Pediatric drug adverse reaction prediction method and system based on deep learning

Abstract

The present application relates to the field of artificial intelligence, and discloses a method for predicting adverse drug reactions in pediatrics based on deep learning. The method includes: determining a first health status feature and a first adverse reaction feature according to the health status feature of the target patient and the characteristics of the drug to be used through an adverse reaction prediction model; determining a second health feature from the first health feature in time series, and determining a second adverse reaction feature associated with the second health feature from the first adverse reaction feature in time series, and determining an adverse reaction exclusion feature through an adverse reaction exclusion model; when the adverse reaction exclusion feature meets a preset condition, outputting a prediction result that the target patient will not have an adverse reaction to the drug to be used; when the adverse reaction exclusion feature does not meet the preset condition, outputting a prediction result that the target patient will have an adverse reaction to the drug to be used. The present application improves the accuracy of pediatric drug adverse reaction prediction through disease condition data.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a method and system for predicting adverse drug reactions in pediatrics based on deep learning. Background Art

[0002] Adverse drug reaction (ADR) refers to harmful and unexpected reactions that occur when normal doses of drugs are used to prevent, diagnose, treat diseases or regulate physiological functions. Pediatric adverse drug reactions are a common problem in clinical medicine. The occurrence of pediatric adverse drug reactions is closely related to many factors such as the age, weight, gender, and underlying diseases of the children. The traditional judgment of adverse drug reactions mainly relies on the doctor's clinical experience and drug instructions. This method is highly subjective and has low accuracy, and it is difficult to meet the actual clinical needs.

[0003] Patent CN109346145B (application number: 201811200585.6) provides an active monitoring method for adverse drug reactions, which obtains the type of drug to be monitored and its drug instructions, and searches for the drug to be monitored in the doctor's prescription medication data, and screens out the medical condition data corresponding to the drug to be monitored. Then, according to the adverse reaction description information in the drug instructions, a text monitoring rule is established for the condition data, and the text monitoring rule is used to monitor whether there are suspicious results of adverse reactions in the condition data, and monitor the signs of adverse reactions that may occur in the patient during the medication process and the suspected adverse reactions. The method in patent CN109346145B only uses text monitoring rules to monitor the condition data to predict adverse reactions, which is limited by the incompleteness of the condition data, resulting in a low accuracy rate in predicting adverse reactions. Summary of the invention

[0004] The purpose of this application is to provide a method and system for predicting adverse drug reactions in pediatrics based on deep learning, which solves the technical problem of low accuracy in predicting adverse reactions due to incomplete medical data, and achieves the technical effect of improving the accuracy of predicting adverse drug reactions in pediatrics through medical data.

[0005] The embodiment of the present application provides a method for predicting adverse drug reactions in pediatrics based on deep learning, the method comprising: obtaining health status characteristics and characteristics of drugs to be used of a target patient; determining, through an adverse reaction prediction model, a first health status characteristic and a first adverse reaction characteristic of the target patient in time series after using the drug according to the health status characteristics and characteristics of the drug to be used of the target patient; determining a second health characteristic from the first health characteristic in time series, and determining a second adverse reaction characteristic associated with the second health characteristic from the first adverse reaction characteristic in time series, and determining an adverse reaction exclusion characteristic according to the characteristics of the drug to be used, the second health characteristic and the second adverse reaction characteristic through an adverse reaction exclusion model; when the adverse reaction exclusion characteristic meets a preset condition, outputting a prediction result that the target patient will not have an adverse reaction to the drug to be used; when the adverse reaction exclusion characteristic does not meet the preset condition, outputting a prediction result that the target patient will have an adverse reaction to the drug to be used.

[0006] In one possible implementation, a second health feature is determined from a first health feature in a time series, and a second adverse reaction feature associated with the second health feature is determined from a first adverse reaction feature in a time series, including: determining a second health feature that meets a first preset condition from the first health feature in a time series; determining a first time period corresponding to the second health feature, and obtaining a second adverse reaction feature within the first time period; wherein the second adverse reaction feature is a sub-feature of the first adverse reaction feature in a time series; when the second adverse reaction feature meets the second preset condition, outputting the second health feature and the second adverse reaction feature; when the second adverse reaction feature does not meet the second preset condition, excluding the second health feature and the second adverse reaction feature.

[0007] In another possible implementation, determining the first time period corresponding to the second health characteristic includes: determining the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic; based on the relative time relationship, determining a first time period that partially overlaps or completely overlaps with the time period where the second health characteristic is located, or determining a first time period after the time period where the second health characteristic is located, and determining the length of the first time period.

[0008] In another possible implementation, determining the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic includes: collecting case information of adverse reactions to medication of children, determining the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic based on the case information of adverse reactions to medication of children; wherein the case information of adverse reactions to medication of children includes the time when the children take medication; determining the average value of the time correspondence between the time when the children take medication and the time of the adverse reaction as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

[0009] In another possible implementation, determining the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic also includes: determining the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic based on the adverse reaction case information of the child's medication, and determining the health status adjustment factor based on the second health characteristic; when the adverse reaction time is after the time period where the second health characteristic is located, the product of the duration between the adverse reaction time and the time period where the second health characteristic is located and the adjustment factor, and the product of the duration of the adverse reaction time and the adjustment factor are used as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

[0010] In another possible implementation, determining the relative time relationship between the second health feature and the second adverse reaction feature further includes: when the adverse reaction time is less than or equal to a preset time period, excluding the adverse reaction time and the second health feature.

[0011] An embodiment of the present application also provides a pediatric drug adverse reaction prediction system based on deep learning, including a unit for executing the method described above.

[0012] An embodiment of the present application also provides a pediatric drug adverse reaction prediction system based on deep learning, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method described above when executing the computer program.

[0013] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0014] An embodiment of the present application also provides a computer program product, including a computer program, which implements the steps of the method described above when executed by a processor.

[0015] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0016] An embodiment of the present application provides a method for predicting adverse drug reactions in pediatrics based on deep learning, the method comprising: obtaining health status characteristics and characteristics of drugs to be used of a target patient; determining, through an adverse reaction prediction model, a first health status characteristic and a first adverse reaction characteristic in time series of the target patient after the target patient uses the drug according to the health status characteristics and characteristics of the drug to be used of the target patient; determining a second health characteristic from the first health characteristic in time series, and determining a second adverse reaction characteristic associated with the second health characteristic from the first adverse reaction characteristic in time series, and determining an adverse reaction exclusion characteristic according to the second health characteristic and the second adverse reaction characteristic through an adverse reaction exclusion model; when the adverse reaction exclusion characteristic meets a preset condition, outputting a prediction result that the target patient will not have an adverse reaction to the drug to be used; when the adverse reaction exclusion characteristic does not meet the preset condition, outputting a prediction result that the target patient will have an adverse reaction to the drug to be used. The embodiment of the present application can first predict the health status characteristics and adverse reaction characteristics of the child in time series after taking the medicine, and then exclude the health status characteristics and adverse reaction characteristics within a specific time range of the health status characteristics and adverse reaction characteristics in time series through an adverse reaction exclusion model, thereby improving the accuracy of predicting the adverse reactions of the child and improving the prediction effect of the adverse reactions of the child. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] Figure 1 A schematic diagram of a process for predicting adverse drug reactions in pediatrics based on deep learning provided in an embodiment of the present application;

[0019] Figure 2 A schematic diagram of the working process of a method for predicting adverse drug reactions in pediatrics based on deep learning provided in an embodiment of the present application;

[0020] Figure 3 A schematic diagram of the logical structure of a pediatric drug adverse reaction prediction system based on deep learning provided in an embodiment of the present application;

[0021] Figure 4 A schematic diagram of the entity structure of a pediatric drug adverse reaction prediction system based on deep learning provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0023] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0024] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0025] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0026] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0027] In the prior art, the method for predicting adverse reactions to children's medications only uses text monitoring rules to monitor disease data to predict adverse reactions. Due to the incompleteness of disease data, the accuracy of adverse reaction prediction is low.

[0028] Based on the above reasons, an embodiment of the present application provides a pediatric drug adverse reaction prediction method based on deep learning, and the method includes: obtaining health status characteristics and characteristics of drugs to be used for the target child; determining, through an adverse reaction prediction model, the first health status characteristics and the first adverse reaction characteristics of the target child in time series after using the drug according to the health status characteristics and characteristics of the drug to be used for the target child; determining the second health characteristic from the first health characteristic in time series, and determining the second adverse reaction characteristic associated with the second health characteristic from the first adverse reaction characteristic in time series, and determining the adverse reaction exclusion characteristic according to the second health characteristic and the second adverse reaction characteristic through an adverse reaction exclusion model; when the adverse reaction exclusion characteristic meets the preset conditions, outputting a prediction result that the target child will not have an adverse reaction to the drug to be used; when the adverse reaction exclusion characteristic does not meet the preset conditions, outputting a prediction result that the target child will have an adverse reaction to the drug to be used. The embodiment of the present application can first predict the health status characteristics and adverse reaction characteristics of the child in time series after taking the medicine, and then exclude the health status characteristics and adverse reaction characteristics within a specific time range of the health status characteristics and adverse reaction characteristics in time series through an adverse reaction exclusion model, thereby improving the accuracy of predicting the adverse reactions of the child and improving the prediction effect of the adverse reactions of the child.

[0029] In some scenarios, a pediatric drug adverse reaction prediction method based on deep learning in an embodiment of the present application can be applied to the prediction of pediatric drug adverse reactions. It can predict adverse reactions to medication based on children's different physical health conditions and improve the safety of children's medication.

[0030] The following is a detailed description of a method for predicting adverse drug reactions in pediatrics based on deep learning provided in an embodiment of the present application with reference to specific examples.

[0031] Figure 1 A flowchart of a method for predicting adverse drug reactions in pediatrics based on deep learning is provided in an embodiment of the present application, such as Figure 1 As shown, the method includes S110 to S140, and S110 to S140 are described in detail below.

[0032] S110. Obtain the health status characteristics of the target child and the characteristics of the drugs to be used.

[0033] When predicting adverse reactions, predictions can be made based on the physical health conditions of different children, so the health status characteristics of the target children and the characteristics of the drugs to be used can be obtained first.

[0034] Exemplarily, health status characteristics may include the child's allergic reaction characteristics, underlying disease characteristics, past medical history characteristics, age, weight, body temperature, physical discomfort and other health status indicators.

[0035] Exemplarily, the characteristics of the drug to be used may include the content of adverse reaction records, pharmacological characteristics, dosage characteristics, and other characteristics of the drug to be used by the child.

[0036] S120. Determine, through the adverse reaction prediction model, the first health status characteristic and the first adverse reaction characteristic of the target child in time sequence after the target child uses the drug according to the health status characteristics of the target child and the characteristics of the drug to be used.

[0037] Figure 2 A schematic diagram of the working process of a method for predicting adverse drug reactions in pediatrics based on deep learning provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, when predicting adverse reactions, predictions can be made through an adverse reaction prediction model, and in order to predict the adverse reactions of the children, the first health status characteristics and the first adverse reaction characteristics of the target children after using the drugs can be determined based on the health status characteristics of the target children and the characteristics of the drugs to be used. The first health status characteristics and the first adverse reaction characteristics correspond to the continuous health status characteristics and adverse reaction characteristics of the children within a period of time from the start of medication.

[0038] For example, the adverse reaction prediction model can be obtained by training a neural network model with the collected health status characteristics of the child, characteristics of the drug to be used, and the continuous health status characteristics and adverse reaction characteristics of the child within a period of time after taking the drug.

[0039] S130. Determine the second health feature from the first health feature in the time series, and determine the second adverse reaction feature associated with the second health feature from the first adverse reaction feature in the time series, and determine the adverse reaction exclusion feature according to the characteristics of the drug to be used, the second health feature and the second adverse reaction feature through the adverse reaction exclusion model.

[0040] like Figure 2 As shown, after predicting the adverse reaction and obtaining the prediction result, the second health feature can be determined from the first health feature in the time series, and the second adverse reaction feature associated with the second health feature can be determined from the first adverse reaction feature in the time series, and the adverse reaction exclusion feature can be determined according to the characteristics of the drug to be used, the second health feature and the second adverse reaction feature through the adverse reaction exclusion model. The adverse reaction exclusion feature can be used to exclude the adverse reaction based on whether the child's medication, the child's health characteristics and the adverse reaction are related.

[0041] Exemplarily, when determining the second health characteristic from the first health characteristic, the screening conditions for the second health characteristic may be conditions of changes in the patient's physical health, such as changes in body temperature. By determining the second health characteristic from the first health characteristic, the adverse reaction prediction results of the patient can be further identified through the second health characteristic.

[0042] Exemplarily, the second adverse reaction feature associated with the second health feature may be a feature corresponding to adverse reaction symptoms such as physical discomfort of the child. The second adverse reaction feature can be combined with the second health feature and the second adverse reaction feature to predict the adverse reaction of the child.

[0043] Exemplarily, when training an adverse reaction exclusion model, a large amount of drug usage characteristics, health characteristics, and adverse reaction characteristics from children's medication case data can be collected, and a neural network model can be trained based on the drug usage characteristics, health characteristics, adverse reaction characteristics, and adverse reaction exclusion characteristics from the children's medication case data to obtain an adverse reaction exclusion model for identifying and eliminating adverse reactions.

[0044] S140: When the adverse reaction exclusion feature meets the preset conditions, the prediction result that the target patient will not have adverse reactions to the drug to be used is output. When the adverse reaction exclusion feature does not meet the preset conditions, the prediction result that the target patient will have adverse reactions to the drug to be used is output.

[0045] like Figure 2 As shown, after obtaining the adverse reaction exclusion feature, when the adverse reaction exclusion feature meets the preset conditions, the prediction result that the target child will not have an adverse reaction to the drug to be used is output, and the second health feature and the second adverse reaction feature can be judged as the adverse reaction of the child to the drug to be used.

[0046] For example, when outputting a prediction result that the target child will not have an adverse reaction to the drug to be used, the second health characteristic and the second adverse reaction characteristic may be used as the adverse reaction of the target child to the drug to be used.

[0047] After obtaining the adverse reaction exclusion feature, when the adverse reaction exclusion feature does not meet the preset conditions, the prediction result of the target child's adverse reaction to the drug is output, indicating that the second health feature and the second adverse reaction feature produced by the child at this time are not adverse reactions to the drug to be used by the target child.

[0048] The beneficial effect brought about by the above implementation method is that the adverse reaction of the child is predicted by the first health status feature in time series and the first adverse reaction feature in time series, thereby improving the accuracy of the adverse reaction prediction.

[0049] The beneficial effect brought about by the above-mentioned implementation method is that after obtaining the first health status characteristic in the time series and the first adverse reaction characteristic in the time series, by identifying the health status and adverse reaction of the key points in the time series and obtaining the second health characteristic and the second adverse reaction characteristic associated with the second health characteristic, the characteristics of the drug to be used, the second health characteristic and the second adverse reaction characteristic associated with the second health characteristic can be identified and excluded, thereby avoiding inaccurate prediction results and improving the accuracy of predicting adverse reactions of children.

[0050] In some implementations, in the above-mentioned S130, the second health feature is determined from the first health feature in time series, and the second adverse reaction feature associated with the second health feature is determined from the first adverse reaction feature in time series, including S131 to S133, and S131 to S133 are specifically described below.

[0051] S131 . Determine a second health feature that meets a first preset condition from the first health features in the time series.

[0052] When determining the second health feature from the first health feature in the time series, the second health feature can be screened according to specific conditions. For example, the second health feature that meets the first preset condition can be determined from the first health feature in the time series. The first preset condition can be a preset condition regarding increased body temperature, decreased body temperature, etc.

[0053] S132: Determine a first time period corresponding to the second health feature, and obtain a second adverse reaction feature in the first time period, wherein the second adverse reaction feature is a sub-feature of the first adverse reaction feature in terms of time sequence.

[0054] When determining the second adverse reaction feature associated with the second health feature from the first adverse reaction feature in time series, the second adverse reaction feature can be obtained according to the time correspondence. Specifically, the first time period corresponding to the second health feature can be determined, and the second adverse reaction feature within the first time period can be obtained, and the second adverse reaction feature is a sub-feature of the first adverse reaction feature in time series.

[0055] After obtaining the second health characteristic and the second adverse reaction characteristic, the adverse reaction of the child can be further predicted and excluded based on the second health characteristic and the second adverse reaction characteristic.

[0056] S133: When the second adverse reaction feature meets the second preset condition, output the second health feature and the second adverse reaction feature. When the second adverse reaction feature does not meet the second preset condition, exclude the second health feature and the second adverse reaction feature.

[0057] During operation, when the second adverse reaction characteristic meets the second preset condition, the second health characteristic and the second adverse reaction characteristic can be output, and the second health characteristic and the second adverse reaction characteristic can be used for subsequent processing.

[0058] Exemplarily, the second preset condition may be that the correlation between the second adverse reaction characteristic and the second health characteristic reaches a preset threshold, or that the adverse reaction symptoms corresponding to the second adverse reaction characteristic meet the symptom requirements of a mild adverse reaction or above.

[0059] During operation, when the second adverse reaction characteristic does not meet the second preset condition, the second health characteristic and the second adverse reaction characteristic can be excluded, thereby avoiding the use of the second health characteristic and the second adverse reaction characteristic that do not meet the requirements for adverse reaction prediction, thereby ensuring the accuracy of the second health characteristic and the second adverse reaction characteristic.

[0060] The beneficial effect brought about by the above-mentioned implementation method is that when determining the second health feature from the first health feature in time series, the second health feature can be screened according to specific conditions, thereby improving the accuracy of identifying the second health feature and improving the accuracy of adverse reaction prediction.

[0061] The beneficial effect brought about by the above-mentioned implementation method is that when determining the second health characteristic and the second adverse reaction characteristic, the second health characteristic and the second adverse reaction characteristic can be output when the second adverse reaction characteristic meets the second preset condition, thereby improving the accuracy of adverse reaction prediction.

[0062] In some implementations, in the above S132, determining the first time period corresponding to the second health characteristic includes S132a to S132b, and S132a to S132b are described in detail below.

[0063] S132a. Determine the relative time relationship between the second adverse reaction feature and the second health feature.

[0064] When obtaining the second adverse reaction characteristic corresponding to the second health characteristic, different adverse reactions will have different manifestations in the physical health characteristics before or after the adverse reaction. Therefore, the second adverse reaction characteristics corresponding to different second health characteristics will have different relative relationships in time. By determining the relative time relationship of the second adverse reaction characteristics corresponding to the second health characteristics, more accurate second health characteristics can be determined for different second adverse reaction characteristics, thereby improving the prediction accuracy of adverse reactions.

[0065] Exemplarily, the relative time relationship may be preset or extracted through clinical experience. For example, the relative time relationship between the adverse reaction of vomiting symptoms and the health characteristics may be: the child may have had the health characteristic of elevated body temperature within an hour before the adverse reaction of vomiting symptoms.

[0066] S132b. Determine, based on the relative time relationship, a first time period that partially or completely overlaps with the time period where the second health characteristic is located, or determine a first time period after the time period where the second health characteristic is located, and determine the duration of the first time period.

[0067] After obtaining the relative time relationship, the first time period corresponding to the time period where the second health characteristic is located can be further determined based on the relative time relationship. The first time period is the time period where the second adverse reaction characteristic corresponding to the second health characteristic is located. By determining the first time period, the second health characteristic and the second adverse reaction characteristic can be accurately identified.

[0068] When determining the first time period corresponding to the second health characteristic based on the relative time relationship, it is possible to determine the first time period that partially overlaps or completely overlaps with the time period where the second health characteristic is located, or determine the first time period after the time period where the second health characteristic is located, and determine the length of the first time period.

[0069] Exemplarily, the duration of the first time period is also determined based on clinical experience.

[0070] The beneficial effect brought about by the above-mentioned implementation method is that it can improve the accuracy and scientificity of determining the second adverse reaction characteristics according to the relative time relationship of the second adverse reaction characteristics corresponding to different second health characteristics, thereby improving the prediction effect of adverse reactions.

[0071] In some implementations, in the above S132a, determining the relative time relationship between the second health feature and the second adverse reaction feature includes S132a1 to S132a2, and S132a1 to S132a2 are specifically described below.

[0072] S132a1. Collect the case information of adverse drug reaction of the child, and determine the second health characteristic and the adverse drug reaction time of the second adverse drug reaction corresponding to the second health characteristic according to the case information of adverse drug reaction of the child. The case information of adverse drug reaction of the child includes the drug use time of the child.

[0073] When determining the relative time relationship, the relative time relationship can be determined through clinical data. First, the case information of adverse drug reactions in children can be collected and processed. Then, the adverse reaction time of the second adverse reaction corresponding to the second health characteristic and the second health characteristic can be determined based on the case information of adverse drug reactions in children. Subsequently, the adverse reaction time of the second adverse reaction corresponding to the second health characteristic and the second health characteristic can be calculated.

[0074] Exemplarily, the case information of adverse drug reaction of a child includes the time when the child takes the drug, and the time when the child takes the drug is used to calculate the length of the adverse reaction time of the second adverse reaction corresponding to the second health feature.

[0075] S132a2. Determine the average value of the time correspondence between the medication time and the adverse reaction time of the child as the relative time relationship of the second adverse reaction feature corresponding to the second health feature.

[0076] In order to improve the accuracy of calculating the adverse reaction time of the second health characteristic and the second adverse reaction corresponding to the second health characteristic, the time when the child takes the medicine can be used as a reference to determine the average value of the time correspondence between the child's medication time and the adverse reaction time, which is used as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

[0077] Exemplarily, the time correspondence between the medication time of the child and the adverse reaction time includes the time length between the end time of the medication time of the child and the start time of the adverse reaction time, and also includes the time length of the adverse reaction time.

[0078] The beneficial effect brought about by the above implementation method is that, by taking the medication time of the child as a reference, the relative time relationship of the second adverse reaction feature corresponding to the second health feature is determined and the average value is calculated, which can further improve the accuracy of predicting adverse reactions.

[0079] In some implementations, in the above S132a, determining the relative time relationship of the second adverse reaction feature corresponding to the second health feature also includes S132a3 to S132a4, and S132a3 to S132a4 are specifically described below.

[0080] S132a3. Determine the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic based on the adverse drug reaction case information of the child, and determine the health status adjustment factor based on the second health characteristic.

[0081] The health characteristics of different children may affect the timing and duration of adverse reactions. When calculating the adverse reaction time, in order to further improve the accuracy of determining the adverse reaction time, the health status adjustment factor can be determined based on the second health characteristic. Then, the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic can be determined based on the child's adverse reaction case information. Then, the adverse reaction time of the second adverse reaction corresponding to the second health characteristic can be adjusted using the health status adjustment factor.

[0082] For example, the health status adjustment factor corresponding to the second health characteristic may be calculated based on the weight and age of the child.

[0083] S132a4. When the adverse reaction time is after the time period of the second health characteristic, the product of the duration between the adverse reaction time and the time period of the second health characteristic and the adjustment factor, and the product of the duration of the adverse reaction time and the adjustment factor are used as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

[0084] When adjusting and calculating the adverse reaction time through the health status adjustment factor, when the adverse reaction time is after the time period of the second health characteristic, the product of the duration between the adverse reaction time and the time period of the second health characteristic and the adjustment factor, and the product of the duration of the adverse reaction time and the adjustment factor can be used as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic. The relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic includes the start time of the adverse reaction time and the duration of the adverse reaction time.

[0085] Exemplarily, the product of the duration between the start time of the adverse reaction time and the start time of the time period where the second health characteristic is located and the adjustment factor can be used as the target time interval between the adverse reaction time and the time period where the second health characteristic is located, and the product of the duration of the adverse reaction time and the adjustment factor can be used as the target time length of the adverse reaction time. The relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic includes the target time interval and target time length of the time relative relationship.

[0086] The beneficial effect brought about by the above-mentioned implementation method is that it can determine the health status adjustment factor based on the second health characteristic of the child, and determine the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic based on the health status adjustment factor, so as to take into account the individual differences in the prediction of adverse reactions in the children.

[0087] In some implementations, in the above S132a, determining the relative time relationship between the second health feature and the second adverse reaction feature further includes: when the adverse reaction time is less than or equal to a preset time period, excluding the adverse reaction time and the second health feature.

[0088] When the adverse reaction time is less than or equal to the preset time period, it means that the measurement of the adverse reaction may be inaccurate or the symptoms are not obvious, and the adverse reaction time and the second health characteristic can be excluded.

[0089] The beneficial effect brought about by the above implementation method is that it can eliminate the situation where the measurement of adverse reactions may be inaccurate or the symptoms are not obvious, thereby improving the accuracy of adverse reaction prediction.

[0090] An embodiment of the present application also provides a pediatric drug adverse reaction prediction system based on deep learning, including a unit for executing the method described above.

[0091] Figure 3 A schematic diagram of the logical structure of a pediatric drug adverse reaction prediction system based on deep learning provided in one embodiment of the present application is shown in FIG. Figure 3 As shown, the system 1 of this embodiment includes a processing unit 11, a storage unit 12 and a transceiver unit 13. The processing unit 11 is used to process data, the storage unit 12 is used to store data, and the transceiver unit 13 is used to send and receive data. The processing unit 11, the storage unit 12 and the transceiver unit 13 cooperate with each other to implement the above method. The beneficial effects brought by the embodiment of the present application have been described in the above method and will not be repeated here.

[0092] An embodiment of the present application also provides a pediatric drug adverse reaction prediction system based on deep learning, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method described above when executing the computer program.

[0093] Figure 4 A schematic diagram of the physical structure of a pediatric drug adverse reaction prediction system based on deep learning provided in one embodiment of the present application, such as Figure 4 As shown, the system 2 of this embodiment includes: at least one processor 20 ( Figure 4 Only one processor 20 is shown in the figure), a memory 21, and a computer program 22 stored in the memory 21 and executable on the at least one processor 20. When the processor 20 executes the computer program 22, the steps in any of the above-mentioned method embodiments are implemented. The beneficial effects brought about by the embodiments of the present application have been described in the above-mentioned methods and will not be repeated here.

[0094] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0095] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0096] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0097] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0098] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0099] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0100] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0101] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0102] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0103] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for predicting adverse drug reactions in pediatrics based on deep learning, characterized in that: The method comprises: Obtain the health status characteristics of the target children and the characteristics of the drugs to be used; Through the adverse reaction prediction model, according to the health status characteristics of the target child and the characteristics of the drug to be used, the first health status characteristics and the first adverse reaction characteristics of the target child in time sequence after the target child uses the drug are determined; The adverse reaction prediction model is obtained by training a neural network model with the collected health status characteristics of the children, the characteristics of the drugs to be used, and the continuous health status characteristics and adverse reaction characteristics of the children within a period of time after taking the drugs. Determine the second health feature from the first health status feature in the time series, and determine the second adverse reaction feature associated with the second health feature from the first adverse reaction feature in the time series, and determine the adverse reaction exclusion feature according to the feature of the drug to be used, the second health feature and the second adverse reaction feature through the adverse reaction exclusion model; When determining the second health feature from the first health status feature, the screening condition for screening the second health feature is the condition under which the physical health of the patient changes. By determining the second health feature from the first health status feature, the adverse reaction prediction result of the patient can be further identified through the second health feature; When training the adverse reaction exclusion model, a large amount of drug usage characteristics, health characteristics and adverse reaction characteristics in the children's medication case data are collected, and the neural network model is trained based on the drug usage characteristics, health characteristics, adverse reaction characteristics and adverse reaction exclusion characteristics in the children's medication case data; When the adverse reaction exclusion feature meets the preset conditions, the prediction result that the target patient will not have adverse reactions to the drug is output; when the adverse reaction exclusion feature does not meet the preset conditions, the prediction result that the target patient will have adverse reactions to the drug is output; Determining a second health feature from a first health status feature in a time series, and determining a second adverse reaction feature associated with the second health feature from a first adverse reaction feature in a time series, including: Determine a second health feature that satisfies a first preset condition from the first health status feature in the time series; The first precondition is the condition under which physical health changes; Determine a first time period corresponding to the second health feature, and obtain a second adverse reaction feature within the first time period; wherein the second adverse reaction feature is a sub-feature of the first adverse reaction feature in time sequence; When determining the second adverse reaction feature associated with the second health feature from the first adverse reaction feature in time series, the second adverse reaction feature is obtained according to the time correspondence, the first time period corresponding to the second health feature is specifically determined, and the second adverse reaction feature in the first time period is obtained, and the second adverse reaction feature is a sub-feature of the first adverse reaction feature in time series; When the second adverse reaction feature meets the second preset condition, output the second health feature and the second adverse reaction feature; when the second adverse reaction feature does not meet the second preset condition, exclude the second health feature and the second adverse reaction feature; The second preset condition is that the correlation between the second adverse reaction characteristic and the second health characteristic reaches a preset threshold, or the adverse reaction symptoms corresponding to the second adverse reaction characteristic meet the symptom requirements of mild or above adverse reactions.

2. The method according to claim 1, characterized in that Determining a first time period corresponding to the second health characteristic includes: Determining a relative time relationship between a second adverse reaction feature and a second health feature; According to the relative time relationship, a first time period partially or completely overlaps with the time period of the second health characteristic is determined, or a first time period after the time period of the second health characteristic is determined, and the duration of the first time period is determined.

3. The method according to claim 2, characterized in that Determining the relative time relationship between the second health characteristic and the second adverse reaction characteristic, including: Collecting the case information of adverse drug reaction of the patient, and determining the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic according to the case information of adverse drug reaction of the patient; wherein the case information of adverse drug reaction of the patient includes the time of drug use of the patient; The average value of the time correspondence between the medication time and the adverse reaction time of the child is determined as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

4. The method according to claim 3, characterized in that Determining the relative time relationship between the second adverse reaction feature and the second health feature also includes: Determine the second health characteristic and the adverse reaction time of the second adverse reaction corresponding to the second health characteristic according to the adverse reaction case information of the child, and determine the health status adjustment factor according to the second health characteristic; When the adverse reaction time is after the time period of the second health characteristic, the product of the duration between the adverse reaction time and the time period of the second health characteristic and the adjustment factor, and the product of the duration of the adverse reaction time and the adjustment factor are taken as the relative time relationship of the second adverse reaction characteristic corresponding to the second health characteristic.

5. The method according to claim 4, characterized in that Determining the relative time relationship between the second adverse reaction feature and the second health feature also includes: When the adverse reaction time is less than or equal to the preset time period, the adverse reaction time and the second health feature are excluded.

6. A pediatric drug adverse reaction prediction system based on deep learning, characterized in that: The method comprises means for performing the method according to any one of claims 1 to 5.

7. A pediatric drug adverse reaction prediction system based on deep learning, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images