A symptom recommendation method, device and computer-readable storage medium

Abstract

The present application provides a symptom recommendation method, device and computer-readable storage medium, including: obtaining user information including user symptoms input by a user, and determining candidate symptoms according to the user symptoms, determining the correlation coefficient between the candidate symptoms and each preset symptom, determining the target symptom from the preset symptoms according to the correlation coefficient, and recommending symptoms according to the target symptom. Since symptom recommendations can be made to users automatically, the reliance on manual experience during self-diagnosis or consultation can be reduced.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a symptom recommendation method, device, and computer-readable storage medium. Background Art

[0002] At present, whether it is the process of patients using applications for self-diagnosis or the process of doctors consulting patients based on applications, they are highly dependent on user experience. Therefore, there is an urgent need for a solution that can recommend symptoms to users to reduce the dependence on user experience during self-diagnosis or consultation. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a symptom recommendation method, device, equipment and storage medium to solve the above-mentioned technical problems.

[0004] The present application embodiment provides a symptom recommendation method, the method comprising:

[0005] Acquire user information; the user information includes user symptoms input by the user;

[0006] determining candidate symptoms according to the user symptoms;

[0007] Determine the correlation coefficient between the candidate symptoms and each preset symptom; the correlation coefficient is a coefficient calculated based on each eigenvector matrix after performing singular value decomposition on the correlation degree matrix to obtain the eigenvector matrix corresponding to each preset symptom; each element in the correlation degree matrix represents the correlation degree between the corresponding preset symptom and the corresponding preset prescription;

[0008] Determining a target symptom from the preset symptoms according to the correlation coefficient;

[0009] Symptom recommendations are made based on the target symptoms.

[0010] In the above implementation process, since symptom recommendations can be made to users based on target symptoms, the reliance on user experience during self-diagnosis or consultation can be reduced.

[0011] Furthermore, the target symptoms include physical signs and symptoms.

[0012] In the above implementation process, it is applicable to the scenario of recommending physical signs and symptoms to users, which improves the universality of this solution.

[0013] Further, the candidate symptoms include a first candidate symptom; and determining the candidate symptoms according to the user symptoms includes:

[0014] Determine a main complaint symptom corresponding to the user's symptom, and use the main complaint symptom as the first candidate symptom;

[0015] or,

[0016] The symptom input last among the user symptoms is used as the first candidate symptom.

[0017] In the above implementation process, the chief symptom is taken as the first candidate symptom, or the last symptom entered among the user's symptoms is taken as the first candidate symptom, ensuring that the candidate symptom is the patient's main or important symptom. Therefore, the target symptom determined based on the candidate symptom is a symptom with a high degree of correlation with the lesion, thereby improving the reliability and accuracy of the recommendation results.

[0018] Furthermore, the candidate symptoms further include a second candidate symptom; and determining the candidate symptoms according to the user symptoms includes:

[0019] A second candidate symptom associated with the first candidate symptom is determined.

[0020] In the above implementation process, the target symptom can be determined based on the first candidate symptom and the second candidate symptom associated with the first candidate symptom, so that the recommended target symptom is related to the lesion and conforms to the actual physical condition of the patient, further improving the accuracy of the recommendation result.

[0021] Furthermore, determining the correlation coefficient between the candidate symptom and each preset symptom includes:

[0022] Determine a first correlation coefficient between the first candidate symptom and each of the preset symptoms, and determine a second correlation coefficient between the second candidate symptom and each of the preset symptoms

[0023] Determining the physical signs and symptoms that need to be recommended from the preset symptoms according to the correlation coefficient includes:

[0024] For each of the preset symptoms, determining a third correlation coefficient between the preset symptom and a candidate symptom group according to the first correlation coefficient and the second correlation coefficient corresponding to the preset symptom; the candidate symptom group is a symptom group consisting of the first candidate symptom and the second candidate symptom;

[0025] The physical signs and symptoms that need to be recommended are screened out from the preset symptoms according to the third correlation coefficient.

[0026] In the above implementation process, the first candidate symptom and the second candidate symptom are taken as a candidate symptom group, and the target symptom with a high correlation with the user symptom is determined from the preset symptoms according to the third correlation coefficient between each preset symptom and the candidate symptom group, thereby improving the accuracy of the recommendation results.

[0027] Further, the determining a second candidate symptom associated with the first candidate symptom includes:

[0028] determining a fourth correlation coefficient between the first candidate symptom and each symptom of the user symptoms except the first candidate symptom;

[0029] Determine the symptom corresponding to the fourth correlation coefficient that is greater than or equal to the preset correlation coefficient threshold as a candidate symptom to be determined;

[0030] From the pending candidate symptoms, the pending candidate symptom corresponding to the maximum value of the fourth correlation coefficient is screened out as the second candidate symptom.

[0031] In the above implementation process, among the user symptoms input by the user, the symptom with the highest correlation with the first candidate symptom is used as the second candidate symptom to ensure that the second candidate symptom is the user's main discomfort symptom, thereby improving the reliability of the final recommendation result.

[0032] Furthermore, determining the correlation coefficient between the candidate symptom and each preset symptom includes:

[0033] The correlation coefficient between the candidate symptom and each preset symptom is queried from a preset correlation coefficient database; the correlation coefficient between each preset symptom is stored in the correlation coefficient database.

[0034] In the above implementation process, there is no need to calculate the correlation coefficient, but the correlation coefficient can be directly queried from a preset correlation coefficient database, thereby improving the operation efficiency.

[0035] Furthermore, each element in the correlation degree matrix is ​​the correlation degree of the preset symptoms with respect to the preset prescription calculated based on a preset medical case database; the medical case database contains a plurality of medical case documents, each of which contains information on a plurality of the preset symptoms and information on the preset prescriptions that can treat the plurality of the preset symptoms; the correlation degree is calculated by the formula Calculated;

[0036] Among them, R si,pj represents the value of the element corresponding to the position of the i-th row and the j-th column in the association degree matrix, si represents the i-th preset symptom, pj represents the j-th preset prescription, n ij represents the number of medical records in the medical records database that contain both the j-th preset prescription and the i-th preset symptom, ∑ k n kj represents the total number of occurrences or symptom categories of all preset symptoms in the medical record database that appear in the same medical record document as the jth preset prescription, ∑ k n kirepresents the total number of occurrences or the number of prescription categories of all preset prescriptions that appear in the same medical record document as the i-th preset symptom in the medical record database, D represents the total number of occurrences or the number of prescription categories of all preset prescriptions in the medical record database; e and Q are preset constants.

[0037] In the above implementation process, a new method for calculating the correlation coefficient between the preset symptoms is provided, and the characteristic vectors corresponding to the preset symptoms are determined based on the degree of correlation between the prescription and the symptoms, thereby improving the accuracy of the calculation results.

[0038] An embodiment of the present application further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement any one of the above methods.

[0039] 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 at least one processor, any one of the above methods is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A flowchart of the symptom recommendation method provided in Example 1 of the present application;

[0042] Figure 2 A schematic diagram of providing symptom recommendations to users provided in Example 1 of the present application;

[0043] Figure 3 A schematic diagram of a process for determining a correlation coefficient provided in Example 1 of the present application;

[0044] Figure 4 A schematic diagram of the structure of a symptom recommendation device provided in Example 2 of the present application;

[0045] Figure 5 This is a schematic diagram of the structure of an electronic device provided in Example 3 of the present application. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0047] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present invention are only for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0048] In the description of the present invention, it should be understood that the numerical labels before the steps do not identify the order of executing the steps, but are only used to facilitate the description of the present invention and distinguish each step, and therefore cannot be understood as a limitation of the present invention.

[0049] Several embodiments will be provided below to specifically introduce the symptom recommendation method, device and computer-readable storage medium.

[0050] Embodiment 1:

[0051] See also Figure 1 As shown, the embodiment of the present application provides a symptom recommendation method, which can be applied to electronic devices, including but not limited to PCs (Personal Computers), mobile phones, tablet computers, laptop computers, etc. The symptom recommendation method provided in the embodiment of the present application may include the following steps:

[0052] S11: Obtain user information; the user information includes user symptoms input by the user.

[0053] S12: Determine candidate symptoms according to the user's symptoms.

[0054] S13: Determine the correlation coefficient between the candidate symptoms and each preset symptom; the correlation coefficient is a coefficient calculated based on each eigenvector matrix after performing singular value decomposition on the correlation degree matrix to obtain the eigenvector matrix corresponding to each preset symptom; each element in the correlation degree matrix represents the degree of correlation between the corresponding preset symptom and the corresponding preset prescription.

[0055] S14: Determine a target symptom from the preset symptoms according to the correlation coefficient.

[0056] S15: Make symptom recommendations based on target symptoms.

[0057] The above steps are described in detail below.

[0058] It should be noted that, in actual application, for step S15, the target symptom may be recommended to a target user, where the target user may be the user in step S11 or another user.

[0059] Exemplarily, the user in the above step S11 may be a patient user, for example, the patient himself or a relevant person who is familiar with the patient's condition, such as a family member of the patient. At this time, in step S15, the target symptoms may be recommended to the patient user to prompt the patient user to perform a self-examination of the symptoms, so that more comprehensive condition information may be collected on the patient side; of course, the target symptoms may also be recommended to the doctor user, so that the doctor may diagnose the patient user based on the recommended target symptoms.

[0060] In addition to user symptoms, the user information in the embodiments of the present application may also include information such as user age and gender. In this way, when the method is applied to a scenario of assisting doctors in consultation, it can assist doctors in determining the cause of the disease and completing diagnosis and treatment.

[0061] Users can enter their symptoms on the application's intelligent consultation platform based on their actual physical condition, for example, entering symptoms such as "cough" and "headache".

[0062] The target symptoms in the embodiments of the present application may include physical symptoms, which refer to symptoms that can be objectively detected by others. They are objectively existing symptoms and do not change with the patient's subjective consciousness. The other person here can be anyone other than the patient, such as a doctor or a patient's companion. The inspection here refers to the diagnosis method of others through non-direct inquiry methods such as naked eyes or touch.

[0063] The basic physical function of the patient, such as diet, sleep, defecation, menstruation, and physical and mental state, such as spirit, body shape, facial image, tongue image, pulse, etc., these symptoms do not change with the patient's subjective consciousness, and can be directly observed or checked by other people other than the patient. Therefore, the physical signs and symptoms in the embodiments of the present application can include at least one of the symptoms related to body shape, facial image, tongue image, pulse, diet, sleep, spirit, menstruation, and defecation. These physical signs and symptoms can be used to judge the cause of the disease, the severity of the disease, the mechanism of the disease, the location of the disease, and the strength of the body's quality, but these physical signs and symptoms are easily ignored by patients and usually require doctor's examination.

[0064] Recommending physical signs and symptoms to patient users can prompt patient users to self-check physical signs and symptoms based on the recommended results. More comprehensive symptom information can be collected without the need for doctors to check such symptoms, which can improve the efficiency of diagnosis and treatment.

[0065] The candidate symptoms in the embodiment of the present application may be one or more.

[0066] In the first example of this embodiment, for step S12, the main complaint symptom corresponding to the user's symptom can be determined and used as the first candidate symptom. The main complaint symptom refers to the main symptom that causes the patient the most pain or causes the patient to visit the doctor this time.

[0067] For example, when a patient user inputs user symptoms according to actual physical conditions, a chief complaint label can be added to the main symptoms in the input user symptoms. The electronic device can determine which symptom in the input user symptoms is the chief complaint symptom by identifying the chief complaint label, and then use the chief complaint symptom as the first candidate symptom of the current patient user.

[0068] In the second example of this embodiment, considering that the patient user does not understand the pathological process, and the chief complaint symptoms are more important in the diagnosis and treatment of the disease, the user information input by the patient user including user symptoms, gender, and age can be pushed to the doctor user, and the doctor user determines the chief complaint symptoms of the patient user from the user symptoms based on the user information. That is, the electronic device receives the chief complaint symptoms determined by the doctor user based on the user symptoms, and then uses the chief complaint symptoms as the first candidate symptom of the current patient user.

[0069] In the third example of this embodiment, the user information input by the patient user can be input into the trained main complaint symptom recommendation model to obtain the main complaint symptom corresponding to the user symptom. The main complaint symptom recommendation model is a model trained based on training sample data, and each training sample data contains information on multiple user symptoms and corresponding main complaint symptom information.

[0070] It should be noted that when there are multiple main complaint symptoms determined, one can be randomly selected from the main complaint symptoms as the first candidate symptom. It is also possible to set a corresponding weight value for each preset symptom in the preset symptom database in advance, and to set the corresponding weight value of each preset symptom according to the frequency of occurrence of each symptom in the medical record database. The higher the frequency of occurrence of the symptom, the higher the corresponding weight value. When multiple main complaint symptoms are determined, the main complaint symptom with the largest corresponding weight value is selected as the first candidate symptom.

[0071] For step S12, the last symptom inputted among the user symptoms may also be used as the first candidate symptom. It should be noted that the method of selecting the first candidate symptom from the user symptoms inputted by the patient user may be flexibly set by the developer. For example, the main complaint symptom may be preferentially used as the first candidate symptom, and when the main complaint symptom cannot be determined, the last symptom inputted among the user symptoms may be used as the first candidate symptom.

[0072] Exemplarily, a second candidate symptom associated with the first candidate symptom may be determined based on the first candidate symptom. The following is an introduction to the method of determining the second candidate symptom.

[0073] In some embodiments, when the determined first candidate symptom is a symptom among the user symptoms input by the user, the fourth correlation coefficient between the first candidate symptom and each symptom among the user symptoms except the first candidate symptom can be determined, and the symptom corresponding to the fourth correlation coefficient greater than or equal to the preset correlation coefficient threshold is determined as the pending candidate symptom, and then the pending candidate symptom corresponding to the maximum value of the fourth correlation coefficient is screened out from the pending candidate symptoms as the second candidate symptom. Of course, if the first candidate symptom is not a symptom among the user symptoms, the fourth correlation coefficient between the first candidate symptom and each user symptom input by the user is determined, and the second candidate symptom can be determined by referring to the above method, which will not be repeated here.

[0074] In some other implementations, the correlation coefficient between the first candidate symptom and each preset symptom in the preset symptom database may be determined, and the preset symptom corresponding to the maximum correlation coefficient may be used as the second candidate symptom.

[0075] It should be noted that a symptom database can be preset, in which multiple preset symptoms are stored. The preset symptoms can be composed of preset physical symptoms and preset non-physical symptoms. In this case, the candidate symptoms and recommended symptoms mentioned in the embodiments of the present application all belong to the symptoms in the database.

[0076] It should be noted that when the preset symptom database contains preset physical symptoms and preset non-physical symptoms, the preset physical symptoms and preset non-physical symptoms can be distinguished by setting labels for the symptoms in advance. For example, an objective attribute label can be set for the preset physical symptoms to characterize them as preset physical symptoms, and symptoms that are not set with objective attribute labels or are set with subjective attribute labels are preset non-physical symptoms. Alternatively, subjective attribute labels can be set for preset non-physical symptoms, and symptoms that are not set with subjective attribute labels are preset physical symptoms.

[0077] It can be understood that when the determined candidate symptoms include a first candidate symptom and a second candidate symptom, a first correlation coefficient between the first candidate symptom and each preset symptom can be determined, and a second correlation coefficient between the second candidate symptom and each preset symptom can be determined.

[0078] In an optional implementation, for each preset symptom in the symptom database, the third correlation coefficient between it and the candidate symptom group can be determined according to its corresponding first correlation coefficient and second correlation coefficient, and the candidate symptom group is a symptom group consisting of the first candidate symptom and the second candidate symptom; the target symptom to be recommended is screened out from the preset symptoms according to the third correlation coefficient. By associating the preset symptoms with the candidate symptom group, the comprehensiveness of the recommendation results can be improved. Exemplarily, the product of the first correlation coefficient and the corresponding preset weight and the product of the second correlation coefficient and the corresponding preset weight can be added to obtain the corresponding third correlation coefficient.

[0079] In another optional implementation, the first target symptom to be recommended can be directly screened out from the preset symptoms according to the first correlation coefficient, and the second target symptom to be recommended can be directly screened out from the preset symptoms according to the second correlation coefficient. The first target symptom and the second target symptom are then recommended to the patient user.

[0080] Of course, if the candidate symptoms only include the first candidate symptom, the target symptom to be recommended can be directly screened out from the preset symptoms based on the first correlation coefficient between the first candidate symptom and each preset symptom.

[0081] It is understandable that a preset number of symptoms can be screened out from the preset symptoms according to the order of correlation coefficients from high to low. If the screened symptoms include non-sign symptoms, the non-sign symptoms are filtered out and the obtained sign symptoms are recommended to the patient user as target symptoms.

[0082] In some embodiments, after the target symptom is recommended to the user, the following steps may also be included:

[0083] Gets the user symptom re-entered by the user for the recommended target symptom.

[0084] The target symptom and the re-entered user symptom are sent to another user or the user who entered the user symptom.

[0085] In an actual application scenario, after the target symptoms are recommended to the patient user, the user symptoms re-entered by the patient user for the recommended target symptoms can be obtained, and then all the collected user symptoms entered by the patient user are sent to the doctor user for the doctor user to diagnose the patient user's disease.

[0086] In some embodiments, after obtaining the target symptoms that need to be recommended to the user, the symptom category to which each target symptom belongs can be determined. Specifically, the target symptoms that need to be recommended to the user can be classified according to a clustering algorithm, and the label names of each category can be determined according to the classification results. Figure 2 As shown, the target symptoms can be placed in the column of the corresponding label name and displayed to the user.

[0087] In a first optional implementation, a correlation coefficient database may be pre-stored in the electronic device. The correlation coefficients between preset symptoms are stored in the correlation database. When the correlation coefficient between two symptoms needs to be determined, the correlation coefficient may be queried from the correlation coefficient database. The correlation coefficient may be directly queried and retrieved from the preset correlation coefficient database, thereby improving operating efficiency.

[0088] In a second optional implementation, the correlation coefficients between symptoms can be calculated in real time during the actual application process.

[0089] Regardless of which of the above implementations is used, the correlation coefficient between two symptoms can be calculated in the following manner. For specific steps, see Figure 3 As shown, including:

[0090] S31: Obtain a correlation degree matrix.

[0091] Each element in the correlation degree matrix represents the correlation degree between the corresponding preset prescription and the corresponding preset symptom. For example, each row of the correlation degree matrix corresponds to a preset symptom, each column of the correlation degree matrix corresponds to a preset prescription, and each element in the matrix can represent the correlation degree between the corresponding preset prescription and the corresponding preset symptom.

[0092] Exemplarily, each element in the correlation degree matrix is ​​the correlation degree of the preset symptoms with respect to the preset prescription calculated based on the preset medical case database; the medical case database contains a plurality of medical case documents, each of which contains information on a plurality of preset symptoms and information on the preset prescriptions that can treat the plurality of preset symptoms; the correlation degree can be calculated by the formula Calculated.

[0093] The meanings of the parameters in the above formula are introduced below.

[0094] R si,pj Represents the value of the element corresponding to the i-th row and j-th column in the association degree matrix.

[0095] si represents the i-th preset symptom, and pj represents the j-th preset prescription.

[0096] n ij It represents the number of medical record documents in the medical record database that contain both the j-th preset prescription and the i-th preset symptom.

[0097] ∑ k n kj It represents the total number of occurrences or the number of symptom categories of all preset symptoms that appear in the same medical case document as the jth preset prescription in the medical case database. For example, if the preset prescription X only appears in two medical case documents, and these two medical case documents contain a preset symptoms and b preset symptoms respectively, and there are c overlapping preset symptoms between these two medical case documents, then the total number of occurrences of all preset symptoms that appear in the same medical case document as the preset prescription X is a+b, and the number of symptom categories of all preset symptoms that appear in the same medical case document as the preset prescription X is a+bc.

[0098] ∑ k n ki It represents the total number of occurrences or the number of prescription categories of all preset prescriptions that appear in the same medical record document as the i-th preset symptom in the medical record database.

[0099] D represents the total number of occurrences of all preset prescriptions in the medical record database or the number of prescription categories. The number of prescription categories is essentially the number of preset prescriptions.

[0100] e and Q are preset constants, e is a preset smoothing coefficient, and Q is a preset adjustment coefficient.

[0101] For example, e can be 2 and Q can be 1.

[0102] After obtaining the correlation degree between each preset prescription and each preset symptom, a correlation degree matrix can be obtained.

[0103] S32: Perform singular value decomposition on the correlation degree matrix to obtain the eigenvector matrix corresponding to each preset symptom.

[0104] In step S32, Perform singular value decomposition on the correlation matrix.

[0105] Among them, R m*n represents the correlation degree matrix, d represents the vector dimension of the decomposed prescription and the preset symptoms, and the specific size of d can be set by the developer based on experience, for example, it can be set to 100. m*d Represents the d-dimensional prescription feature vector matrix of m prescriptions, V n*d A d-dimensional feature vector matrix representing n preset symptoms. Matrix V n*d Each row vector of represents the eigenvector matrix corresponding to the preset symptom.

[0106] S33: Calculate the correlation coefficient between each preset symptom according to each eigenvector matrix.

[0107] In step S33, the correlation coefficient between the preset symptoms can be calculated according to the cosine distance between the eigenvector matrices. For example, the correlation coefficient between the preset symptoms can be calculated according to the following formula:

[0108]

[0109] Among them, Dis ij represents the correlation coefficient between the i-th preset symptom and the j-th preset symptom, V ik represents the value of the k-th dimension of the eigenvector matrix corresponding to the i-th preset symptom, V jk Represents the value of the k-th dimension of the eigenvector matrix corresponding to the j-th preset symptom.

[0110] In some embodiments, after the correlation coefficient between the preset symptoms is obtained in step S33, the correlation coefficient can be stored in a correlation coefficient database.

[0111] Embodiment 2:

[0112] This application embodiment provides a symptom recommendation device, see Figure 4 As shown, including:

[0113] The acquisition module 501 is used to acquire user information; the user information includes user symptoms input by the user.

[0114] The first determination module 502 is used to determine candidate symptoms according to user symptoms.

[0115] The second determination module 503 is used to determine the correlation coefficient between the candidate symptoms and each preset symptom; the correlation coefficient is a coefficient calculated based on each eigenvector matrix after performing singular value decomposition on the correlation degree matrix to obtain the eigenvector matrix corresponding to each preset symptom; each element in the correlation degree matrix represents the degree of correlation between the corresponding preset symptom and the corresponding preset prescription.

[0116] The third determination module 504 determines the target symptom from the preset symptoms according to the correlation coefficient.

[0117] The recommendation module 505 is used to make symptom recommendations based on the target symptoms.

[0118] In an exemplary embodiment, the target symptom includes physical symptoms, which include at least one of symptoms related to body shape, facial appearance, tongue appearance, pulse appearance, diet, sleep, spirit, menstruation, and defecation.

[0119] In an exemplary embodiment, the first determination module 502 is used to determine the main complaint symptom corresponding to the user's symptoms, and use the main complaint symptom as the first candidate symptom; or, use the last symptom input among the user's symptoms as the first candidate symptom.

[0120] In an exemplary embodiment, the first determination module 502 is configured to determine a second candidate symptom associated with the first candidate symptom.

[0121] In an exemplary embodiment, the first determination module 502 is used to determine the fourth correlation coefficient between the first candidate symptom and each of the user's symptoms except the first candidate symptom, and determine the symptom corresponding to the fourth correlation coefficient that is greater than or equal to a preset correlation coefficient threshold as the pending candidate symptom, and from the pending candidate symptoms, select the pending candidate symptom corresponding to the maximum value of the fourth correlation coefficient as the second candidate symptom.

[0122] In an exemplary embodiment, the second determination module 503 is used to determine the first correlation coefficient between the first candidate symptom and each preset symptom, and to determine the second correlation coefficient between the second candidate symptom and each preset symptom, and for each preset symptom, determine the third correlation coefficient between it and the candidate symptom group according to its corresponding first correlation coefficient and second correlation coefficient; the candidate symptom group is a symptom group composed of the first candidate symptom and the second candidate symptom, and the target symptom to be recommended is screened out from the preset symptoms according to the third correlation coefficient.

[0123] In an exemplary embodiment, the second determination module 503 is used to query the correlation coefficient between the candidate symptom and each preset symptom from a preset correlation coefficient database; the correlation coefficient between each preset symptom is stored in the correlation coefficient database.

[0124] In an exemplary embodiment, the symptom recommendation device may also include a calculation module for obtaining a correlation degree matrix, performing singular value decomposition on the correlation degree matrix, obtaining eigenvector matrices corresponding to each preset symptom, calculating the correlation coefficient between each preset symptom based on each eigenvector matrix, and storing the correlation coefficient in a correlation coefficient database.

[0125] Each element in the correlation degree matrix is ​​the correlation degree of the preset symptoms with respect to the preset prescription calculated based on the preset medical case database; the medical case database contains multiple medical case documents, each of which contains information on multiple preset symptoms and information on preset prescriptions that can treat the multiple preset symptoms; the correlation degree can be calculated by the formula Calculated;

[0126] Among them, R si,pj represents the value of the element corresponding to the position of the i-th row and the j-th column in the association degree matrix, si represents the i-th preset symptom, pj represents the j-th preset prescription, nij represents the number of medical records in the medical records database that contain both the j-th preset prescription and the i-th preset symptom, ∑ k n kj represents the total number of occurrences or symptom categories of all preset symptoms in the medical record database that appear in the same medical record document as the jth preset prescription, ∑ k n ki represents the total number of occurrences or the number of prescription categories of all preset prescriptions that appear in the same medical record document as the i-th preset symptom in the medical record database, D represents the total number of occurrences or the number of prescription categories of all preset prescriptions in the medical record database; e and Q are preset constants.

[0127] In an exemplary embodiment, the acquisition module 501 is also used to obtain the user symptoms re-entered by the user for the recommended target symptoms after the recommendation module recommends the target symptoms to the user. The target symptom recommendation device may also include a sending module for sending all user symptoms entered by the user to another user.

[0128] It should be understood that, for the sake of brevity, some of the contents described in the first embodiment will not be repeated in this embodiment.

[0129] Embodiment three:

[0130] Based on the same inventive concept, the present application embodiment provides an electronic device, see Figure 5 As shown, it includes a processor 501 and a memory 502, wherein the memory 502 stores a computer program, and the processor 501 executes the computer program to implement the steps of the method in the above-mentioned embodiment 1, which will not be repeated here.

[0131] Understandably, Figure 5 The structure shown is for illustration only, and the device may also include Figure 5 More or fewer components as shown, or with Figure 5 Different configurations are shown.

[0132] The processor 501 may be an integrated circuit chip with signal processing capabilities. The processor 501 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It may implement or execute various methods, steps, and logic block diagrams disclosed in the embodiments of the present application.

[0133] The memory 502 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable read-only memory (EPROM), electrically erasable read-only memory (EEPROM), etc.

[0134] This embodiment also provides a computer-readable storage medium, such as a floppy disk, a CD, a hard disk, a flash memory, a USB flash drive, a secure digital (SD) card, a multimedia (MMC) card, etc., in which one or more programs for implementing the above steps are stored. These one or more programs can be executed by one or more processors to implement the steps of the methods in the above embodiments, which will not be repeated here.

[0135] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0136] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method.

[0137] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A symptom recommendation method, characterized in that: include: Acquire user information; the user information includes user symptoms input by the user; determining candidate symptoms according to the user symptoms; Determine the correlation coefficient between the candidate symptoms and each preset symptom; the correlation coefficient is a coefficient calculated based on each eigenvector matrix after performing singular value decomposition on the correlation degree matrix to obtain the eigenvector matrix corresponding to each preset symptom; each element in the correlation degree matrix represents the correlation degree between the corresponding preset symptom and the corresponding preset prescription; Determining a target symptom from the preset symptoms according to the correlation coefficient; Make symptom recommendations based on the target symptoms described; Wherein, each element in the correlation degree matrix is ​​the correlation degree of the preset symptoms with respect to the preset prescription calculated based on the preset medical case database; the medical case database contains a plurality of medical case documents, each of which contains information of a plurality of the preset symptoms and information of the preset prescriptions that can treat the plurality of the preset symptoms; the correlation degree is calculated by the formula Calculated; Among them, R si,pj represents the value of the element corresponding to the position of the i-th row and the j-th column in the association degree matrix, si represents the i-th preset symptom, pj represents the j-th preset prescription, n ij represents the number of medical records in the medical records database that contain both the j-th preset prescription and the i-th preset symptom, ∑ k n kj represents the total number of occurrences or symptom categories of all preset symptoms in the medical record database that appear in the same medical record document as the jth preset prescription, ∑ k n ki represents the total number of occurrences or the number of prescription categories of all preset prescriptions that appear in the same medical record document as the i-th preset symptom in the medical record database, D represents the total number of occurrences or the number of prescription categories of all preset prescriptions in the medical record database; e and Q are preset constants.

2. The symptom recommendation method according to claim 1, characterized in that: The target symptoms include physical signs.

3. The symptom recommendation method according to claim 1, characterized in that: The candidate symptoms include a first candidate symptom; and determining the candidate symptoms according to the user symptoms includes: Determine a main complaint symptom corresponding to the user's symptom, and use the main complaint symptom as the first candidate symptom; or, The symptom input last among the user symptoms is used as the first candidate symptom.

4. The symptom recommendation method according to claim 3, characterized in that: The candidate symptoms also include a second candidate symptom; and determining the candidate symptoms according to the user symptoms includes: A second candidate symptom associated with the first candidate symptom is determined.

5. The symptom recommendation method according to claim 4, characterized in that: Determining the correlation coefficient between the candidate symptom and each preset symptom includes: Determining a first correlation coefficient between the first candidate symptom and each of the preset symptoms, and determining a second correlation coefficient between the second candidate symptom and each of the preset symptoms; Determining the target symptom to be recommended from the preset symptoms according to the correlation coefficient includes: For each of the preset symptoms, determining a third correlation coefficient between the preset symptom and a candidate symptom group according to the first correlation coefficient and the second correlation coefficient corresponding to the preset symptom; the candidate symptom group is a symptom group consisting of the first candidate symptom and the second candidate symptom; The target symptom to be recommended is screened out from the preset symptoms according to the third correlation coefficient.

6. The symptom recommendation method according to claim 4, characterized in that: The determining a second candidate symptom associated with the first candidate symptom comprises: determining a fourth correlation coefficient between the first candidate symptom and each symptom of the user symptoms except the first candidate symptom; Determine the symptom corresponding to the fourth correlation coefficient that is greater than or equal to the preset correlation coefficient threshold as a candidate symptom to be determined; From the pending candidate symptoms, the pending candidate symptom corresponding to the maximum value of the fourth correlation coefficient is screened out as the second candidate symptom.

7. The symptom recommendation method according to claim 1, characterized in that: Determining the correlation coefficient between the candidate symptom and each preset symptom includes: The correlation coefficient between the candidate symptom and each preset symptom is queried from a preset correlation coefficient database; the correlation coefficient between each preset symptom is stored in the correlation coefficient database.

8. An electronic device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program to implement the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by at least one processor, the method according to any one of claims 1 to 7 is implemented.

Images