Medical information recommendation method, device, medium and equipment

Abstract

The present disclosure relates to a method, device, medium and equipment for recommending medical information. The method includes: obtaining the self-description content of the current user; identifying the physiological characteristics of the current user from the self-description content of the current user; determining the probability of the physiological state associated with the physiological characteristics of the current user according to a predetermined graph structure, wherein the graph structure includes the association relationship and weight relationship between the physiological characteristics and the physiological state; and determining the recommended medical information according to the probability of the physiological state associated with the physiological characteristics of the current user. In this way, medical information can be recommended to the user efficiently and accurately based on the user's self-description, without wasting the user's medical time, reducing the pressure on hospital staff and reducing the probability of misdiagnosis.

Description

Technical Field

[0001] The present disclosure relates to the field of medical automatic control technology, and in particular to a method, apparatus, medium and device for recommending medical information. Background Art

[0002] When a user visits a doctor, confirming the doctor's information (for example, the department of treatment) is the first and indispensable step in the diagnosis and treatment process. If the doctor's information is incorrect, the subsequent diagnosis will undoubtedly be difficult, waste a lot of time, and even increase the probability of misdiagnosis.

[0003] Taking the medical department as an example, there are two main methods for users to determine the department they need to register for. One method is to search online. However, due to the complexity of online data and the risk of users mistakenly believing they have a specific illness, it is easy to select the wrong department. The other method is to go to the hospital front desk to inquire. However, due to the high volume of inquiries, it takes time to queue up, which not only wastes the user's time but also increases the pressure on hospital staff. Therefore, how to efficiently and accurately locate the department they need to see is an urgent need for both users and hospital staff. Summary of the Invention

[0004] The purpose of the present disclosure is to provide a medical information recommendation method, device, medium and equipment, which can efficiently and accurately recommend medical information to users based on the user's self-description.

[0005] In order to achieve the above objectives, the present disclosure provides a method for recommending medical information, the method comprising:

[0006] Get the current user's self-description content;

[0007] Identifying physiological characteristics of the current user from the self-description content of the current user;

[0008] determining a probability of a physiological state associated with a physiological characteristic of the current user according to a predetermined graph structure, wherein the graph structure includes an association relationship and a weight relationship between the physiological characteristic and the physiological state;

[0009] Recommended medical consultation information is determined according to the probability of the physiological state associated with the physiological characteristics of the current user.

[0010] Optionally, determining the probability of the physiological state associated with the physiological characteristics of the current user according to a predetermined graph structure includes:

[0011] Determining a physiological state associated with the physiological characteristics of the current user in the graph structure;

[0012] For each physiological state associated with the physiological characteristics of the current user, calculate the sum of weights contributed by the physiological characteristics associated with the physiological state in the physiological characteristics of the current user to the physiological state, as the weight of the physiological state;

[0013] The weights of the physiological states associated with the physiological characteristics of the current user are normalized to obtain the probability of each physiological state associated with the physiological characteristics of the current user.

[0014] Optionally, the atlas structure further includes an inferred relationship between vital sign parameters and physiological characteristics, and the method further includes:

[0015] Identifying the vital sign parameters of the current user from the self-description content of the current user;

[0016] If the physical sign parameter of the current user is greater than the corresponding threshold, the physiological feature inferred in the atlas structure due to the physical sign parameter being greater than the corresponding threshold is determined as the physiological feature of the current user.

[0017] Optionally, after determining the physiological state associated with the physiological characteristics of the current user in the graph structure, the method further includes:

[0018] Identify personal information of the current user from the self-description of the current user, the personal information including gender and age;

[0019] The physiological state associated with the physiological characteristics of the current user is screened according to the identified personal information.

[0020] Optionally, the medical information is a medical department, and the determining of the recommended medical information based on the probability of the physiological state associated with the physiological characteristics of the current user includes:

[0021] Determining the department corresponding to the physiological state associated with the physiological characteristics of the current user according to a predetermined correspondence relationship, wherein the predetermined correspondence relationship is a correspondence relationship between the physiological state and the department;

[0022] For each determined department, determine the sum of the probabilities of the physiological states corresponding to the department and associated with the physiological characteristics of the current user as a first probability of the department;

[0023] The recommended department for consultation is determined based on the first probability of each department.

[0024] Optionally, after determining the department corresponding to the physiological state associated with the physiological characteristics of the current user according to the predetermined corresponding relationship, the method further includes:

[0025] Identify personal information of the current user from the self-description of the current user, the personal information including gender and age;

[0026] The identified departments are screened based on the identified personal information.

[0027] Optionally, determining the recommended department for consultation based on the first probability of each department includes:

[0028] The department with the highest first probability is determined as the recommended department for consultation.

[0029] Optionally, determining the recommended department for consultation based on the first probability of each department includes:

[0030] Acquiring historical data, wherein the historical data includes multiple records, each record including self-description content and corresponding department;

[0031] Filtering multiple records whose self-description contents are similar to the self-description contents of the current user from the historical data;

[0032] The recommended department for consultation is determined based on the first probability of each department and the screened records.

[0033] Optionally, determining the recommended medical department according to the first probability of each department and the screened records includes:

[0034] The second probability of each department is calculated according to the following formula:

[0035]

[0036] For each department, calculate the target probability of the department according to the first probability and the second probability of the department;

[0037] The department with the highest target probability is determined as the recommended treatment department;

[0038] Among them, P 2i is the second probability of the i-th department in the screened records, S i is the maximum value of the similarity of the self-description content in one or more records corresponding to the i-th department in the filtered records, and n is the number of all departments in the filtered records.

[0039] Optionally, calculating the target probability of each department according to the first probability and the second probability of the department includes:

[0040] For each department, a weighted sum of the first probability and the second probability of the department is determined as the target probability of the department.

[0041] Optionally, the weight of the second probability is determined according to the following formula:

[0042]

[0043]

[0044] 1>K p1 >0

[0045] 1>K>0

[0046] 1>K1>0;

[0047] Wherein, K is the weight of the second probability, N is the number of all records in the historical data, and N m is the preset record upper limit, N min is the preset record lower limit, K1 is a constant, K p1 is a constant, and C is a constant.

[0048] The present disclosure also provides a medical information recommendation device, the device comprising:

[0049] The first acquisition module is used to obtain the self-description content of the current user;

[0050] A first recognition module is used to recognize the physiological characteristics of the current user from the self-description content of the current user;

[0051] a first determining module, configured to determine a probability of a physiological state associated with a physiological characteristic of the current user according to a predetermined graph structure, wherein the graph structure includes an association relationship and a weight relationship between the physiological characteristics and the physiological state;

[0052] The second determining module is configured to determine recommended medical consultation information according to the probability of the physiological state associated with the physiological characteristics of the current user.

[0053] The present disclosure also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the steps of the above method provided by the present disclosure when the program is executed by a processor.

[0054] The present disclosure also provides an electronic device, comprising:

[0055] a memory having a computer program stored thereon;

[0056] A processor is used to execute the computer program in the memory to implement the steps of the above method provided by the present disclosure.

[0057] The above technical solution predetermines a graph structure based on prior knowledge, including the associations and weights between physiological characteristics and physiological states. Based on this graph structure, the probability of the physiological state associated with the identified user's physiological characteristics is determined. Recommended medical information is then determined based on the determined physiological state probabilities. This allows for efficient and accurate recommendations based on the user's self-reported information, minimizing the user's time, alleviating the burden on hospital staff, and reducing the probability of misdiagnosis.

[0058] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following detailed description, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings:

[0060] Figure 1 is a flow chart of a method for recommending medical information provided by an exemplary embodiment;

[0061] Figure 2 is a schematic diagram of a graph structure provided by an exemplary embodiment;

[0062] Figure 3 is a flowchart of a method for recommending medical information provided by another exemplary embodiment;

[0063] Figure 4 is a flowchart of a method for recommending medical information provided by another exemplary embodiment;

[0064] Figure 5 is a flowchart of a method for recommending medical information provided by another exemplary embodiment;

[0065] Figure 6 is a block diagram of a medical information recommendation device provided by an exemplary embodiment;

[0066] Figure 7 It is a block diagram of an electronic device shown in an exemplary embodiment. DETAILED DESCRIPTION

[0067] The following describes the specific embodiments of the present disclosure in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present disclosure and are not intended to limit the present disclosure.

[0068] Figure 1 FIG. 1 is a flowchart of a method for recommending medical information provided by an exemplary embodiment. Figure 1 As shown, the method may include steps 11 to 14.

[0069] In step 11, the self-description content of the current user is obtained.

[0070] The self-report content can be edited text content, and users can edit their medical conditions into text. For older users who may not be good at typing, users can orally describe their medical conditions, and the integrated speech recognition module can realize speech-to-text conversion.

[0071] In step 12, the physiological characteristics of the current user are identified from the self-description content of the current user.

[0072] In the present disclosure, the user may be a patient who is preparing to see a doctor, the physiological characteristics may be what are commonly referred to as symptoms, the physiological state mentioned below may be what are commonly referred to as diseases, and the medical information may be the medical department or doctor to be seen, etc.

[0073] After obtaining the user's self-description, we can perform structured extraction on it. For example, we can use entity recognition to identify physiological features within the self-description. This entity recognition method can be implemented using Bert+BiLSTM+CRF. Physiological features can include symptoms such as "dizziness," "headache," "fatigue," "memory loss," and "high blood pressure." We can leverage prior knowledge to pre-construct a dictionary of physiological features and store common symptom terms in this dictionary. During entity recognition, if a word in the self-description can be found in this dictionary, it is identified as a physiological feature of the current user.

[0074] In step 13, the probability of the physiological state associated with the physiological characteristics of the current user is determined according to a predetermined graph structure, where the graph structure includes an association relationship and a weight relationship between the physiological characteristics and the physiological state.

[0075] The probability of a physiological state associated with the current user's physiological characteristics is the probability that the current user is in that physiological state, or in other words, the probability that the current user suffers from a disease associated with the symptoms identified in their self-reported content. For example, if the current user's self-reported content identifies the symptoms (physiological characteristics) of "dizziness," "headache," and "high blood pressure," and the diseases (physiological states) associated with these symptoms are "hypertension" and "hypoglycemia," this step determines the probability that the current user suffers from "hypertension" and "hypoglycemia," respectively.

[0076] The atlas structure needs to be constructed in advance. The atlas structure includes the association relationship and weight relationship between physiological characteristics and physiological states. Among them, a physiological state can have multiple physiological characteristics associated with it, and a physiological characteristic can also have multiple physiological states associated with it. For example, the physiological characteristics of "dizziness", "headache" and "high blood pressure" can correspond to the same physiological state "hypertension", and the physiological characteristic of "dizziness" can correspond to at least two physiological states "hypertension" and "hypoglycemia". In addition, there can be hierarchical relationships in the atlas structure. For example, if "long-term cough" is identified, according to the hierarchical relationship of the atlas, "long-term cough" can be classified as the superordinate "cough", but if "cough" is identified, it cannot be classified as "long-term cough".

[0077] Figure 2 is a schematic diagram of a graph structure provided by an exemplary embodiment. Figure 2 In the embodiment, multiple symptoms such as "dizziness", "headache", "fatigue", "memory loss" and "high blood pressure" are all associated with the disease of "hypertension". "Long-term dizziness" and "short-term dizziness" can both be classified as "dizziness".

[0078] In addition, the graph structure also includes the weight relationship between physiological characteristics and physiological states set according to prior knowledge ( Figure 2 (not shown), that is, the contribution of a physiological feature to the physiological state associated with it, and the size of the weight reflects the possibility of determining a physiological state by a physiological feature.

[0079] For example, if "high blood pressure" is associated with five symptoms: "dizziness," "headache," "fatigue," "memory loss," and "high blood pressure," the weights of these five symptoms on "high blood pressure" are 3, 1, 1, 2, and 5, respectively. If the probability of "dizziness" alone determining "high blood pressure" is 1, then the probability of "high blood pressure" determining "high blood pressure" is 5. If the user's self-report includes these five symptoms, then all five symptoms contribute to the user's "high blood pressure."

[0080] It is understood that when the same physiological characteristic is associated with multiple physiological states, it may contribute different weights to each associated physiological state. For example, "dizziness" has a weight of 3 for "high blood pressure" and a weight of 1 for "cold." Furthermore, the sum of the weights of all physiological characteristics associated with a physiological state can be greater than 1, equal to 1, or less than 1. Therefore, this weight only reflects the magnitude of the probability and does not have normalized restrictions.

[0081] The probability of the physiological state associated with the physiological feature of the current user depends on the weight relationship between the physiological feature of the current user and the physiological state associated with it.

[0082] In step 14 , recommended medical consultation information is determined based on the probability of the physiological state associated with the physiological characteristics of the current user.

[0083] The medical information can be the medical department or the doctor. The specific association between the physiological state and the medical information can be set based on prior knowledge, and the recommended medical information can be determined based on the probability of the physiological state.

[0084] The probability of a physiological state is the likelihood that the user is in that physiological state. A user can have one or more physiological states. The recommended medical records can also be one or more. If there is only one physiological state, the medical records associated with that physiological state can be directly determined as the recommended medical records. If there are multiple physiological states, the recommended medical records can be determined based on the probability of each physiological state. For example, the recommended medical records may be the medical records corresponding to the one or two physiological states with the highest probability among the multiple physiological states.

[0085] The above technical solution predetermines a graph structure based on prior knowledge, including the associations and weights between physiological characteristics and physiological states. Based on this graph structure, the probability of the physiological state associated with the identified user's physiological characteristics is determined. Recommended medical information is then determined based on the determined physiological state probabilities. This allows for efficient and accurate recommendations based on the user's self-reported information, minimizing the user's time, alleviating the burden on hospital staff, and reducing the probability of misdiagnosis.

[0086] Figure 3 FIG. 1 is a flow chart of a method for recommending medical information provided by another exemplary embodiment. Figure 3 As shown, in Figure 1 On the basis of the above, step 13 of determining the probability of the physiological state associated with the physiological characteristics of the current user according to the predetermined graph structure may include steps 131 to 133.

[0087] In step 131 , a physiological state associated with the physiological characteristics of the current user in the graph structure is determined.

[0088] In the predetermined graph structure, the association between physiological characteristics and physiological states is set based on prior knowledge. Therefore, it is easy to find the physiological state associated with the physiological characteristics of the current user, and usually multiple associated physiological states can be found.

[0089] In step 132 , for each physiological state associated with the physiological characteristics of the current user, the sum of the weights of the physiological characteristics associated with the physiological state in the physiological characteristics of the current user that contribute to the physiological state is calculated as the weight of the physiological state.

[0090] For example, the associated physiological states found include "hypertension" and "hypoglycemia". For "hypertension", although there are ten physiological characteristics associated with it in the atlas structure, only the sum of the weights of the physiological characteristics of the current user for "hypertension" among these ten physiological characteristics is taken, that is, only the sum of the weights of the five physiological characteristics of "dizziness", "headache", "fatigue", "memory loss" and "high blood pressure" identified in the current user's self-description content for "hypertension" is taken: 3+1+1+2+5=12, which is the weight of "high blood pressure". For "hypoglycemia", the sum of the weights of the current user's physiological characteristics (associated with "hypoglycemia") for "hypoglycemia" is calculated, which is the weight of "hypoglycemia".

[0091] The weight of each physiological state reflects the possibility that the current user is in the physiological state.

[0092] In step 133 , the weights of the physiological states associated with the physiological characteristics of the current user are normalized to obtain the probability of each physiological state associated with the physiological characteristics of the current user.

[0093] That is, the weights of all physiological states associated with the physiological characteristics of the current user are summed up, and the ratio of the weight of each physiological state to the summed result is the probability of the physiological state.

[0094] For example, based on the patient's self-report, the physiological states determined are: "hypertension" with a weight of C1, "diabetes" with a weight of C2, and "renal hypertension" with a weight of C3. Normalizing the weights of these three physiological states yields the probability of "hypertension" as C1 / (C1+C2+C3), the probability of "diabetes" as C2 / (C1+C2+C3), and the probability of "renal hypertension" as C3 / (C1+C2+C3).

[0095] The normalized probabilities of each physiological state reflect the proportion of each physiological state when the total amount is 1.

[0096] In this embodiment, the probability of each physiological state of the current user is obtained by searching the graph structure and performing simple calculations. The method is simple and has high accuracy.

[0097] In another embodiment, the graph structure further includes an inferred relationship between a vital sign parameter and a physiological characteristic. The method further includes: identifying the vital sign parameter of the current user from the self-description of the current user; if the vital sign parameter of the current user is greater than a corresponding threshold, determining the physiological characteristic inferred from the vital sign parameter being greater than the corresponding threshold in the graph structure as the physiological characteristic of the current user.

[0098] A dictionary of physical signs can be pre-built using prior knowledge, and common physical sign terms can be stored in this dictionary. Entity recognition is performed on the self-description. If a word in the self-description can be found in this dictionary, it is identified as the current user's physical sign. The number following a physical sign word in the physical sign dictionary in the self-description is the physical sign parameter. For example, the physical sign parameter for the sign "body temperature" is identified as "38.5°C" in the self-description.

[0099] In the graph structure, the inferred relationship between physical parameters and physiological characteristics can be included. For example, if "body temperature" is greater than 37.2℃, it is inferred that the physiological characteristic of "fever" exists. In this case, even if the entity recognition of the self-description content does not identify the physiological characteristic of "fever", "fever" will be classified as the physiological characteristic of the current user and treated the same as other physiological characteristics obtained by entity recognition. For example, in Figure 2 In the example, when the “blood pressure value” in the self-description content is greater than a predetermined value, it can be inferred that the current user has the physiological characteristic of “high blood pressure”.

[0100] In this embodiment, the physical sign parameters in the current user's self-description content can be identified as physiological characteristics, avoiding information omission, thereby making the recommended medical information more accurate.

[0101] In another embodiment, after step 131 of determining the physiological state associated with the physiological characteristics of the current user in the graph structure, the method further includes: identifying the personal information of the current user from the self-description content of the current user, the personal information including gender and age; and screening the physiological state associated with the physiological characteristics of the current user based on the identified personal information.

[0102] Some physiological conditions are only common among people of a certain age or gender. For example, a male cannot suffer from breast hyperplasia, and an elderly person cannot suffer from polio. Therefore, based on the age and gender identified in the self-reported content, physiological conditions that clearly do not match the current user can be eliminated.

[0103] In this embodiment, the personal information identified in the self-description content can be used to filter out incorrect physiological states, reduce the amount of calculation, and improve the accuracy of recommended medical information.

[0104] Figure 4 FIG. 1 is a flowchart of a method for recommending medical information provided by another exemplary embodiment. The medical information is the medical department. Figure 4 As shown, in Figure 1 On the basis of the above, step 14 of determining the recommended medical consultation information according to the probability of the physiological state associated with the physiological characteristics of the current user may include steps 141 to 143.

[0105] In step 141 , the department corresponding to the physiological state associated with the physiological characteristics of the current user is determined according to a predetermined correspondence relationship.

[0106] Predetermined correspondences are relationships between physiological states and departments, and are pre-established based on prior knowledge. For example, in this correspondence, the departments corresponding to "hypertension" are "Cardiovascular Medicine," "Hypertension Center," and "Geriatrics," while the department corresponding to "diabetes" is "Endocrinology."

[0107] The department corresponding to the physiological state associated with the physiological characteristics of the current user can be found in the corresponding relationship. Regardless of whether it is in the corresponding relationship or for the physiological state of the current user, one physiological state can correspond to multiple departments, and multiple physiological states can also correspond to one department.

[0108] In step 142 , for each determined department, the sum of the probabilities of the physiological states corresponding to the department and associated with the physiological characteristics of the current user is determined as the first probability of the department.

[0109] As described above, one department may correspond to multiple physiological states. For each department determined in step 141, the physiological state associated with the physiological characteristics of the current user is selected from all the corresponding physiological states, and the probabilities of the selected physiological states are summed to obtain the first probability of the department.

[0110] For example, "Cardiovascular Medicine" corresponds to ten physiological states in the above correspondence, but the current user only has two physiological states "hypertension" and "premature beats" corresponding to "Cardiovascular Medicine", then the first probability of "Cardiovascular Medicine" is the sum of the probabilities of "hypertension" and "premature beats".

[0111] When the medical information is a medical department, the first probability reflects the degree of conformity between each department and the current user's condition determined by the self-reported content.

[0112] In addition, the first probability of each department can also be normalized. For example, the departments corresponding to the physiological state associated with the current user's physiological characteristics include "cardiology department" and "endocrinology department", P 心血管内科 、P 内分泌科 are the first probabilities of “Cardiovascular Department” and “Endocrinology Department” before normalization, respectively. Then the first probability of “Cardiovascular Department” after normalization is The first probability of "Endocrinology" after normalization is

[0113] In step 143 , a recommended consultation department is determined based on the first probabilities of the various departments.

[0114] In this embodiment, the degree of conformity between each department and the condition determined by the self-description content is determined through simple probability calculation, so that the medical department that meets the current condition of the user can be easily determined with high accuracy.

[0115] In another embodiment, after step 141 of determining the department corresponding to the physiological state associated with the physiological characteristics of the current user based on a predetermined correspondence, the method further includes: identifying the personal information of the current user from the self-description content of the current user, the personal information including gender and age; and screening the determined department based on the identified personal information.

[0116] Specifically, specific departments can be filtered based on age and gender. For example, if the user is 25, "Geriatrics" can be excluded, and if the user is female, "Andrology" can be excluded. Therefore, based on the age and gender identified in the self-description, departments that clearly do not match the current user can be eliminated.

[0117] In this embodiment, the personal information identified in the self-description content can be used to filter out incorrect departments, reduce the amount of calculation, and improve the accuracy of the recommended medical departments.

[0118] In yet another embodiment, step 143 of determining the recommended department for consultation based on the first probabilities of the various departments includes: determining the department with the highest first probability as the recommended department for consultation.

[0119] The first probability is the highest, indicating that the department is most suitable for the condition described in the self-report. The department with the highest first probability is directly determined as the recommended treatment department. The method is simple and the processing speed is fast.

[0120] In another embodiment, the medical department recommended to the current user may be determined by referring to the medical departments that have been successfully completed and correctly recommended in the history. Figure 5 FIG. 1 is a flowchart of a method for recommending medical information provided by another exemplary embodiment. Figure 5 As shown, in Figure 4 On the basis of the first probability of each department, step 143 of determining the recommended treatment department may include steps 1431 to 1433.

[0121] In step 1431, historical data is obtained, which includes multiple records, each record including self-description content and corresponding department.

[0122] The historical data may include corresponding departments that have been confirmed to be correctly recommended. Such historical data has a relatively high reference value.

[0123] In step 1432, multiple records with self-description contents similar to the self-description content of the current user are filtered out from the historical data.

[0124] For example, a real-time search engine (ElasticSearch, ES) can be used to query the top 30 records. Alternatively, the similarity between the self-description content in each record in the historical data and the self-description content of the current user can be calculated using a cosine similarity algorithm, and the top 30 similarity records can be listed.

[0125] In step 1433, the recommended consultation department is determined based on the first probability of each department and the screened records.

[0126] The departments in the filtered records are recommended medical departments based on historical records. The recommended medical departments are based on the combination of the first probability of each department and the filtered records. Based on the current calculation rules, the historical records are combined with a variety of reference factors to make the recommendation results more accurate and reduce errors.

[0127] In yet another embodiment, step 1433 of determining a recommended department for consultation based on the first probabilities of the respective departments and the screened records includes:

[0128] The second probability of each department is calculated according to the following formula (1): for each department, the target probability of the department is calculated based on the first probability and the second probability of the department; the department with the largest target probability is determined as the recommended department for treatment.

[0129]

[0130] Among them, P 2i is the second probability of the i-th department in the screened records, S i is the maximum value of the similarity of the self-description content in one or more records corresponding to the i-th department in the filtered records, and n is the number of all departments in the filtered records.

[0131] For example, the 30 filtered records correspond to three departments: A, B, and C. The maximum similarity value for the self-descriptions of the 8 records corresponding to Department A is 80, the maximum similarity value for the self-descriptions of the 12 records corresponding to Department B is 90, and the maximum similarity value for the self-descriptions of the 10 records corresponding to Department B is 100. Therefore, the second probability for Department A is 80 ÷ (80 + 90 + 100), the second probability for Department B is 90 ÷ (80 + 90 + 100), and the second probability for Department C is 100 ÷ (80 + 90 + 100).

[0132] It can be seen that, among multiple records corresponding to a department, the greater the maximum value of the self-description content similarity is, the more likely the department is to be applicable to the current user, and the greater its second probability is.

[0133] The first probability and second probability of a department reflect the probability that the department is suitable for the current user, as determined by the graph structure and historical data, respectively. In this embodiment, the first and second probabilities are combined to calculate the target probability to determine the recommended department, further increasing the accuracy of the department recommendation.

[0134] In another embodiment, the above-mentioned calculation of the target probability of each department based on the first probability and the second probability of the department may include: for each department, determining the weighted sum of the first probability and the second probability of the department as the target probability of the department.

[0135] By setting weight values, the participation of the first probability and the second probability in calculating the target probability is allocated. Through a simple weighted summation method, the processing speed is fast and the result is accurate.

[0136] In yet another embodiment, the weight of the second probability is determined according to the following formula (2):

[0137]

[0138]

[0139] 1>K p1 >0

[0140] 1>K>0

[0141] 1>K1>0;

[0142] Among them, K is the weight of the second probability, the weight of the first probability is 1-K, N is the number of all records in the historical data, N m is the preset record upper limit, N min is the preset record lower limit, K1 is a constant, K p1 is a constant, and C is a constant.

[0143] When N min When >N>0, the number of records in the historical data is less than the record lower limit. It can be considered that the number of records in the historical data is small, and a smaller preset constant K1 can be directly used as the weight of the second probability.

[0144] When N>N m When the number of records in the historical data is greater than the upper limit of the record, it can be considered that the number of records in the historical data is large. m >N>N min When the number is large, the weight of the second probability can be increased by C(N) on the basis of K1. m -N min), when the number is appropriate, the weight of the second probability can be increased by C(NN on the basis of K1 min ), so that the weight of the second probability C(N m -N min )+K1 is smaller than the weight of the second probability C(NN min )+K1. In this way, when the number of records in the historical data is appropriate, the second probability has a greater weight.

[0145] And, through Keep the weight K of the second probability between 0 and 1-K p1 So, just set the appropriate K p1 The value of can make the weight K of the second probability not too large, thereby playing a certain protective role on the weight of the first probability.

[0146] By using the above formula (2), the weight of the second probability can be reasonably determined for situations where there are different numbers of records in the historical data, thereby improving the accuracy of the recommended department.

[0147] Figure 6 FIG. 1 is a block diagram of a medical information recommendation device provided by an exemplary embodiment. Figure 6 As shown, the medical information recommendation device 600 may include a first acquisition module 601 , a first identification module 602 , a first determination module 603 and a second determination module 604 .

[0148] The first acquisition module 601 is used to acquire the self-description content of the current user.

[0149] The first recognition module 602 is used to recognize the physiological characteristics of the current user from the self-description content of the current user.

[0150] The first determination module 603 is used to determine the probability of the physiological state associated with the physiological characteristics of the current user according to a predetermined graph structure, where the graph structure includes the association relationship and weight relationship between the physiological characteristics and the physiological state.

[0151] The second determining module 604 is configured to determine recommended medical consultation information based on the probability of the physiological state associated with the physiological characteristics of the current user.

[0152] Optionally, the first determination module 603 includes a first determination submodule, a first calculation submodule and a normalization submodule.

[0153] The first determination submodule is used to determine the physiological state associated with the physiological characteristics of the current user in the graph structure.

[0154] The first calculation submodule is used to calculate, for each physiological state associated with the physiological characteristics of the current user, the sum of the weights of the physiological characteristics associated with the physiological state in the physiological characteristics of the current user that contribute to the physiological state as the weight of the physiological state.

[0155] The normalization submodule is used to normalize the weights of the physiological states associated with the physiological characteristics of the current user to obtain the probability of each physiological state associated with the physiological characteristics of the current user.

[0156] Optionally, the atlas structure also includes an inferred relationship between physical sign parameters and physiological characteristics. The medical information recommendation device 600 also includes a second identification module and a third determination module.

[0157] The second recognition module is used to recognize the vital sign parameters of the current user from the self-description content of the current user.

[0158] The third determination module is used to determine the physiological characteristics inferred in the atlas structure due to the physical parameters being greater than the corresponding threshold as the physiological characteristics of the current user if the physical parameters of the current user are greater than the corresponding threshold.

[0159] Optionally, the medical information recommendation device 600 further includes a third identification module and a first screening module.

[0160] The third identification module is used to identify the personal information of the current user from the self-description content of the current user, and the personal information includes gender and age.

[0161] The first screening module is used to screen the physiological state associated with the physiological characteristics of the current user based on the identified personal information.

[0162] Optionally, the second determining module 604 includes a second determining submodule, a third determining submodule and a fourth determining submodule.

[0163] The second determining submodule is used to determine the department corresponding to the physiological state associated with the physiological characteristics of the current user according to a predetermined corresponding relationship, where the predetermined corresponding relationship is a corresponding relationship between the physiological state and the department.

[0164] The third determination submodule is configured to determine, for each determined department, the sum of the probabilities of the physiological states corresponding to the department and associated with the physiological characteristics of the current user as the first probability of the department.

[0165] The fourth determination submodule is used to determine the recommended consultation department according to the first probability of each department.

[0166] Optionally, the medical information recommendation device 600 further includes a third identification module and a second screening module.

[0167] The third identification module is used to identify the personal information of the current user from the self-description content of the current user, and the personal information includes gender and age.

[0168] The second screening module is used to screen the determined departments based on the identified personal information.

[0169] Optionally, the fourth determination submodule is used to determine the department with the highest first probability as the recommended department for consultation.

[0170] Optionally, the fourth determining submodule includes a fifth determining submodule, a screening submodule and a sixth determining submodule.

[0171] The fifth determination submodule is used to obtain historical data, which includes multiple records, each of which includes self-description content and the corresponding department.

[0172] The filtering submodule is used to filter out multiple records whose self-description contents are similar to the self-description contents of the current user from historical data.

[0173] The sixth determination submodule is used to determine the recommended treatment department based on the first probability of each department and the screened records.

[0174] Optionally, the sixth determining submodule includes a second calculating submodule, a third calculating submodule and a seventh determining submodule.

[0175] The second calculation submodule is used to calculate the second probability of each department according to the following formula:

[0176]

[0177] Among them, P 2i is the second probability of the i-th department in the screened records, S i is the maximum value of the similarity of the self-description content in one or more records corresponding to the i-th department in the filtered records, and n is the number of all departments in the filtered records.

[0178] The third calculation submodule is used to calculate the target probability of each department according to the first probability and the second probability of the department.

[0179] The seventh determination submodule is used to determine the department with the highest target probability as the recommended treatment department.

[0180] Optionally, the third calculation submodule is configured to determine, for each department, a weighted sum of the first probability and the second probability of the department as the target probability of the department.

[0181] Optionally, the weight of the second probability is determined according to the following formula:

[0182]

[0183]

[0184] 1>K p1 >0

[0185] 1>K>0

[0186] 1>K1>0;

[0187] Among them, K is the weight of the second probability, N is the number of all records in the historical data, N m is the preset record upper limit, N min is the preset record lower limit, K1 is a constant, K p1 is a constant, and C is a constant.

[0188] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0189] The above technical solution predetermines a graph structure based on prior knowledge, including the associations and weights between physiological characteristics and physiological states. Based on this graph structure, the probability of the physiological state associated with the identified user's physiological characteristics is determined. Recommended medical information is then determined based on the determined physiological state probabilities. This allows for efficient and accurate recommendations based on the user's self-reported information, minimizing the user's time, alleviating the burden on hospital staff, and reducing the probability of misdiagnosis.

[0190] The present disclosure also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which implements the steps of the above method provided by the present disclosure when the program is executed by a processor.

[0191] The present disclosure further provides an electronic device comprising a memory and a processor. The memory stores a computer program. The processor is configured to execute the computer program in the memory to implement the steps of the above method provided by the present disclosure.

[0192] Figure 7 FIG. 7 is a block diagram of an electronic device 700 according to an exemplary embodiment. Figure 7 As shown, the electronic device 700 may include: a processor 701 , a memory 702 , and may further include one or more of a multimedia component 703 , an input / output (I / O) interface 704 , and a communication component 705 .

[0193] The processor 701 is used to control the overall operation of the electronic device 700 to complete all or part of the steps in the above-mentioned medical information recommendation method. The memory 702 is used to store various types of data to support the operation of the electronic device 700. Such data may include, for example, instructions for any application or method operating on the electronic device 700, as well as application-related data, such as contact information, sent and received messages, pictures, audio, video, etc. The memory 702 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 703 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 702 or sent via the communication component 705. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 704 provides an interface between the processor 701 and other interface modules. The above-mentioned other interface modules may be a keyboard, a mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 705 is used for wired or wireless communication between the electronic device 700 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IOT, eMTC, or other 5G, etc., or a combination of one or more thereof, is not limited here. Therefore, the corresponding communication component 705 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0194] In an exemplary embodiment, the electronic device 700 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned medical information recommendation method.

[0195] In another exemplary embodiment, a computer-readable storage medium including program instructions is further provided. When executed by a processor, the program instructions implement the steps of the aforementioned medical information recommendation method. For example, the computer-readable storage medium may be the aforementioned memory 702 including the program instructions. The program instructions may be executed by the processor 701 of the electronic device 700 to implement the aforementioned medical information recommendation method.

[0196] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings. However, the present disclosure is not limited to the specific details of the above embodiments. Within the technical concept of the present disclosure, various simple modifications can be made to the technical solutions of the present disclosure, and these simple modifications all fall within the scope of protection of the present disclosure.

[0197] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0198] In addition, the various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. A method for recommending medical information, characterized in that: The method comprises: Get the current user's self-introduction content; Identifying physiological characteristics of the current user from the self-reported content of the current user; Determining the probability of the physiological state associated with the physiological characteristics of the current user according to a predetermined graph structure, wherein the graph structure includes an association relationship and a weight relationship between the physiological characteristics and the physiological state; Determining recommended medical consultation information according to the probability of the physiological state associated with the physiological characteristics of the current user, wherein the medical consultation information is a medical department; The step of determining the recommended medical consultation information according to the probability of the physiological state associated with the physiological characteristics of the current user includes: Determining the department corresponding to the physiological state associated with the physiological characteristics of the current user according to a predetermined corresponding relationship, wherein the predetermined corresponding relationship is a corresponding relationship between the physiological state and the department; For each determined department, determine the sum of the probabilities of the physiological states corresponding to the department and associated with the physiological characteristics of the current user as the first probability of the department; The second probability of each department is calculated according to the following formula: ； For each department, a weighted sum of the first probability and the second probability of the department is determined as the target probability of the department; The department with the highest target probability is determined as the recommended consultation department; in, The first The second probability of each department, The records that are filtered out correspond to The maximum value of the similarity of the self-report content in one or more records of a department, is the number of all departments in the filtered records; The weight of the second probability is determined according to the following formula: ； in, is the weight of the second probability, is the number of all records in the historical data, is the preset record limit, is the preset record lower limit, is a constant, is a constant, Is a constant.

2. The method according to claim 1, characterized in that The determining, according to a predetermined graph structure, a probability of a physiological state associated with a physiological characteristic of the current user comprises: Determining a physiological state associated with a physiological characteristic of the current user in the graph structure; For each physiological state associated with the physiological characteristics of the current user, respectively, calculate the sum of weights of contributions of the physiological characteristics associated with the physiological state in the physiological characteristics of the current user to the physiological state as the weight of the physiological state; The weights of the physiological states associated with the physiological characteristics of the current user are normalized to obtain the probability of each physiological state associated with the physiological characteristics of the current user.

3. The method according to claim 1, characterized in that: The atlas structure also includes an inferred relationship between the vital sign parameters and the physiological characteristics, and the method further includes: Identifying the vital sign parameters of the current user from the self-description content of the current user; If the physical sign parameter of the current user is greater than the corresponding threshold, the physiological feature inferred in the graph structure due to the physical sign parameter being greater than the corresponding threshold is determined as the physiological feature of the current user.

4. The method according to claim 2, characterized in that: After determining the physiological state associated with the physiological characteristics of the current user in the graph structure, the method further includes: Identify the personal information of the current user from the self-description content of the current user, the personal information including gender and age; The physiological state associated with the physiological characteristics of the current user is screened according to the identified personal information.

5. The method according to claim 1, characterized in that After determining the department corresponding to the physiological state associated with the physiological characteristics of the current user according to the predetermined corresponding relationship, the method further includes: Identify the personal information of the current user from the self-description content of the current user, the personal information including gender and age; The identified departments are screened based on the identified personal information.

6. The method according to claim 1, characterized in that The method further comprises: Acquire historical data, wherein the historical data includes a plurality of records, each record including self-description content and a corresponding department; A plurality of records having self-description contents similar to the self-description contents of the current user are screened out from the historical data.

7. A medical information recommendation device, characterized in that: The device comprises: The first acquisition module is used to acquire the self-description content of the current user; A first recognition module, used to recognize the physiological characteristics of the current user from the self-description content of the current user; A first determination module, configured to determine a probability of a physiological state associated with a physiological feature of the current user according to a predetermined graph structure, wherein the graph structure includes an association relationship and a weight relationship between the physiological feature and the physiological state; A second determination module is used to determine recommended medical information according to the probability of the physiological state associated with the physiological characteristics of the current user, wherein the medical information is a medical department; The second determination module includes a second determination submodule, a third determination submodule, a second calculation submodule, a third calculation submodule and a seventh determination submodule. The second determination submodule is used to determine the department corresponding to the physiological state associated with the physiological characteristics of the current user according to a predetermined corresponding relationship, where the predetermined corresponding relationship is a corresponding relationship between the physiological state and the department; The third determination submodule is used to determine, for each determined department, the sum of the probabilities of the physiological states corresponding to the department and associated with the physiological characteristics of the current user as the first probability of the department; The second calculation submodule is used to calculate the second probability of each department according to the following formula: ； in, The first The second probability of each department, The records that are filtered out correspond to The maximum value of the similarity of the self-report content in one or more records of a department, is the number of all departments in the filtered records; The third calculation submodule is used to determine, for each department, a weighted sum of the first probability and the second probability of the department as the target probability of the department; The seventh determination submodule is used to determine the department with the highest target probability as the recommended consultation department; The weight of the second probability is determined according to the following formula: ； in, is the weight of the second probability, is the number of all records in the historical data, is the preset record limit, is the preset record lower limit, is a constant, is a constant, Is a constant.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 6 are implemented.

9. An electronic device, characterized in that: include: a memory having a computer program stored thereon; A processor, configured to execute the computer program in the memory to implement the steps of the method according to any one of claims 1 to 6.

Images