency correlation degree, drug reduction application correlation degree and other correlation degree, among which, the health damage probability correlation degree includes the kidney damage probability correlation degree.

[0064] Step S302: Calculate the first feature correlation value based on the correlation degree of health damage probability, time correlation degree, drug use frequency correlation degree, drug reduction application correlation degree, other correlation degree and preset weight value.

[0065] Specifically, the other correlations mentioned above can be user age correlations or user gender correlations.

[0066] Specifically, the correlation coefficients for the probability of health damage, time, frequency of drug use, reduced drug use, and other correlation coefficients can be a value between 0 and 1.

[0067] Specifically, the first feature correlation value can be calculated based on the correlation of health damage probability, time correlation, drug use frequency correlation, drug reduction application correlation, other correlations and preset weight values. Each correlation can be pre-set with a corresponding weight value. The first feature correlation value is obtained by multiplying the correlation of health damage probability, time correlation, drug use frequency correlation, drug reduction application correlation, other correlations and corresponding weight values ​​and then adding them together.

[0068] In some embodiments, a first feature correlation value is calculated based on health damage probability correlation, time correlation, drug usage frequency correlation, drug dosage reduction correlation, other correlations, and preset weight values, including:

[0069] Step S401: Based on the correlation degree of health damage probability, time correlation degree, drug use frequency correlation degree, drug reduction application correlation degree, other correlation degree and preset weight value, calculate the first feature correlation degree value by weighted average method.

[0070] Specifically, this application uses a weighted average method to calculate the correlation value of the first feature, providing data support for subsequent early warnings and improving the accuracy of early warnings.

[0071] In some embodiments, multi-level early warning thresholds are determined using a clustering algorithm based on a second feature correlation value, including:

[0072] Step S501: Based on the second feature correlation value, clustering is performed using a clustering algorithm to obtain the clustering results;

[0073] Step S502: Based on the clustering results and preset thresholds, perform association rule mining to obtain multi-level early warning thresholds.

[0074] In some embodiments, historical user medical data does not include medical data related to underlying diseases. Before issuing an early warning for adverse drug reactions of the target user based on a first feature correlation value and a multi-level early warning threshold, the data further includes:

[0075] Step S601: If the target user has underlying diseases, match the target user's underlying diseases with a preset drug dictionary for underlying diseases to obtain the target user's adverse drug reaction warning result. The underlying diseases include, but are not limited to, at least one of cardiovascular diseases, respiratory diseases, digestive diseases, endocrine diseases, and nervous diseases.

[0076] The aforementioned pre-defined basic disease drug dictionary may include, but is not limited to, the names of various basic diseases and the corresponding adverse drug reactions. The adverse drug reaction information includes the drug name, adverse reaction level, and adverse reaction manifestation.

[0077] This application improves the accuracy of early warning by differentiating underlying diseases and providing separate early warnings for adverse drug reactions to these diseases.

[0078] In some embodiments, based on a first feature correlation value and a multi-level warning threshold, an early warning is issued for adverse drug reactions of a target user, including:

[0079] Step S701: Match the first feature correlation value with the multi-level warning threshold to obtain the adverse drug reaction warning table for the target user. The adverse drug reaction warning table includes, but is not limited to, the drug name, drug warning level and adverse drug reaction description. The drug warning level includes low risk, medium risk and high risk.

[0080] Step S702: Based on the adverse drug reaction warning table and preset warning rules, issue early warnings for adverse drug reactions of target users.

[0081] The aforementioned preset warning rules can be used to issue warnings only when the drug warning level is high-risk.

[0082] The above-mentioned approach, based on the adverse drug reaction warning list and preset warning rules, can provide early warnings for adverse drug reactions to target users. This involves screening drugs with high-risk warning levels on the adverse drug reaction warning list and notifying the target users of the corresponding drug name, warning level, and adverse drug reaction description via SMS or other means, thereby achieving the purpose of providing early warnings for adverse drug reactions to target users.

[0083] Additionally, refer to Figure 2 One embodiment of this application provides a drug adverse reaction early warning system, including a data acquisition module 1100, a feature extraction module 1200, a feature correlation value output module 1300, a multi-level early warning threshold determination module 1400, and an early warning module 1500, wherein:

[0084] The data acquisition module 1100 is used to acquire medical data of the target user and medical data of historical users. The medical data of the target user includes medical order forms, drug instructions and clinical data.

[0085] The feature extraction module 1200 is used to extract the first multi-source data features of the target user's medical data; and to extract the second multi-source data features of the historical user's medical data.

[0086] The feature correlation value output module 1300 is used to input the first multi-source data features into the trained LightGBM prediction model to obtain the first feature correlation value; and to input the second multi-source data features into the trained LightGBM prediction model to obtain the second feature correlation value.

[0087] The multi-level early warning threshold determination module 1400 is used to determine the multi-level early warning threshold based on the second feature correlation value and through a clustering algorithm.

[0088] The early warning module 1500 is used to provide early warnings of adverse drug reactions for target users based on the first feature correlation value and multi-level early warning thresholds.

[0089] This system acquires medical data from target users and historical users, including medical orders, drug instructions, and clinical data. It extracts first multi-source data features from the target user's medical data and second multi-source data features from the historical user's medical data. The first multi-source data features are input into a trained LightGBM prediction model to obtain a first feature correlation value. The second multi-source data features are also input into the trained LightGBM prediction model to obtain a second feature correlation value. Based on the second feature correlation value, a clustering algorithm determines multi-level warning thresholds. Finally, based on the first feature correlation value and the multi-level warning thresholds, the system provides timely warnings about adverse drug reactions for target users.

[0090] It should be noted that the system embodiments described above are based on the same inventive concept as the method embodiments described above. Therefore, the relevant content of the method embodiments described above is also applicable to the system embodiments described above, and will not be repeated here.

[0091] Figure 3 A schematic diagram of the rule mining hardware structure provided in an embodiment of this application is shown.

[0092] The adverse drug reaction early warning device may include a processor 301 and a memory 302 storing computer program instructions.

[0093] Specifically, the processor 301 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0094] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 302 is non-volatile solid-state memory.

[0095] In some embodiments, memory 302 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Thus, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0096] The processor 301 reads and executes computer program instructions stored in the memory 302 to implement any of the adverse drug reaction early warning methods in the above embodiments.

[0097] In one example, the adverse drug reaction early warning device may also include a communication interface 303 and a bus 310. For example, Figure 3As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 310 and complete communication with each other.

[0098] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0099] Bus 310 includes hardware, software, or both, that couples components of the adverse drug reaction warning device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 310 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0100] This adverse drug reaction early warning device can execute the adverse drug reaction early warning method in the embodiments of this application based on a three-dimensional design model, thereby achieving a combination of... Figure 1 and Figure 2 The methods and systems for early warning of adverse drug reactions are described.

[0101] Furthermore, in conjunction with the adverse drug reaction early warning method in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the adverse drug reaction early warning methods in the above embodiments.

[0102] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0103] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0104] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0105] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0106] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for early warning of adverse drug reactions, characterized in that, The adverse drug reaction early warning method includes: Acquire target user medical data and historical user medical data, wherein the target user medical data includes medical order forms, drug instructions and clinical data; Extract the first multi-source data features from the target user's medical data; extract the second multi-source data features from the historical user's medical data; The first multi-source data features are input into the trained LightGBM prediction model to obtain the first feature correlation value; the second multi-source data features are input into the trained LightGBM prediction model to obtain the second feature correlation value. Based on the second feature correlation value, a multi-level early warning threshold is determined by a clustering algorithm; Based on the first feature correlation value and the multi-level early warning threshold, an early warning is issued for adverse drug reactions of the target user.

2. The method for early warning of adverse drug reactions according to claim 1, characterized in that, The first multi-source data feature extracted from the target user's medical data includes: Extract drug usage time characteristics, drug usage frequency characteristics, drug dosage reduction characteristics, and other characteristics from the target user's medical data; The first multi-source data feature is obtained by integrating the drug usage time feature, the drug usage frequency feature, the drug dosage reduction application feature, and the other features.

3. The method for early warning of adverse drug reactions according to claim 2, characterized in that, The step of inputting the first multi-source data features into the trained LightGBM prediction model to obtain the first feature correlation value includes: The first multi-source data features are input into the trained LightGBM prediction model to obtain the correlation between the target user's drug and health damage probability, time correlation, drug usage frequency correlation, drug reduction application correlation and other correlations, among which the health damage probability correlation includes the kidney damage probability correlation. The first feature correlation value is calculated based on the health damage probability correlation, the time correlation, the drug usage frequency correlation, the drug dosage reduction correlation, the other correlations, and the preset weight value.

4. The method for early warning of adverse drug reactions according to claim 3, characterized in that, The calculation of the first feature correlation value based on the health damage probability correlation, the time correlation, the drug usage frequency correlation, the drug dosage reduction correlation, the other correlations, and the preset weight value includes: Based on the health damage probability correlation, the time correlation, the drug usage frequency correlation, the drug dosage reduction correlation, the other correlations, and the preset weight value, the first feature correlation value is calculated using a weighted average method.

5. The method for early warning of adverse drug reactions according to claim 1, characterized in that, The step of determining multi-level early warning thresholds based on the second feature correlation value using a clustering algorithm includes: Based on the second feature correlation value, clustering is performed using a clustering algorithm to obtain the clustering results; Based on the clustering results and preset thresholds, association rule mining is performed to obtain the multi-level early warning thresholds.

6. The method for early warning of adverse drug reactions according to claim 1, characterized in that, The historical user medical data does not include medical data related to underlying diseases. Before issuing an early warning for adverse drug reactions of the target user based on the first feature correlation value and the multi-level early warning threshold, the following steps are also included: If the target user has underlying medical conditions, the system matches the target user's underlying medical conditions with a preset drug dictionary for underlying medical conditions to obtain a warning result for adverse drug reactions of the target user. The underlying medical conditions include, but are not limited to, at least one of cardiovascular diseases, respiratory diseases, digestive diseases, endocrine diseases, and nervous system diseases.

7. The method for early warning of adverse drug reactions according to claim 6, characterized in that, The method of issuing early warnings for adverse drug reactions of the target user based on the first feature correlation value and the multi-level early warning threshold includes: The first feature correlation value is matched with the multi-level early warning threshold to obtain the adverse drug reaction warning table for the target user. The adverse drug reaction warning table includes, but is not limited to, drug name, drug warning level and adverse drug reaction description. The drug warning level includes low risk, medium risk and high risk. Based on the adverse drug reaction warning table and preset warning rules, early warnings are issued for adverse drug reactions of the target user.

8. A drug adverse reaction early warning system, characterized in that, The adverse drug reaction early warning system includes: The data acquisition module is used to acquire medical data of target users and medical data of historical users, wherein the medical data of target users includes medical order forms, drug instructions and clinical data; The feature extraction module is used to extract the first multi-source data features of the target user's medical data and the second multi-source data features of the historical user's medical data. The feature correlation value output module is used to input the first multi-source data features into the trained LightGBM prediction model to obtain the first feature correlation value; and to input the second multi-source data features into the trained LightGBM prediction model to obtain the second feature correlation value. A multi-level early warning threshold determination module is used to determine multi-level early warning thresholds based on the second feature correlation value using a clustering algorithm. The early warning module is used to issue early warnings for adverse drug reactions of the target user based on the first feature correlation value and the multi-level early warning threshold.

9. A drug adverse reaction early warning device, characterized in that, It includes at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which, when executed by the at least one control processor, enable the at least one control processor to perform a method for early warning of adverse drug reactions as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions for causing a computer to perform a method for early warning of adverse drug reactions as described in any one of claims 1 to 7.

Images