Ear disease prediction method and system based on big data

Abstract

The invention discloses an ear disease prediction method and system based on big data, and relates to the technical field of data analysis, and the method comprises the following steps: obtaining sudden ear disease types and corresponding symptoms, updating the sudden ear disease types according to a first disease type, classifying the same sudden ear disease types in third data, and determining the corresponding symptoms of the sudden ear disease types in the third data; and obtaining a first classification, calculating a first time period, fitting the sub-equation according to a least square method, and predicting a first number of the next third data and a time interval between the next third data and the current time point. According to the method, medical big data is analyzed and mined, patient satisfaction is improved, a time sequence analysis method is used, the disease development trend of a patient is captured, the future disease possibility is further predicted, an equation of sub-data and time is fitted through a least square method, a first number of next third data is predicted, and the probability of the disease in the future is predicted. And the time interval between the next third data and the current time point is calculated, so that the occurrence time of the ear disease is predicted.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to an ear disease prediction method and system based on big data. Background Art

[0002] In recent years, artificial intelligence and machine learning technologies have been widely applied in the field of otolaryngology. Especially in the diagnosis, treatment, and prognosis prediction of ear diseases, these technologies assist medical decision-making by deeply mining clinical data, analyzing radiomics features, establishing disease prediction models, etc.

[0003] Currently, in the Chinese invention patent with the publication number CN 111933288 A, a congenital deafness disease prediction method, system, and terminal based on CNN are disclosed. This method analyzes common gene loci of deafness, designs an adaptive machine learning algorithm model relying on a big data platform, and realizes fast and accurate deafness gene data analysis and prediction through a large amount of training. However, in the related technologies, it is not considered that people are under great life pressure nowadays, which is likely to cause sudden deafness. There is a lack of applicability in prediction, and the prediction of ear diseases is not combined with big data analysis and expert opinions, lacking the authority and accuracy of ear disease prediction. The data to be analyzed is not reasonably screened according to various factors, which is not conducive to the rapid response of calculation, is prone to prediction errors, and has certain limitations. Summary of the Invention

[0004] The technical problem solved by the present invention is that in the related technologies, it is not considered that people are under great life pressure nowadays, which is likely to cause sudden deafness. There is a lack of applicability in prediction, and the prediction of ear diseases is not combined with big data analysis and expert opinions, lacking the authority and accuracy of ear disease prediction. The data to be analyzed is not reasonably screened according to various factors, which is not conducive to the rapid response of calculation, is prone to prediction errors, and has certain limitations.

[0005] To solve the above technical problem, the present invention provides the following technical solutions: In the first aspect, an ear disease prediction method based on big data includes the following steps:

[0006] Step S100: Obtain the patient number, match the patient's case according to the patient number, extract the relevant data of the patient's case, where the relevant data includes the visit number, past medical history, current symptoms, and diagnosis code, and obtain the patient's self-test record;

[0007] Step S200: Count the number of the patient's visit numbers, denoted as the number of visits. Set the first sequence according to the order of the visit numbers, set the second sequence according to the past medical history and current symptoms, count the number of times each word appears in the second sequence, denoted as the first count, set the fourth sequence according to the diagnosis code, and set the fifth sequence according to the time interval between visits;

[0008] Step S300: Set the position of the third sequence in the second sequence as the first label, set the position of the fourth element in the first sequence as the second label, perform a first comparison between the position of the second label and the position of the fourth label, delete invalid data according to the first comparison result, delete the corresponding positions in the third element, the fourth element, and the fifth element, and update the value of the next position corresponding to the position in the fifth element, and obtain the data corresponding to the updated second label and the data corresponding to the third label, denoted as the first data and the second data;

[0009] Step S400: Obtain the types of sudden ear diseases and their corresponding symptoms, perform a second comparison between the types of sudden ear diseases and the types of sudden diseases, and insert the types of diseases into the fifth sequence according to the corresponding number of medical visits according to the second comparison result, or retrieve the expert platform, obtain the fifth element adjacent to the current symptoms of the sudden disease in the fifth sequence, obtain the patient self-test record corresponding to the adjacent fifth element, input the current symptoms of the sudden disease and the patient self-test record corresponding to the fifth element into the expert platform, obtain the corresponding type of disease, denoted as the first type of disease, update the types of sudden diseases according to the first type of disease, and insert the first type of disease into the fifth sequence according to the corresponding number of medical visits;

[0010] Step S500: Obtain the fifth sequence after inserting the types of diseases and the first type of disease, denoted as the sixth sequence, classify the same types of sudden ear diseases in the third data according to the time sequence, obtain the first classification, perform a first numbering on the first sub-classification, calculate the time interval between the sub-data of the first classification, denoted as the first time period, and fit the equation of the sub-data and time according to the least squares method, denoted as the sub-equation;

[0011] Step S600: Obtain the current time point, calculate the time interval between the adjacent third data and the current time point, predict the first number of the next third data according to the first number and the time interval, and the time interval between the next third data and the current time point.

[0012] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S100 includes the following sub-steps:

[0013] Step S101: Obtain the patient number, retrieve the case database, input the patient number into the case database, and match the patient case corresponding to the patient number;

[0014] Step S102: Extract relevant data of the patient's case. The relevant data includes the visit number, past medical history, current symptoms, and diagnosis code. The diagnosis code is the type of sudden disease determined by the doctor, including car accident injury, fall from height injury, burn, fracture due to fall, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergy, food allergy, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding, and sudden ear disease. There is only one and it corresponds to the visit number;

[0015] Different visit numbers indicate that the same patient visits at different times;

[0016] Step S103: Obtain the patient's self-test records. The patient's self-test records represent the degree of hearing impairment and physical sign data self-tested by the patient at each time interval. The physical sign data includes heart rate, respiratory rate, blood pressure, blood glucose, body temperature, and pain level.

[0017] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the extraction logic of the relevant data of the patient's case includes:

[0018] Set the visit number, past medical history, current symptoms, and diagnosis code as keywords of the natural language extractor. According to the keywords, identify the paragraph box of the text, extract the text after the keywords corresponding to the paragraph box, and set the keywords as the labels of the text to obtain the key data of the patient's case. Remove redundant words and punctuation marks in the key data. The redundant words include prepositions, conjunctions, articles, and pronouns to obtain the relevant data of the patient's case.

[0019] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S200 includes the following sub-steps:

[0020] Step S201: Count the number of the patient's visit numbers, denoted as the number of visits;

[0021] Step S202: Set the first sequence according to the order of the visit numbers, and set the second sequence according to the past medical history and current symptoms. The number of data in the first sequence and the number of data in the second sequence are equal to the number of visits. Count the number of times each word appears in the second sequence, denoted as the first count. Set the third sequence according to the first count. Count the number of times the diagnosis code appears, denoted as the second count. Set the fourth sequence according to the second count. Set the fifth sequence according to the time interval between visits. The time interval between visits represents the number of days from this visit to the previous visit.

[0022] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S202 further includes the following sub-steps:

[0023] Step S2021, set in the order of the visit numbers as the first sequence. The elements included in the first sequence are denoted as first elements, and the first sequence is expressed as X = {x1, x2,..., x T};

[0024] Step S2022, set the past medical history and current symptoms as the second sequence. The elements included in the second sequence are denoted as second elements, and the second sequence is expressed as Y = {y1, y2,..., y T}, where T represents the number of visits of the patient, and each second element represents the past medical history and current symptoms corresponding to the number of visits;

[0025] Step S2023, count the number of times each word appears in the second sequence, denoted as the first count. Retain the words with the value of the first count being 1, and delete the words with the value of the first count not being 1. Set the retained words in the order of the subscripts of the original second elements as the third sequence, and number the third sequence with the third number, where the third number is a natural number. Establish the first mapping relationship between the third label and the subscript of the original second element. The elements included in the third sequence are denoted as third elements, and the third sequence is expressed as K = {k1, k2,..., k M}, where M ranges from 1 to T, and M is a natural number;

[0026] Step S2024, count the number of times the diagnostic code appears, denoted as the second count. Retain the words with the value of the second count being 1, and delete the words with the value of the second count not being 1. Obtain the subscripts of the first elements corresponding to the retained words. Set the retained words in the order of the subscripts of the corresponding first elements as the fourth sequence, and number the fourth sequence with the fourth number, where the fourth number is a natural number. Establish the second mapping relationship between the fourth label and the subscript of the original first element. The elements included in the fourth sequence are denoted as fourth elements, and the fourth sequence is expressed as C = {c1, c2,..., C N}, where C ranges from 1 to T, and C is a natural number;

[0027] Step S2025, set the time interval between visits as the fifth sequence, and the fifth sequence is expressed as L = {l1, l2,..., l U}, where U ranges from 1 to T - 1, and U is a natural number.

[0028] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S300 includes the following sub-steps:

[0029] Step S301: Obtain the subscript of the third element, get the original subscript of the second element according to the first mapping relationship, mark the corresponding position of the third element in the second sequence in red, and set the first label, where the first label is the current symptom of the sudden onset;

[0030] Step S302: Obtain the subscript of the fourth element, get the original subscript of the first element according to the second mapping relationship, mark the corresponding position of the fourth element in the first sequence in red, and set the second label, where the second label is the type of disease of the sudden onset;

[0031] Step S303: Perform a first comparison between the position of the second label and the position of the fourth label to obtain a first comparison result. Delete the invalid data according to the first comparison result, delete the corresponding positions in the third element, the fourth element, and the fifth element, and update the value of the next position corresponding to the position in the fifth element;

[0032] The first comparison result is that the position of the second label is the same as the position of the fourth label, and the position of the second label is different from the position of the fourth label;

[0033] When the first comparison result is that the position of the second label is different from the position of the fourth label, delete the corresponding positions in the third element, the fourth element, and the fifth element;

[0034] When the first comparison result is that the position of the second label is the same as the position of the fourth label, and update the value of the next position corresponding to the position in the fifth element. The update logic of the value of the next position includes:

[0035] Obtain the fifth element from the corresponding position in the fifth element to the next position corresponding to the position in the fifth element, calculate the sum value of the fifth element, denoted as the first sum value. Obtain the fifth element from the previous position corresponding to the position in the fifth element to the fifth element corresponding to the position in the fifth element, calculate the sum value of the fifth element, denoted as the second sum value. Calculate the sum value of the first sum value and the second sum value, denoted as the third sum value. Replace the first sum value with the third sum value to obtain the new value of the next position corresponding to the position in the fifth element;

[0036] Step S304: Obtain the data corresponding to the updated second label and the data corresponding to the third label, denoted as the first data and the second data.

[0037] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S400 includes the following sub-steps:

[0038] Step S401: Obtain the types of sudden ear diseases and their corresponding symptoms. The types of sudden ear diseases include mild sudden ear diseases, moderate sudden ear diseases, severe sudden ear diseases, and extremely severe sudden ear diseases;

[0039] The symptoms of mild sudden ear diseases are unilateral hearing loss. The symptoms of moderate sudden ear diseases include unilateral or bilateral hearing loss, tinnitus, accompanied by dizziness and sleep disorders. The symptoms of severe sudden ear diseases are bilateral hearing loss, tinnitus, dizziness, ear fullness, and occasionally accompanied by sleep disorders. The symptoms of extremely severe sudden ear diseases are deafness, dizziness, ear fullness, anxiety, sleep disorders, nausea, vomiting, and cold sweats;

[0040] Step S402: Conduct a second comparison between the types of sudden ear diseases and the types of sudden diseases to obtain a second comparison result. The second comparison result includes that the type of sudden disease belongs to the types of sudden ear diseases and that the type of sudden disease does not belong to the types of sudden ear diseases;

[0041] Step S403: When the second comparison result is that the type of sudden disease belongs to the types of sudden ear diseases, insert the type of disease into the fifth sequence according to the corresponding number of visits;

[0042] When the second comparison result is that the type of sudden disease does not belong to the types of sudden ear diseases, retrieve the expert platform, obtain the fifth element adjacent to the current symptoms of the sudden disease in the fifth sequence, obtain the patient self-test records corresponding to the adjacent fifth element, input the current symptoms of the sudden disease and the patient self-test records corresponding to the fifth element into the expert platform, and obtain the corresponding type of disease, denoted as the first type of disease. The first type of disease includes car accident injuries, falls from heights, burns, fractures from falls, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergies, food allergies, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding, and sudden ear diseases. The adjacent self-test records are expressed as the previous element and the next element of the current symptoms of the sudden disease in the fifth sequence;

[0043] Step S403: Update the type of sudden disease according to the first type of disease, and insert the first type of disease into the fifth sequence according to the corresponding number of visits.

[0044] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S500 includes the following sub-steps:

[0045] Step S501: Obtain the fifth sequence after inserting the types of diseases and the first type of disease, denoted as the sixth sequence. The types of diseases and the first type of disease are both expressed as the types of sudden ear diseases, and express the types of diseases and the first type of disease as the third data;

[0046] Step S502: Classify the same types of sudden deafness in the third data in chronological order to obtain the first classification, and assign the first serial number to the first classification. The first serial number is a natural number.

[0047] Step S503: Calculate the time interval between the sub-data of the first classification, denoted as the first time period. The calculation logic for the time interval between the sub-data of the first classification includes:

[0048] Obtain the fifth element between the sub-data, calculate the sum value of the fifth element, denoted as the fourth sum value, and set the fourth sum value as the time interval between the sub-data.

[0049] Step S503: Fit the equation of the sub-data and time according to the least squares method, denoted as the sub-equation, and assign the second serial number to the sub-equation. The second serial number is a natural number and corresponds to the first serial number.

[0050] As a preferred solution of the ear disease prediction method based on big data according to the present invention, wherein: the step S600 includes the following sub-steps:

[0051] Step S601: Obtain the current time point, and according to the sixth sequence, obtain the adjacent third data of the current time point. The adjacent third data represents the third data with the closest time and its first serial number.

[0052] Step S602: Calculate the time interval between the adjacent third data and the current time point, and predict the first serial number of the next third data and the time interval between the next third data and the current time point according to the first serial number and the time interval. The prediction logic for the first serial number of the next third data includes:

[0053] Obtain the first serial number of the adjacent third data, and according to the serial number order, obtain the first serial number of the next third data.

[0054] The calculation logic for the time interval between the next third data and the current time point is:

[0055] Obtain the current time point, obtain the first serial number of the next third data, retrieve the sub-equation corresponding to the first serial number of the next third data, calculate the time point of the next third data, denoted as the future time point, calculate the difference between the future time point and the current time point, denoted as the first difference, and set the first difference as the time interval between the next third data and the current time point.

[0056] In the second aspect, an ear disease prediction system based on big data includes a collection module, an analysis module, and a calculation module.

[0057] The acquisition module is used to obtain the patient number, match the patient's case according to the patient number, extract the relevant data of the patient's case, where the relevant data includes the visit number, past medical history, current symptoms, and diagnosis code, and obtain the patient's self-test record;

[0058] The analysis module is used to count the number of visit numbers of the patient, denoted as the number of visits, set the first sequence according to the order of the visit numbers, set the second sequence according to the past medical history and current symptoms, count the number of times each word appears in the second sequence, denoted as the first count, set the fourth sequence according to the diagnosis code, set the fifth sequence according to the time interval between visits, set the position of the third sequence in the second sequence as the first label, set the position of the fourth element in the first sequence as the second label, perform the first comparison between the position of the second label and the position of the fourth label, delete the invalid data according to the first comparison result, delete the corresponding positions in the third element, the fourth element, and the fifth element, and update the value of the next position corresponding to the position in the fifth element, obtain the data corresponding to the updated second label and the third label, denoted as the first data and the second data, obtain the types of sudden ear diseases and their corresponding symptoms, perform the second comparison between the types of sudden ear diseases and the types of sudden diseases, and insert the types of diseases into the fifth sequence according to the corresponding number of visits according to the second comparison result, or call the expert platform, obtain the fifth element adjacent to the current symptoms of the sudden onset in the fifth sequence, obtain the patient self-test record corresponding to the adjacent fifth element, input the current symptoms of the sudden onset and the patient self-test record corresponding to the fifth element into the expert platform, and obtain the corresponding type of disease, denoted as the first type of disease, update the types of sudden diseases according to the first type of disease, and insert the first type of disease into the fifth sequence according to the corresponding number of visits;

[0059] The calculation module obtains the fifth sequence after inserting the types of diseases and the first type of disease, denoted as the sixth sequence, classifies the same types of sudden ear diseases in the third data in chronological order to obtain the first classification, assigns the first number to the first sub-classification, calculates the time interval between the sub-data of the first classification, denoted as the first time period, fits the equation of the sub-data and time according to the least squares method, denoted as the sub-equation, obtains the current time point, calculates the time interval between the adjacent third data and the current time point, and predicts the first number of the next third data and the time interval between the next third data and the current time point according to the first number and the time interval.

[0060] Advantages of the present invention: By analyzing and mining medical big data, individualized diagnosis and treatment plans are realized, patient satisfaction is improved. Medical health big data analysis helps doctors predict possible diseases of patients, which is of great significance for the early prevention and control of diseases. Using time series analysis methods to capture the development trend of patients' diseases, and then predict possible future diseases, which is particularly important for the prediction of ear diseases. By fitting the equation of sub-data and time using the least squares method, predict the first number of the next third data and the time interval between the next third data and the current time point, so as to realize the prediction of the occurrence time of ear diseases. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 FIG. is a schematic flowchart of the basic process of the ear disease prediction method based on big data provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be made in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments.

[0063] Embodiment, referring to Figure 1 , an embodiment of the present invention provides an ear disease prediction method based on big data, including the following steps:

[0064] Step S100, obtain the patient number, match the patient's case according to the patient number, extract the relevant data of the patient's case, the relevant data includes the visit number, past medical history, current symptoms, and diagnosis code, and obtain the patient's self-test record;

[0065] Step S200, count the number of the patient's visit numbers, denoted as the number of visits, set the first sequence according to the order of the visit numbers, set the second sequence according to the past medical history and current symptoms, count the number of times each word appears in the second sequence, denoted as the first number, set the fourth sequence according to the diagnosis code, and set the fifth sequence according to the time interval between visits;

[0066] Step S300, set the position of the third sequence in the second sequence as the first label, set the position of the fourth element in the first sequence as the second label, perform the first comparison between the position of the second label and the position of the fourth label, delete the invalid data according to the first comparison result, delete the corresponding positions in the third element, the fourth element, and the fifth element, and update the value of the next position corresponding to the position in the fifth element, and obtain the data corresponding to the updated second label and the data corresponding to the third label, denoted as the first data and the second data;

[0067] Step S400: Obtain the types of sudden ear diseases and their corresponding symptoms, conduct a second comparison between the types of sudden ear diseases and the types of the sudden diseases, insert the types of diseases into the fifth sequence according to the corresponding number of medical visits based on the result of the second comparison, or retrieve the expert platform, obtain the fifth element adjacent to the current symptoms of the sudden disease in the fifth sequence, obtain the patient self-test records corresponding to the adjacent fifth element, input the current symptoms of the sudden disease and the patient self-test records corresponding to the fifth element into the expert platform to obtain the corresponding type of disease, denoted as the first type of disease, update the type of the sudden disease according to the first type of disease, and insert the first type of disease into the fifth sequence according to the corresponding number of medical visits.

[0068] Step S500: Obtain the fifth sequence after inserting the types of diseases and the first type of disease, denoted as the sixth sequence. Classify the same types of sudden ear diseases in the third data according to the time sequence to obtain the first classification, assign the first number to the first sub-classification, calculate the time interval between the sub-data of the first classification, denoted as the first time period, and fit the equation of the sub-data and time by the least squares method, denoted as the sub-equation.

[0069] Step S600: Obtain the current time point, calculate the time interval between the adjacent third data and the current time point, predict the first number of the next third data and the time interval between the next third data and the current time point according to the first number and the time interval.

[0070] The present invention realizes individualized diagnosis and treatment plans by analyzing and mining medical big data, improves patient satisfaction. The analysis of medical and health big data helps doctors predict the possible diseases of patients, which is of great significance for the early prevention and control of diseases. Using the time series analysis method to capture the development trend of patients' diseases, and then predict the possible future diseases, which is particularly important for the prediction of ear diseases. By fitting the equation of sub-data and time by the least squares method, predict the first number of the next third data and the time interval between the next third data and the current time point, so as to realize the prediction of the occurrence time of ear diseases.

[0071] The step S100 includes the following sub-steps:

[0072] Step S101: Obtain the patient number, retrieve the case database, input the patient number into the case database, and match the patient case corresponding to the patient number.

[0073] Step S102: Extract relevant data of the patient's case. The relevant data includes the visit number, past medical history, current symptoms, and diagnosis code. The diagnosis code is the type of sudden disease determined by the doctor, including car accident injury, fall from height injury, burn, fracture due to fall, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergy, food allergy, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding, and sudden ear disease. There is only one such code and it corresponds to the visit number;

[0074] Different visit numbers indicate that the same patient visits at different times;

[0075] Step S103: Obtain the patient's self-test records. The patient's self-test records represent the degree of hearing impairment and physical sign data self-tested by the patient at each time interval. The physical sign data includes heart rate, respiratory rate, blood pressure, blood sugar, body temperature, and pain level.

[0076] The extraction logic of the relevant data of the patient's case includes:

[0077] Set the visit number, past medical history, current symptoms, and diagnosis code as the keywords of the natural language extractor. Identify the paragraph box of the text according to the keywords, extract the text after the keywords corresponding to the paragraph box, and set the keywords as the labels of the text to obtain the key data of the patient's case. Remove redundant words and punctuation marks in the key data. The redundant words include prepositions, conjunctions, articles, and pronouns to obtain the relevant data of the patient's case.

[0078] In specific implementation, by obtaining the patient number and matching the patient's case, integrate all medical information of the patient, including the visit number, past medical history, current symptoms, and diagnosis code, to provide comprehensive data support for subsequent data analysis and disease prediction. Detailed records of the patient's past medical history and current symptoms help doctors fully understand the patient's health status and provide more accurate medical services for the patient. By integrating the patient's visit records and self-test records, construct a continuous and complete patient health record to provide a more accurate data basis for disease prediction and analysis.

[0079] The step S200 includes the following sub-steps:

[0080] Step S201: Count the number of the patient's visit numbers, denoted as the number of visits;

[0081] Step S202: Set the first sequence according to the order of the visit numbers, and set the second sequence according to the past medical history and current symptoms. The number of data in the first sequence and the number of data in the second sequence are equal to the number of visits. Count the number of occurrences of each word in the second sequence, denoted as the first count. Set the third sequence according to the first count. Count the number of occurrences of the diagnosis codes, denoted as the second count. Set the fourth sequence according to the second count. Set the fifth sequence according to the time interval between visits, where the time interval between visits is expressed as the number of days from the current visit to the previous visit.

[0082] Step S202 further includes the following sub-steps:

[0083] Step S2021: Set the first sequence according to the order of the visit numbers. The elements included in the first sequence are denoted as the first elements, and the first sequence is expressed as X = {x1, x2,..., xT}.

[0084] Step S2022: Set the past medical history and current symptoms as the second sequence. The elements included in the second sequence are denoted as the second elements, and the second sequence is expressed as Y = {y1, y2,... yT}, where T represents the number of visits of the patient, and each second element represents the past medical history and current symptoms corresponding to the number of visits.

[0085] Step S2023: Count the number of occurrences of each word in the second sequence, denoted as the first count. Retain the words with a first count value of 1, and delete the words with a first count value not equal to 1. Set the retained words as the third sequence according to the order of the subscripts of the original second elements, and assign a third number to the third sequence. The third number is a natural number. Establish a first mapping relationship between the third label and the subscript of the original second element. The elements included in the third sequence are denoted as the third elements, and the third sequence is expressed as K = {k1, k2..., kM}, where M ranges from 1 to T, and M is a natural number.

[0086] Step S2024: Count the number of occurrences of the diagnosis codes, denoted as the second count. Retain the words with a second count value of 1, and delete the words with a second count value not equal to 1. Obtain the subscripts of the first elements corresponding to the retained words. Set the retained words as the fourth sequence according to the order of the corresponding subscripts of the first elements, and assign a fourth number to the fourth sequence. The fourth number is a natural number. Establish a second mapping relationship between the fourth label and the subscript of the original first element. The elements included in the fourth sequence are denoted as the fourth elements, and the fourth sequence is expressed as C = {c1, c2,... CN}, where C ranges from 1 to T, and C is a natural number.

[0087] Step S2025, set the visit time interval as the fifth sequence, and the fifth sequence is expressed as L = {l1, l2,... lU}, where U is distributed between 1 and T - 1, and U is a natural number.

[0088] In specific implementation, by counting the number of patient visit numbers, the visit frequency of patients can be understood, providing basic data for analyzing the health status and disease development trend of patients. Setting the first sequence according to the order of visit numbers, the past medical history and current symptoms as the second sequence, and the visit time interval as the fifth sequence helps to construct a comprehensive patient health record, providing structured data for disease prediction and analysis. By retaining the words with the first occurrence count of 1 and deleting the words with the first occurrence count not equal to 1, valuable effective data for disease prediction and analysis can be screened out, improving the accuracy and efficiency of the prediction model. Establishing the first mapping relationship between the third label and the subscript of the original second element, and the second mapping relationship between the fourth label and the subscript of the original first element helps to quickly locate and extract relevant data in subsequent analysis.

[0089] The step S300 includes the following sub-steps:

[0090] Step S301, obtain the subscript of the third element, obtain the subscript of the original second element according to the first mapping relationship, highlight the position corresponding to the third element in the second sequence, and set the first label, where the first label is the sudden current symptom;

[0091] Step S302, obtain the subscript of the fourth element, obtain the subscript of the original first element according to the second mapping relationship, highlight the position corresponding to the fourth element in the first sequence, and set the second label, where the second label is the sudden disease type;

[0092] Step S303, perform the first comparison between the position of the second label and the position of the fourth label to obtain the first comparison result. According to the first comparison result, delete the invalid data, delete the corresponding positions in the third element, the fourth element, and the fifth element, and update the value of the next position corresponding to the position in the fifth element;

[0093] The first comparison result is that the position of the second label is the same as the position of the fourth label, and the position of the second label is different from the position of the fourth label;

[0094] When the first comparison result is that the position of the second label is different from the position of the fourth label, delete the corresponding positions in the third element, the fourth element, and the fifth element;

[0095] When the first comparison result shows that the positions of the second tag and the fourth tag are the same, update the value of the next position corresponding to the corresponding position in the fifth element. The update logic of the value of the next position includes:

[0096] Obtain the fifth element from the corresponding position in the fifth element to the next position corresponding to the corresponding position in the fifth element, calculate the sum value of the fifth element, denoted as the first sum value. Obtain the fifth element from the previous position corresponding to the corresponding position in the fifth element to the fifth element corresponding to the corresponding position, calculate the sum value of the fifth element, denoted as the second sum value. Calculate the sum value of the first sum value and the second sum value, denoted as the third sum value. Replace the first sum value with the third sum value to obtain the new value of the next position corresponding to the corresponding position in the fifth element;

[0097] Step S304: Obtain the data corresponding to the updated second tag and the data corresponding to the third tag, denoted as the first data and the second data.

[0098] In specific implementation, by highlighting the positions of the third element and the fourth element in their respective sequences and setting corresponding tags, the sudden current symptoms and disease types can be visually identified and marked, facilitating subsequent data analysis and processing. Delete the corresponding positions in the third element, the fourth element, and the fifth element related to invalid data, clean the data set, and remove the noise data that may affect the performance of the prediction model. When the positions of the second tag and the fourth tag are the same, by updating the value of the fifth element, the symptoms and disease types can be accurately associated, providing more accurate time-correlated data for disease prediction.

[0099] The step S400 includes the following sub-steps:

[0100] Step S401: Obtain the types of sudden ear diseases and their corresponding symptoms. The types of sudden ear diseases include mild sudden ear disease, moderate sudden ear disease, severe sudden ear disease, and extremely severe sudden ear disease;

[0101] The symptoms of mild sudden ear disease are unilateral hearing loss. The symptoms of moderate sudden ear disease include unilateral or bilateral hearing loss, tinnitus, accompanied by dizziness, and sleep disorders. The symptoms of severe sudden ear disease are bilateral hearing loss, tinnitus, dizziness, ear fullness, and occasionally accompanied by sleep disorders. The symptoms of extremely severe sudden ear disease are deafness, dizziness, ear fullness, anxiety, sleep disorders, nausea, vomiting, and cold sweats;

[0102] Step S402: Conduct a second comparison between the types of sudden ear diseases and the types of sudden diseases to obtain a second comparison result. The second comparison result includes that the type of sudden disease belongs to the type of sudden ear disease and that the type of sudden disease does not belong to the type of sudden ear disease;

[0103] Step S403, when the second comparison result indicates that the type of the sudden disease belongs to the type of sudden ear disease, insert the disease type into the fifth sequence according to the corresponding number of visits;

[0104] When the second comparison result indicates that the type of the sudden disease does not belong to the type of sudden ear disease, retrieve the expert platform, obtain the fifth element adjacent to the current symptoms of the sudden disease in the fifth sequence, obtain the self-test records of the patients corresponding to the adjacent fifth element, input the current symptoms of the sudden disease and the self-test records of the patients corresponding to the fifth element into the expert platform, and obtain the corresponding disease type, denoted as the first disease type. The first disease type includes car accident injuries, falls from heights, burns, fractures from falls, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergies, food allergies, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding, and sudden ear diseases. The adjacent self-test records are represented as the previous element and the next element of the current symptoms of the sudden disease in the fifth sequence;

[0105] Step S403, update the type of the sudden disease according to the first disease type, and insert the first disease type into the fifth sequence according to the corresponding number of visits.

[0106] In specific implementation, by comparing the type of sudden ear disease with the type of the sudden disease, it can quickly identify whether the patient has sudden ear disease, improving the efficiency and accuracy of diagnosis. When the disease type does not belong to sudden ear disease, by retrieving the expert platform and the self-test records of the patients, more data resources are integrated, improving the comprehensiveness and depth of diagnosis. Utilizing the knowledge and experience of the expert platform and combining the self-test records of the patients can assist doctors in making more accurate diagnoses, especially when facing complex or atypical cases. Inserting the disease type into the fifth sequence according to the corresponding number of visits takes into account the patient's visit frequency, providing a more accurate time basis for disease prediction and health management.

[0107] The step S500 includes the following sub-steps:

[0108] Step S501, obtain the fifth sequence after inserting the disease type and the first disease type, denoted as the sixth sequence. Both the disease type and the first disease type are represented as the type of sudden ear disease, and represent the disease type and the first disease type as the third data;

[0109] Step S502, classify the same types of sudden ear diseases in the third data in chronological order to obtain the first classification, and assign the first classification a first number, where the first number is a natural number;

[0110] Step S503, calculate the time interval between the sub-data of the first classification, denoted as the first time period. The calculation logic for the time interval between the sub-data of the first classification includes:

[0111] Obtain the fifth element between the sub-data, calculate the sum value of the fifth element, denote it as the fourth sum value, and set the fourth sum value as the time interval between the sub-data;

[0112] Step S503: Fit the equation of the sub-data and time according to the least squares method, denote it as the sub-equation, and perform a second numbering on the sub-equation. The second numbering is a natural number and corresponds to the first numbering.

[0113] In specific implementation, by obtaining the fifth sequence after inserting the disease types and the first disease type, form the sixth sequence, integrate all the data related to sudden deafness, provide a complete data basis for subsequent time series analysis, analyze the frequency and law of the occurrence of ear diseases by calculating the time interval between the sub-data of the first classification, provide a time basis for disease prediction, use the least squares method to fit the equation of the sub-data and time, and establish a mathematical model to describe the time trend of the occurrence of ear diseases, so as to improve the accuracy of prediction.

[0114] The step S600 includes the following sub-steps:

[0115] Step S601: Obtain the current time point, and according to the sixth sequence, obtain the adjacent third data of the current time point. The adjacent third data represents the third data with the closest time and its first numbering.

[0116] Step S602: Calculate the time interval between the adjacent third data and the current time point, and predict the first numbering of the next third data and the time interval between the next third data and the current time point according to the first numbering and the time interval. The prediction logic of the first numbering of the next third data includes:

[0117] Obtain the first numbering of the adjacent third data, and according to the numbering order, obtain the first numbering of the next third data;

[0118] The calculation logic of the time interval between the next third data and the current time point is:

[0119] Obtain the current time point, obtain the first numbering of the next third data, retrieve the sub-equation corresponding to the first numbering of the next third data, calculate the time point of the next third data, denote it as the future time point, calculate the difference between the future time point and the current time point, denote it as the first difference, and set the first difference as the time interval between the next third data and the current time point.

[0120] In specific implementation, by obtaining the current time point and adjacent third data, the occurrence of ear diseases is analyzed and predicted in real time to improve the timeliness of prediction. By calculating the time interval between the adjacent third data and the current time point, the occurrence time of the next third data is predicted more accurately. By predicting the time point of the next third data, the continuity of the time series is maintained, providing support for the long-term trend analysis of diseases.

[0121] The present invention realizes individualized diagnosis and treatment plans by analyzing and mining medical big data, improving patient satisfaction. Medical health big data analysis helps doctors predict possible diseases of patients, which is of great significance for the early prevention and control of diseases. Using time series analysis methods to capture the trend of the development of patients' diseases, and then predicting possible future diseases, which is particularly important for the prediction of ear diseases. By fitting the equation of sub-data and time by the least squares method, the first number of the next third data and the time interval between the next third data and the current time point are predicted, so as to realize the prediction of the occurrence time of ear diseases.

[0122] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 specified in one block or multiple blocks.

[0123] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An ear disease prediction method based on big data, characterized in that: The following steps are involved: Step S100, obtaining the patient number, matching the patient case according to the patient number, extracting the relevant data of the patient case, the relevant data including the consultation number, past medical history, current symptoms and diagnosis code, and obtaining the patient's self-test record; Step S200, counting the number of the patient's consultation numbers, recorded as the number of consultations, setting a first sequence according to the order of the consultation numbers, setting a second sequence according to the past medical history and current symptoms, counting the number of times each word in the second sequence appears, recorded as the first number, setting a fourth sequence according to the diagnosis code, and setting a fifth sequence according to the time interval between consultations; Step S300, setting a first label at the position of the third sequence in the second sequence, setting a second label at the position of the fourth element in the first sequence, performing a first comparison between the position of the second label and the position of the fourth label, deleting invalid data according to the first comparison result, deleting the corresponding position in the third element, the corresponding position in the fourth element, and the corresponding position in the fifth element, and updating the value of the next position of the corresponding position in the fifth element, obtaining the updated data corresponding to the second label and the data corresponding to the third label, and recording them as the first data and the second data; Step S400, obtaining the type of sudden ear disease and its corresponding symptoms, performing a second comparison between the type of sudden ear disease and the type of sudden disease, and inserting the type of disease into the fifth sequence according to the corresponding number of visits according to the second comparison result, or calling the expert platform to obtain the fifth element adjacent to the sudden current symptom in the fifth sequence, obtaining the patient self-test record corresponding to the adjacent fifth element, inputting the sudden current symptom and the patient self-test record corresponding to the fifth element into the expert platform, obtaining the corresponding disease type, recording it as the first disease type, updating the sudden disease type according to the first disease type, and inserting the first disease type into the fifth sequence according to the corresponding number of visits; Step S500, obtaining a fifth sequence after inserting the disease type and the first disease type, recorded as the sixth sequence, classifying the same sudden ear disease types in the third data in chronological order to obtain a first classification, assigning a first number to the first sub-classification, calculating the time interval between the sub-data of the first classification, recorded as the first time period, and fitting an equation between the sub-data and time according to the least squares method, recorded as a sub-equation; Step S600, obtaining the current time point, calculating the time interval between adjacent third data and the current time point, predicting the first number of the next third data and the time interval between the next third data and the current time point according to the first number and the time interval.

2. The ear disease prediction method based on big data according to claim 1, characterized in that: Said Step S100 includes the following sub-steps: Step S101, obtaining a patient number, retrieving a case database, inputting the patient number into the case database, and matching a patient case corresponding to the patient number; Step S102, extracting relevant data of the patient's case, the relevant data including the consultation number, past medical history, current symptoms and diagnosis code, the diagnosis code is the type of sudden illness determined by the doctor, including traffic accident injury, fall injury, burn, fracture, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergy, food allergy, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding and sudden ear disease, there is only one code and it corresponds to the consultation number; Different visit numbers indicate that the same patient visited the hospital at different times; Step S103, obtaining the patient's self-test record, wherein the patient's self-test record represents the hearing loss degree and physical sign data of the patient's self-test at each time interval, wherein the physical sign data includes heart rate, respiratory rate, blood pressure, blood sugar, body temperature and pain degree.

3. The ear disease prediction method based on big data according to claim 2, characterized in that: The logic for extracting relevant data of the patient case includes: The consultation number, past medical history, current symptoms and diagnosis code are set as keywords of the natural language extractor, the paragraph frame of the text is identified according to the keywords, the text after the keyword corresponding to the paragraph frame is extracted, and the keyword is set as the label of the text to obtain the key data of the patient case, and the redundant words and punctuation marks in the key data are removed, and the redundant words include prepositions, conjunctions, articles and pronouns to obtain the relevant data of the patient case.

4. The ear disease prediction method based on big data according to claim 1, characterized in that: The step S200 includes the following sub-steps: Step S201, counting the number of patients' consultation numbers, recorded as the number of consultations; Step S202, setting a first sequence according to the order of the consultation numbers, setting a second sequence according to the past medical history and current symptoms, the number of data in the first sequence and the number of data in the second sequence are equal to the number of consultations, counting the number of times each word in the second sequence appears, recorded as the first number, setting a third sequence according to the first number, counting the number of times the diagnosis code appears, recorded as the second number, setting a fourth sequence according to the second number, and setting a fifth sequence according to the time interval between consultations, the time interval between consultations being expressed as the number of days from the current visit to the last visit.

5. The ear disease prediction method based on big data according to claim 4, characterized in that: The step S202 further includes the following sub-steps: Step S2021: Set the first sequence in the order of the consultation numbers. The elements included in the first sequence are recorded as the first elements. The first sequence is represented by X = {x1, x2, ..., x T }; Step S2022: Set the past medical history and current symptoms as a second sequence. The elements included in the second sequence are recorded as second elements. The second sequence is represented by Y = {y1, y2, ...y T }, where T represents the number of visits of the patient, and each second element represents the medical history and current symptoms corresponding to the number of visits; Step S2023, count the number of times each word in the second sequence appears, record it as the first number, retain the words whose first number is 1, delete the words whose first number is not 1, set the retained words to the third sequence according to the order of the subscripts of the original second elements, and perform a third numbering on the third sequence, the third numbering being a natural number, establishing a first mapping relationship between the third number and the subscript of the original second element, the elements included in the third sequence are recorded as the third elements, and the third sequence is represented by K={k1, k2..., k M }, where M is distributed between 1 and T, and M is a natural number; Step S2024, count the number of times the diagnostic code appears, record it as the second number, retain the words whose second number is 1, delete the words whose second number is not 1, obtain the subscripts of the first elements corresponding to the retained words, set the retained words as a fourth sequence according to the order of the subscripts of the corresponding first elements, and perform a fourth numbering on the fourth sequence, the fourth numbering being a natural number, establish a second mapping relationship between the fourth number and the subscript of the original first element, the elements included in the fourth sequence are recorded as fourth elements, and the fourth sequence is represented by C={c1, c2, ...C N }, where C is distributed between 1 and T, and C is a natural number; Step S2025, setting the time interval of the medical consultation to a fifth sequence, wherein the fifth sequence is represented by L = {l1, l2, ...l U }, where U is distributed between 1 and T-1, and U is a natural number.

6. The ear disease prediction method based on big data according to claim 1, characterized in that: The step S300 includes the following sub-steps: Step S301, obtaining the corner mark of the third element, obtaining the original corner mark of the second element according to the first mapping relationship, marking the corresponding position of the third element in the second sequence in red, and setting a first label, wherein the first label is the current symptom of the sudden event; Step S302, obtaining the fourth element corner mark, obtaining the original first element corner mark according to the second mapping relationship, marking the fourth element in the corresponding position in the first sequence in red, and setting a second label, wherein the second label is the type of sudden disease; Step S303, performing a first comparison between the position of the second tag and the position of the fourth tag to obtain a first comparison result, deleting invalid data according to the first comparison result, deleting the corresponding position in the third element, the corresponding position in the fourth element, and the corresponding position in the fifth element, and updating the value of the next position of the corresponding position in the fifth element; The first comparison result is that the position of the second tag is the same as the position of the fourth tag, and the position of the second tag is different from the position of the fourth tag; When the first comparison result is that the position of the second tag is different from the position of the fourth tag, deleting the corresponding position in the third element, the corresponding position in the fourth element, and the corresponding position in the fifth element; When the first comparison result is that the position of the second tag is the same as the position of the fourth tag, and the value of the next position of the corresponding position in the fifth element is updated, the update logic of the value of the next position includes: Get the fifth element from the corresponding position in the fifth element to the next position of the corresponding position in the fifth element, calculate the sum of the fifth elements, record it as the first sum, get the fifth element from the previous position of the corresponding position in the fifth element to the corresponding position in the fifth element, calculate the sum of the fifth elements, record it as the second sum, calculate the sum of the first sum and the second sum, record it as the third sum, replace the first sum according to the third sum, and obtain the value of the next position of the corresponding position in the new fifth element; Step S304, obtaining updated data corresponding to the second tag and data corresponding to the third tag, which are recorded as first data and second data.

7. The ear disease prediction method based on big data according to claim 1, characterized in that: The step S400 includes the following sub-steps: Step S401, obtaining the types of sudden ear diseases and their corresponding symptoms, wherein the types of sudden ear diseases include mild sudden ear diseases, moderate sudden ear diseases, severe sudden ear diseases and extremely severe sudden ear diseases; The symptoms of mild sudden ear disease are unilateral hearing loss, the symptoms of moderate sudden ear disease include unilateral or bilateral hearing loss, tinnitus, accompanied by dizziness, and sleep disorders, the symptoms of severe sudden ear disease include bilateral hearing loss, tinnitus, dizziness, ear fullness, and occasionally accompanied by sleep disorders, and the symptoms of extremely severe sudden ear disease include deafness, dizziness, ear fullness, anxiety, sleep disorders, nausea, vomiting, and cold sweats; Step S402, performing a second comparison between the sudden ear disease category and the sudden disease category to obtain a second comparison result, wherein the second comparison result includes that the sudden disease category belongs to the sudden ear disease category, and the sudden disease category does not belong to the sudden ear disease category; Step S403, when the second comparison result shows that the sudden disease type belongs to the sudden ear disease type, the disease type is inserted into the fifth sequence according to the corresponding number of visits; When the second comparison result is that the sudden disease type does not belong to the sudden ear disease type, call the expert platform to obtain the fifth element adjacent to the sudden current symptom in the fifth sequence, obtain the patient self-test record corresponding to the adjacent fifth element, input the sudden current symptom and the patient self-test record corresponding to the fifth element into the expert platform, obtain the corresponding disease type, and record it as the first disease type. The first disease type includes traffic accident injury, fall injury, burn, fall fracture, arrhythmia, angina pectoris, myocardial infarction, sudden cardiac death, cerebral hemorrhage, cerebral infarction, drug allergy, food allergy, acute appendicitis, acute gastrointestinal perforation, gastrointestinal bleeding and sudden ear disease. The adjacent self-test records are represented by the previous element and the next element of the sudden current symptom in the fifth sequence; Step S403, updating the sudden disease type according to the first disease type, and inserting the first disease type in the fifth sequence according to the corresponding number of visits.

8. The ear disease prediction method based on big data according to claim 1, characterized in that: The step S500 includes the following sub-steps: Step S501, obtaining a fifth sequence after inserting the disease type and the first disease type, recorded as the sixth sequence, the disease type and the first disease type are both represented as sudden ear disease types, and the disease type and the first disease type are represented as third data; Step S502, classifying the same types of sudden ear diseases in the third data in chronological order to obtain a first category, and assigning a first number to the first category, wherein the first number is a natural number; Step S503, calculating the time interval between the sub-data of the first category, recorded as the first time period, the calculation logic of the time interval between the sub-data of the first category includes: Obtain the fifth element between the sub-data, calculate the sum of the fifth element, record it as the fourth sum, and set the fourth sum as the time interval between the sub-data; Step S503, fitting the equation between the sub-data and the time according to the least square method, recording it as a sub-equation, and assigning a second number to the sub-equation, where the second number is a natural number and the second number corresponds to the first number.

9. The ear disease prediction method based on big data according to claim 1, characterized in that: The step S600 includes the following sub-steps: Step S601, obtaining the current time point, and obtaining the adjacent third data of the current time point according to the sixth sequence, wherein the adjacent third data represents the third data closest in time and its first number; Step S602, calculating the time interval between the adjacent third data and the current time point, predicting the first number of the next third data and the time interval between the next third data and the current time point according to the first number and the time interval, wherein the prediction logic of the first number of the next third data includes: Obtain the first number of the adjacent third data, and obtain the first number of the next third data according to the sequence of numbers; The calculation logic of the time interval between the next third data and the current time point is: Get the current time point, get the first number of the next third data, call the sub-equation with the first number of the next third data, calculate the time point of the next third data, record it as the future time point, calculate the difference between the future time point and the current time point, record it as the first difference, and set the first difference as the time interval between the next third data and the current time point.

10. An ear disease prediction system based on big data, the system being used to execute the ear disease prediction method based on big data according to claim 1, characterized in that: It includes acquisition module, analysis module and calculation module; The acquisition module is used to obtain the patient number, match the patient case according to the patient number, extract the relevant data of the patient case, the relevant data includes the consultation number, past medical history, current symptoms and diagnosis code, and obtain the patient's self-test record; The analysis module is used to count the number of patient consultation numbers, recorded as the number of consultations, set a first sequence according to the order of the consultation numbers, set a second sequence according to the past medical history and current symptoms, count the number of times each word in the second sequence appears, recorded as the first number, set a fourth sequence according to the diagnosis code, set a fifth sequence according to the time interval between consultations, set a first label for the position of the third sequence in the second sequence, set a second label for the position of the fourth element in the first sequence, perform a first comparison between the position of the second label and the position of the fourth label, delete invalid data according to the first comparison result, delete the corresponding position in the third element, the corresponding position in the fourth element, and the corresponding position in the fifth element, and update the value of the next position of the corresponding position in the fifth element to obtain the updated first label. The data corresponding to the second label and the data corresponding to the third label are recorded as the first data and the second data, the sudden ear disease type and the corresponding symptoms are obtained, the sudden ear disease type is compared with the sudden disease type for the second time, and the disease type is inserted into the fifth sequence according to the corresponding number of visits according to the second comparison result, or the expert platform is called to obtain the fifth element adjacent to the sudden current symptom in the fifth sequence, and the patient self-test record corresponding to the adjacent fifth element is obtained, and the sudden current symptom and the patient self-test record corresponding to the fifth element are input into the expert platform to obtain the corresponding disease type, which is recorded as the first disease type, and the sudden disease type is updated according to the first disease type, and the first disease type is inserted into the fifth sequence according to the corresponding number of visits; The calculation module obtains the fifth sequence after inserting the disease type and the first disease type, recorded as the sixth sequence, classifies the same sudden ear disease types in the third data in chronological order to obtain the first classification, gives the first sub-classification a first number, calculates the time interval between the sub-data of the first classification, recorded as the first time period, fits the equation of the sub-data and time according to the least squares method, recorded as the sub-equation, obtains the current time point, calculates the time interval between the adjacent third data and the current time point, and predicts the first number of the next third data and the time interval between the next third data and the current time point according to the first number and the time interval.

Images