A method and system for outpatient auxiliary examination analysis based on big data

Abstract

The present invention provides an outpatient auxiliary examination and analysis method and system based on big data, which belongs to the field of information analysis and solves the problem of slow outpatient consultation speed. The specific steps are as follows: Step S1: Obtain historical medical records and the total number of residents, and calculate the comprehensive incidence coefficient group of each disease; Step S2: Obtain physician information, determine the pathological classification of different diseases and the inducing relationship between different diseases in combination with historical medical records, and build a disease diagnosis model; Step S3: Obtain the patient's judgment parameters; Input the judgment parameters into the disease diagnosis model, read the determined disease, pending treatment time and determined inducing disease output by the model as an initial diagnosis report; Step S4: Summarize the initial diagnosis report and provide feedback; The present invention performs data mining and data induction on the historical data of the target hospital, determines the incidence and inducing relationship of different diseases, and then uses machine learning to judge the patient's disease, thereby improving the efficiency and speed of outpatient consultation.

Description

Technical Field

[0001] The present invention discloses an outpatient auxiliary examination analysis method and system based on big data, and relates to the field of information analysis. Background Art

[0002] The existing analysis methods or systems for outpatient auxiliary examinations have the following deficiencies:

[0003] Data quality varies: During outpatient auxiliary examinations, some examination data may be inaccurate or missing due to human error, equipment failure, and other reasons. At the same time, most systems related to outpatient auxiliary examinations often directly import and analyze medical record data from local provincial large hospitals when mining historical medical records. These systems lack regional characteristics and regional specificity, and have poor medical record analysis capabilities at the municipal or county level.

[0004] High technical complexity: Building a big data outpatient auxiliary examination and analysis system involves multiple technologies such as big data processing, data mining, and artificial intelligence. These technologies are not only highly complex, but also rapidly updated, which places extremely high demands on the hospital's technical team. If the hospital cannot keep up with the pace of technological development, it will be difficult to give full play to the advantages of the system.

[0005] Reliance on external data sources: Existing analysis systems often require external data to enrich and improve patients' medical information; however, the quality and reliability of these external data are often difficult to guarantee, and there are differences in data formats and standards between different data sources, which increases the difficulty of data integration and analysis; in addition, if the external data is interrupted or changed, it will have an adverse impact on the normal operation of the system. Summary of the invention

[0006] In view of the shortcomings of the prior art, the purpose of the present invention is to provide an outpatient auxiliary examination and analysis method and system based on big data, aiming to solve the problem of slow outpatient service speed.

[0007] In order to achieve the above object, the present invention is implemented through the following technical solutions: A method for outpatient auxiliary examination and analysis based on big data includes:

[0008] Step S1: Obtain all historical medical records of the target hospital and the total number of residents in the target area; extract the names of the symptoms in the historical medical records, summarize all historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents;

[0009] Step S2: Obtain all physician information of the target hospital; obtain the symptom manifestations of the corresponding disease in the disease database according to the historical medical records, and determine the pathological classification of different diseases and the induction relationship between different diseases in combination with the onset symptoms of all diseases in the historical medical records; construct a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease;

[0010] Step S3: Obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model as the initial diagnosis report;

[0011] Step S4: Summarize the initial diagnosis report and provide feedback.

[0012] Furthermore, the specific steps of step S1 are as follows:

[0013] Step S11: Obtain all historical medical records of the target hospital; extract all disease names from the historical medical records, count the number of all disease names, and record it as dn; obtain the total number of residents in the target area, and record it as pn;

[0014] Step S12: Summarize all historical medical records of patients with the first disease; calculate the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease, and obtain a comprehensive incidence rate coefficient group of the first disease;

[0015] Step S13: Repeat the same steps of calculating the comprehensive morbidity coefficient group of the first disease, calculate the comprehensive morbidity coefficient groups of the second to dnth diseases, and enter step S2.

[0016] Furthermore, the specific steps of step S12 are as follows:

[0017] Step S121: Record the patient suffering from the first disease as the target patient;

[0018] Count the number of historical medical records of all target patients, recorded as fb;

[0019] Calculate the overall incidence of the first disease, denoted as Ir; the calculation formula for Ir is as follows:

[0020] Ir=fb / pn;wherein, Ir≤1;

[0021] Step S122: Count the number of historical medical records of male target patients, recorded as mb; count the number of historical medical records of female target patients, recorded as wb;

[0022] Calculate the male incidence of the first disease, recorded as Mr; the calculation formula for Mr is as follows:

[0023] Mr=mb / pn;wherein, Mr≤1;

[0024] Calculate the female incidence of the first disease, denoted as Wr; the calculation formula for Wr is as follows:

[0025] Wr=wb / pn;Wr≤1;

[0026] Among them, the relationship between mb, fb and fb satisfies: mb+fb=fb; the relationship between Mr, Wr and Ir satisfies: Mr+Wr=Ir;

[0027] Mr and Wr were used as the sex incidence of the first disease;

[0028] Step S123: Count the number of historical medical records of infants and young children as target patients, recorded as fpn1;

[0029] The number of historical medical records of target patients who were adolescents was counted, recorded as fpn2;

[0030] The number of historical medical records of target patients who were adults was counted, recorded as fpn3;

[0031] The number of historical medical records of the target patient being middle-aged was counted, recorded as fpn4;

[0032] The number of historical medical records of the target patient being elderly was counted, recorded as fpn5;

[0033] Among them, fpn1~fpn5 are all greater than or equal to 0; calculate the sum of fpn1~fpn5, denoted as fpn; the relationship between fpn and fb satisfies: fpn≥fb;

[0034] Step S124: The infant morbidity rate of the first disease is recorded as Fbb1; the calculation formula of Fbb1 is: Fbb1 = (Tr*fpn1) / fb;

[0035] The adolescent incidence of the first disease is recorded as Fbb2; the calculation formula of Fbb2 is: Fbb2 = (Tr*fpn2) / fb;

[0036] The adult incidence of the first disease is recorded as Fbb3; the calculation formula of Fbb3 is: Fbb3 = (Tr*fpn3) / fb;

[0037] The middle-aged incidence of the first disease is recorded as Fbb4; the calculation formula of Fbb4 is: Fbb4 = (Tr*fpn4) / fb;

[0038] The elderly incidence of the first disease is recorded as Fbb5; the calculation formula of Fbb5 is: Fbb5 = (Tr*fpn5) / fb;

[0039] Fbb1 to Fbb5 were used as the age incidence of the first disease;

[0040] Step S125: Calculate the age group coefficients a1-a5 of the first disease;

[0041] The calculation formula of a1 is: a1 = fpn1 / fpn; the calculation formula of a2 is: a2 = fpn2 / fpn; the calculation formula of a3 is: a3 = fpn3 / fpn; the calculation formula of a4 is: a4 = fpn4 / fpn; the calculation formula of a5 is: a5 = fpn5 / fpn;

[0042] The sick time of all infants and young children is counted and recorded as atⅠ1~atⅠ fpn1 ; Recovery time, recorded as atⅡ1~atⅡ fpn1 ;

[0043] The duration of illness of all adolescents is counted and recorded as btⅠ1~btⅠ fpn2 ; Recovery time, recorded as btⅡ1~btⅡ fpn2 ;

[0044] The duration of illness of all adults was counted and recorded as ctⅠ1~ctⅠ fpn3 ; Recovery time, recorded as ctⅡ1~ctⅡ fpn3 ;

[0045] The total duration of illness in middle-aged people is recorded as dtⅠ1~dtⅠ fpn4 ; Recovery time, recorded as dtⅡ1~dtⅡ fpn4 ;

[0046] The duration of illness of all elderly people is counted and recorded as etⅠ1~etⅠ fpn5 ; Recovery time, recorded as etⅡ1~etⅡ fpn5 ;

[0047] Step S126: Calculate the comprehensive treatment time of the first disease, recorded as tt; the calculation formula of tt is as follows:

[0048] ;

[0049] Step S127: Summarize steps S121 to S126 to obtain the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease as the comprehensive incidence rate coefficient group of the first disease.

[0050] Furthermore, the specific steps of step S2 are as follows:

[0051] Step S21: extracting the medical record information of each patient from the historical medical records;

[0052] The patient's medical record information includes: patient name, disease name, symptoms, illness duration and attending physician;

[0053] Step S22: Obtain the names of doctors in the target hospital and the corresponding departments of each doctor;

[0054] Link to the disease database to obtain the symptoms corresponding to the disease names in the historical medical records;

[0055] Step S23: Based on the data obtained in steps S21 to S22, construct a patient table, a disease table, and a physician table in the database; the corresponding SQL statement is as follows:

[0056] Patient table:

[0057] Create table Patients(

[0058] Patient name String, disease name String, disease status String,

[0059] Illness time String, attending physician String,

[0060] primary key (patient name),

[0061] foreign key (disease name) referances disease (disease name)

[0062] Disease table:

[0063] Create table Disease(

[0064] Symptom name String, Symptom String,

[0065] primary key (disease name)

[0066] Physician table:

[0067] Create table Physician(

[0068] Doctor's name String, Department String,

[0069] primary key (Doctor's name)

[0070] Step S24: judging the pathological classification of different symptoms and the inducing relationship between different symptoms according to the patient table, the disease table and the physician;

[0071] Step S25: constructing a disease diagnosis model based on the random forest model;

[0072] Symptom diagnosis model: used to determine the patient's diagnosed symptoms, medical department, pending treatment time and diagnosed induced diseases based on the patient's symptom characteristics.

[0073] Furthermore, the specific steps of step S24 are as follows:

[0074] Step S241: Read the number dn of all disease names;

[0075] Step S242: Based on the patient table and the physician table, the pathological classification of different diseases is output using SQL statement 2. SQL statement 1 is as follows:

[0076] select patient.disease name, physician.department count(select patient.attending physician, physician.physician name from patient inner join patient.attending physician=physician.physician name) from physician group by patient.disease name;

[0077] The output of SQL statement 1 is the pathological classification of the 1st to pnth diseases;

[0078] Step S243: determining the inducing relationship between the 1st to pnth diseases;

[0079] Step S2431: Obtain the first illness time corresponding to the 1st to pnth diseases, recorded as ft1, ft2, ft3~ft pn ;

[0080] Step S2432: Obtain all the symptoms corresponding to the 1st to pnth diseases, and obtain sets Bm1, Bm2, Bm3 to Bm4. pn ;

[0081] Step S2433: Obtain all symptom manifestations corresponding to the 1st to pnth symptom, and obtain sets Bi1, Bi2, Bi3 to Bi4. pn ;

[0082] Step S2434: determining the inducing relationship of the first symptom among the first to pn th symptom;

[0083] Step S2435: Repeat the same steps of determining the induction relationship of the first disease to determine the induction relationship of the second to pnth diseases.

[0084] Furthermore, the specific steps of step S2434 are as follows:

[0085] Step S24341: determining whether there is an inducing relationship between the first symptom and the second symptom;

[0086] The first disease is recorded as disease A, and the second disease is recorded as disease B;

[0087] In the comprehensive incidence coefficient group of the first disease, read the overall incidence of disease A, recorded as P (A); in the comprehensive incidence coefficient group of the second disease, read disease B;

[0088] Step S24342: Count the number of people who suffer from both disease A and disease B, recorded as pn AB ;

[0089] Calculate the probability of having both disease A and disease B, recorded as P(AB); the calculation formula of P(AB) is: P(AB) = pn AB / pn;

[0090] Step S24343: Calculate the probability of first contracting disease A and then contracting disease B, recorded as P(B|A); the calculation formula of P(B|A) is: P(B|A)=P(AB) / P(A);

[0091] Calculate the probability of contracting disease B before contracting disease A, recorded as P(A|B); the calculation formula of P(A|B): P(A|B) = P(AB) / P(B);

[0092] Step S24344: defining relationship equations 1 to 2 that the first symptom and the second symptom have an inducing relationship;

[0093] Relationship 1: ft1≤ftB; where ftB represents ft2;

[0094] Relation 2: ; Where BmB and BiB represent the set Bm2 and the set Bi2 respectively

[0095] Relation 3: P(B|A)+P(A|B)<1;

[0096] Step S24345: Determine whether all of equations 1 to 3 are satisfied;

[0097] If all of these conditions are met, it means that there is an induction relationship between the first symptom and the second symptom, that is, the first symptom induces the second symptom;

[0098] If none of them are met or partially met, it means that there is no inductive relationship between the first and second symptoms;

[0099] Step S24346: Expand the representation range of disease B to the third to pnth diseases, and expand the representation range of ftB to the set ft3 to the set ft pn , the range of BmB and BiB is extended to the set Bm3 to the set Bm pn and set Bi3~set Bi pn ; Repeat steps S24341 to S24345 to determine whether there is an inducing relationship between the first symptom and the third to pnth symptoms, and summarize them.

[0100] Furthermore, the specific steps of step S25 are as follows:

[0101] Step S251: define the processing logic of the decision tree; assume that the model input is X, where X represents the symptom characteristics of the patient;

[0102] Step S252: Let X and set Bi i The matching degree is mat(X_i);

[0103] The logical expression of mat(X_i): mat(X_i) = Word2Vec(X, Bi i ), where i ranges from 1 to pn, Bi i represents the set of symptom manifestations corresponding to the ith symptom; Word2Vec represents the Word2Vec model;

[0104] Use logical expressions to calculate the relationship between X and the set Bi1 to the set Bi pn The matching degree is obtained by mat(X_1)~mat(X_pn);

[0105] Step S253: Obtain a judgment threshold, denoted as β; define relation 4: mat(X_i)>β;

[0106] Substitute mat(X_1)~mat(X_pn) into equation 4, and summarize the right index of mat(X_i) that satisfies the matching degree of equation 4 as the search sequence number; read the corresponding disease according to the search sequence number, record it as the pending disease, and count the number of pending diseases, record it as uni;

[0107] Step S254: determine the patient's gender and age, and read the gender incidence rate and age incidence rate corresponding to the pending disease;

[0108] Definition selection formula 5: g(j) = mat(X_j)*(fpb(j)*Pr(j)) 1 / 2 ; Wherein, the value range of j is: 1~uni; g(j) represents the selection coefficient of the jth undetermined disease, mat(X_j) represents the matching degree of the disease manifestation set corresponding to X and the jth undetermined disease, fpb(j) represents the gender incidence rate of the jth undetermined disease (corresponding to the patient's gender), and Pr(j) represents the age incidence rate of the jth undetermined disease (corresponding to the patient's age);

[0109] According to the selection formula 5, the selection coefficients g(1) to g(uni) of the undetermined diseases from No. 1 to No. uni are calculated;

[0110] Step S255: Compare the sizes of g(1) to g(uni), read the disease with the largest selection coefficient as the determined disease; call SQL statement 1 according to the determined disease to obtain the pathological classification of the determined disease, that is, the medical department; read the comprehensive treatment time corresponding to the determined disease as the pending treatment time;

[0111] Step S256: read all the induced diseases corresponding to the determined disease, and record them as pending induced diseases; count the number of pending induced diseases, and record them as unf;

[0112] Step S257: Read the gender incidence rate and age incidence rate corresponding to the pending induced disease according to the patient's gender and age;

[0113] Definition selection formula 6: g`(q) = (fpb`(q)*Pr`(q)) 1 / 2 ; Wherein, the value range of q is: 1~unf; g`(q) represents the alternative coefficient of the qth undetermined induced disease, fpb`(q) represents the gender incidence rate of the qth undetermined induced disease, and Pr`(q) represents the age incidence rate of the qth undetermined induced disease;

[0114] According to the selection formula 6, the candidate coefficients g`(1)~g`(unf) of the first to unf undetermined diseases are calculated;

[0115] Step S258: Compare the sizes of g`(1) to g`(unf), and read the induced disease with the largest candidate coefficient as the determined induced disease;

[0116] Step S259: Output the determined symptom, medical department, pending treatment time and determined induced disease corresponding to the patient's symptom feature X.

[0117] An outpatient auxiliary examination and analysis system based on big data includes:

[0118] Medical record acquisition module: used to obtain all historical medical records of the target hospital and the total number of residents in the target area; extract the names of the symptoms in the historical medical records, summarize all the historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents;

[0119] Model building module: used to obtain all the physician information of the target hospital; obtain the symptom manifestations of the corresponding disease in the disease database according to the historical medical records, and judge the pathological classification of different diseases and the induction relationship between different diseases by combining the symptoms of all diseases in the historical medical records; build a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease;

[0120] Patient analysis module: used to obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model (symptom diagnosis model) as the initial diagnosis report;

[0121] Physician interaction module: used to summarize the initial diagnosis report and provide feedback to the physician.

[0122] Compared with the prior art, the present invention has the following beneficial effects:

[0123] Improve diagnostic accuracy and efficiency: The present invention provides doctors with comprehensive and accurate diagnostic basis by deeply mining the hidden rules in medical data; using machine learning artificial intelligence technology, the system can automatically analyze the patient's various examination data, quickly identify symptoms, and assist doctors in formulating more accurate treatment plans.

[0124] Improve the quality of medical services: When the present invention conducts data mining based on the existing historical medical records in the target area, when conducting diagnostic analysis on patients in the same area, the present invention comprehensively analyzes the patient's symptoms and possible future symptoms based on the patient's symptom manifestations, age, gender and precipitating diseases, to help doctors better judge the patient's condition and adjust the treatment plan.

[0125] Enhance hospital competitiveness: The present invention reduces doctors' workload by providing intelligent auxiliary diagnosis services, allowing them to have more time to focus on patients' personalized needs, provide more considerate and comprehensive medical services, attract more patients to seek treatment, and thus enhance the hospital's visibility and reputation. BRIEF DESCRIPTION OF THE DRAWINGS

[0126] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0127] Figure 1 It is a schematic diagram of the method of the present invention;

[0128] Figure 2 It is a schematic diagram of the system of the present invention;

[0129] Figure 3 This is a schematic diagram of the database design of the present invention. DETAILED DESCRIPTION

[0130] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0131] Embodiment 1

[0132] See also Figure 1, a big data-based outpatient auxiliary examination analysis method includes:

[0133] Step S1: Obtain all historical medical records of the target hospital (last year), obtain the total number of residents in the target area (permanent population); extract the names of the symptoms in the historical medical records, summarize all the historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents;

[0134] It should be noted that the "target hospital" in the present invention refers to the hospital that uses the present invention (a method and system for outpatient auxiliary examination analysis based on big data); the "target area" refers to the city-level area where the "target hospital" is located;

[0135] The time span of the "historical medical records" in the present invention is 1 year, and the "historical medical records" in the present invention only count the historical medical records of the "permanent population" in the target area; if the floating population in the target area has lived in the target area for more than 3 years and stayed in the target area for at least 9 months each year, the floating population is also regarded as the "permanent population in the target area";

[0136] The "historical medical records" in the present invention can be obtained by associating with the "Hospital Information System (HIS)" of the target hospital; if the target hospital is a newly built hospital and there is no record in the hospital information system (of the target hospital), the "historical medical records" are changed to: obtaining the historical medical records of other hospitals (non-target hospitals) in the target area last year.

[0137] The specific steps of step S1 are as follows:

[0138] Step S11: Obtain all historical medical records of the target hospital (last year); extract all disease names from the historical medical records, count the number of all disease names, and record it as dn; obtain the total number of residents in the target area, and record it as pn;

[0139] Step S12: Summarize all historical medical records of patients with the first disease; calculate the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease, and obtain a comprehensive incidence rate coefficient group of the first disease;

[0140] It should be noted that the "first symptom" in the present invention refers to the disease corresponding to the "first symptom name"; the "second symptom" refers to the disease corresponding to the "second symptom name"; and so on, the "dnth symptom" refers to the disease corresponding to the "dnth symptom name".

[0141] Step S121: Record the patient suffering from the first disease as the target patient;

[0142] (Based on the premise that “each patient’s medical history is counted only once”) Count the number of medical history records of all target patients, recorded as fb;

[0143] Calculate the overall incidence of the first disease, denoted as Ir; the calculation formula for Ir is as follows:

[0144] Ir=fb / pn;wherein, Ir≤1;

[0145] Step S122: (Based on the premise that “only one historical medical record of each patient is counted”) count the number of historical medical records of “the target patient is male”, recorded as mb; count the number of historical medical records of “the target patient is female”, recorded as wb;

[0146] Calculate the male incidence of the first disease, recorded as Mr; the calculation formula for Mr is as follows:

[0147] Mr=mb / pn;wherein, Mr≤1;

[0148] Calculate the female incidence of the first disease, denoted as Wr; the calculation formula for Wr is as follows:

[0149] Wr=wb / pn;Wr≤1;

[0150] Among them, the relationship between mb, fb and fb satisfies: mb+fb=fb; the relationship between Mr, Wr and Ir satisfies: Mr+Wr=Ir;

[0151] Mr and Wr were used as the sex incidence of the first disease;

[0152] It should be noted that step S121 and the "precondition of 'only count the historical medical records of each patient once'" in step S121 mean that for multiple historical medical records of the same patient, "the 'number of historical medical records of all target patients' in step S121" and "the 'number of historical medical records of male target patients' or 'the number of historical medical records of female target patients' in step S122" are counted only once;

[0153] Example 1: If patient A has 5 historical medical records of "suffering from the first disease", then in step S121, the present invention only counts 1 historical medical record of patient A;

[0154] Example 1: Male patient A has 5 medical records of "suffering from the first disease", and female patient B has 3 medical records of "suffering from the first disease";

[0155] In step S122, for male patient A, the present invention only counts the number of historical medical records of male target patients once;

[0156] In step S122, the present invention only counts the number of historical medical records of female patient B whose target patient is female once.

[0157] Step S123: (Based on the premise that “historical medical records of each patient can be counted multiple times”) Please refer to Table 1, and count the number of historical medical records of infants and young children as target patients, recorded as fpn1;

[0158] The number of historical medical records of target patients who were adolescents was counted, recorded as fpn2;

[0159] The number of historical medical records of target patients who were adults was counted, recorded as fpn3;

[0160] The number of historical medical records of the target patient being middle-aged was counted, recorded as fpn4;

[0161] The number of historical medical records of the target patient being elderly was counted, recorded as fpn5;

[0162] Age range Group division (0，5] Infants and young children [6，18] teenager [19，40] adult [41，60) middle aged ≥60 elderly

[0163] Table 1 (Age classification table)

[0164] Among them, fpn1~fpn5 are all greater than or equal to 0; calculate the sum of fpn1~fpn5, denoted as fpn; the relationship between fpn and fb satisfies: fpn≥fb;

[0165] It should be noted that "Table 1" in the present invention is only for example demonstration, and users or relevant technical personnel can adjust the age classification table according to actual needs;

[0166] It should also be noted that the "precondition that 'each patient's historical medical records can be counted multiple times'" in step S123 means that for multiple historical medical records of the same patient suffering from the same disease at different age groups, "fpn1 to fpn5 in step S123" are counted multiple times, but for multiple historical medical records of the same patient suffering from the same disease at the same age group, "fpn1 to fpn5 in step S123" are counted only once;

[0167] Example 1: Patient A has a history of "suffering from the first disease" once in adolescence, and patient A has a history of "suffering from the first disease" once in middle age. For patient A, fpn2 and fpn4 in step S123 are counted once each;

[0168] Example 2: Patient B has a history of "suffering from the first disease" three times when he was a teenager. For patient B, fpn2 in step S123 is only counted once.

[0169] Step S124: The infant morbidity rate of the first disease is recorded as Fbb1; the calculation formula of Fbb1 is: Fbb1 = (Tr*fpn1) / fb;

[0170] The adolescent incidence of the first disease is recorded as Fbb2; the calculation formula of Fbb2 is: Fbb2 = (Tr*fpn2) / fb;

[0171] The adult incidence of the first disease is recorded as Fbb3; the calculation formula of Fbb3 is: Fbb3 = (Tr*fpn3) / fb;

[0172] The middle-aged incidence of the first disease is recorded as Fbb4; the calculation formula of Fbb4 is: Fbb4 = (Tr*fpn4) / fb;

[0173] The elderly incidence of the first disease is recorded as Fbb5; the calculation formula of Fbb5 is: Fbb5 = (Tr*fpn5) / fb;

[0174] Fbb1 to Fbb5 were used as the age incidence of the first disease;

[0175] Step S125: Calculate the age group coefficients a1-a5 of the first disease;

[0176] The calculation formula of a1 is: a1 = fpn1 / fpn; the calculation formula of a2 is: a2 = fpn2 / fpn; the calculation formula of a3 is: a3 = fpn3 / fpn; the calculation formula of a4 is: a4 = fpn4 / fpn; the calculation formula of a5 is: a5 = fpn5 / fpn;

[0177] (Based on the premise that "each patient can have histories of medical records counted multiple times"), count the illness time of all (target patients) infants and young children, and record it as atⅠ1~atⅠ fpn1 ; Recovery time, recorded as atⅡ1~atⅡ fpn1 ;

[0178] The duration of illness of all (target patients) adolescents is counted and recorded as btⅠ1~btⅠ fpn2 ; Recovery time, recorded as btⅡ1~btⅡ fpn2 ;

[0179] The duration of illness of all (target patients) adults is counted and recorded as ctⅠ1~ctⅠ fpn3 ; Recovery time, recorded as ctⅡ1~ctⅡ fpn3 ;

[0180] Count the duration of illness of all middle-aged patients (target patients), recorded as dtⅠ1~dtⅠ fpn4 ; Recovery time, recorded as dtⅡ1~dtⅡ fpn4 ;

[0181] Count the duration of illness of all elderly patients (target patients), recorded as etⅠ1~etⅠ fpn5 ; Recovery time, recorded as etⅡ1~etⅡ fpn5;

[0182] Step S126: Calculate the comprehensive treatment time of the first disease, recorded as tt; the calculation formula of tt is as follows:

[0183] ;

[0184] Step S127: Summarize the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease obtained in steps S121 to S126 as a comprehensive incidence rate coefficient group of the first disease;

[0185] Step S13: Repeat the same steps of calculating the comprehensive morbidity coefficient group of the first disease (step S12) to calculate the comprehensive morbidity coefficient groups of the second to dnth diseases.

[0186] Step S2: Obtain all physician information of the target hospital; obtain the symptom manifestations of the corresponding disease in the disease database according to the historical medical records, and determine the pathological classification of different diseases and the induction relationship between different diseases in combination with the onset symptoms of all diseases in the historical medical records; construct a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease;

[0187] The specific steps of step S2 are as follows:

[0188] Step S21: extracting the medical record information of each patient from the historical medical records;

[0189] The patient's medical record information includes: patient name, disease name, symptoms, illness duration and attending physician;

[0190] Step S22: Obtain the names of doctors in the target hospital and the corresponding departments of each doctor;

[0191] Link the disease database to obtain the symptom manifestations corresponding to (all) "symptom names" in historical medical records;

[0192] It should be noted that the "medical database" in step S22 refers to a "database or data storage system used to digitally store and manage disease information"; common medical databases include: National Health Database, Centers for Disease Control and Prevention database, Electronic Health Record System (HERs), etc.

[0193] Step S23: Please refer to Figure 3 Based on the data obtained in steps S21 to S22, a "patient" table (relational table), a "disease" table (relational table) and a "physician" table (relational table) are constructed in the database; the corresponding SQL statements are as follows:

[0194] Patients table:

[0195] Create table Patients(

[0196] Patient name String, disease name String, disease status String,

[0197] Illness time String, attending physician String,

[0198] primary key (patient name),

[0199] foreign key (disease name) referances disease (disease name)

[0200] Disease table:

[0201] Create table Disease(

[0202] Symptom name String, Symptom String,

[0203] primary key (disease name)

[0204] Physician table:

[0205] Create table Physician(

[0206] Doctor's name String, Department String,

[0207] primary key (Doctor's name)

[0208] Step S24: judging the pathological classification of different diseases and the inducing relationship between different diseases according to the "patient" table (relationship table), the "disease" table (relationship table) and the "physician" table (relationship table);

[0209] Step S241: Read the number of all disease names dn; (or use SQL statement 1' to count the number of all "disease names" in the "disease" table. SQL statement 1' is as follows:

[0210] select count(*) from information_schema.columns where table_name='disease';)

[0211] Step S242: Based on the "patient" table and the "physician" table, use SQL statement 2 to output the pathological classification of different diseases. SQL statement 1 is as follows:

[0212] select patient.disease name, physician.department count(select patient.attending physician, physician.physician name from patient inner join patient.attending physician=physician.physician name) from physician group by patient.disease name;

[0213] The output of SQL statement 1 is the pathological classification of the 1st to pnth diseases;

[0214] Step S243: determining the inducing relationship between the 1st to pnth diseases (i.e., disease-inducing disease relationship);

[0215] Step S2431: Use SQL statement 2 to obtain the first illness time corresponding to the 1st to pnth diseases. SQL statement 2 is as follows:

[0216] select disease name,MIN(illness time) from patient group by disease name;

[0217] Summarize the output data of SQL statement 2 to obtain the first illness time corresponding to the 1st to pnth diseases, recorded as ft1, ft2, ft3~ft pn ; (where ft1 is the first time the first disease occurs; ft2 is the first time the second disease occurs; ft3 is the first time the third disease occurs; and so on, ft pn Indicates the first time of onset of the pnth disease)

[0218] Step S2432: Use SQL statement 3 to obtain all the symptoms corresponding to the 1st to pnth diseases in the "patient" table. SQL statement 4 is as follows:

[0219] select disease name, symptoms from patient group by disease name;

[0220] Summarize the output data of SQL statement 3 to obtain the (all) patient symptom sets corresponding to the 1st to pnth diseases in the "patient" table, recorded as set Bm1, set Bm2, set Bm3~set Bm pn ; (wherein, set Bm1 represents all the symptoms of the first disease in the "patient" table; set Bm2 represents all the symptoms of the second disease in the "patient" table; and so on, set Bm pn Indicates all symptoms of the pnth disease in the "Patient" table)

[0221] Step S2433: Use SQL statement 4 to obtain all symptom manifestations corresponding to the 1st to pnth symptom in the "Disease" table. SQL statement 4 is as follows:

[0222] select disease name, disease manifestation from patient group by disease name;

[0223] Summarize the output data of SQL statement 4 to obtain the (all) sets of symptom manifestations corresponding to the 1st to pnth symptom in the "Disease" table, recorded as set Bi1, set Bi2, set Bi3~set Bi pn ; (wherein, set Bi1 represents all the symptoms of the first disease in the "Disease" table; set Bi2 represents all the symptoms of the second disease in the "Disease" table; set Bi3 represents all the symptoms of the third disease in the "Disease" table; and so on, set Bi pn Indicates all symptoms of the pnth disease in the "Disease" table)

[0224] Step S2434: among the first to pnth diseases, determining the inducing relationship between the first disease (and the second to pnth diseases);

[0225] Step S24341: determining whether there is an inducing relationship between the first symptom and the second symptom;

[0226] The first disease is recorded as disease A, and the second disease is recorded as disease B; (i.e. the tentative range of disease B is: the second disease)

[0227] In the comprehensive incidence coefficient group of the first disease, read the overall incidence of disease A (the first disease), recorded as P(A); in the comprehensive incidence coefficient group of the second disease, read the overall incidence of disease B (the second disease), recorded as P(B);

[0228] Step S24342: Using SQL statement 5, count the number of patients with both disease A and disease B in the "patient" table, denoted as pn AB ; SQL statement 5 is as follows:

[0229] select count(select patient name from patient where disease name=disease A and disease name=disease B) as pn AB from patient;

[0230] Calculate the probability of having both disease A and disease B, recorded as P(AB); the calculation formula of P(AB) is: P(AB) = pn AB / pn;

[0231] Step S24343: Calculate the probability of first contracting disease A and then contracting disease B, recorded as P(B|A); the calculation formula of P(B|A) is: P(B|A)=P(AB) / P(A);

[0232] Calculate the probability of contracting disease B before contracting disease A, recorded as P(A|B); the calculation formula of P(A|B): P(A|B) = P(AB) / P(B);

[0233] Step S24344: defining relationship equations 1 to 2 of “there is an induction relationship between the first symptom and the second symptom (i.e., the first symptom induces the second symptom)”;

[0234] Relationship 1: ft1≤ftB; where ftB represents ft2; (i.e., the tentative range of ftB is: ft2)

[0235] Relation 2: ; Among them, BmB and BiB represent the set Bm2 and the set Bi2 respectively (that is, the tentative representation range of the set BmB is: set Bm2; the tentative representation range of the set BiB is: set Bi2)

[0236] Relation 3: P(B|A)+P(A|B)<1;

[0237] Step S24345: Determine whether all of equations 1 to 3 are satisfied;

[0238] If all of these conditions are met, it means that there is an induction relationship between the first symptom and the second symptom, that is, the first symptom induces the second symptom;

[0239] If none of them are met or partially met, it means that there is no inductive relationship between the first and second symptoms;

[0240] Step S24346: Expand the representation range of disease B to the third to pnth diseases, and expand the representation range of ftB to the set ft3 to the set ft pn , the range of BmB and BiB is extended to the set Bm3 to the set Bm pn and set Bi3~set Bi pn Repeat steps S24341 to S24345 to determine whether there is an inducing relationship between the first symptom and the third to pn th symptom, and summarize the results;

[0241] Step S2435: repeating the same steps of determining the inducing relationship of the first disease (i.e., steps S24341 to S24346) to determine the inducing relationship of the second to pnth diseases;

[0242] Step S25: constructing a disease diagnosis model based on the random forest model;

[0243] Symptom diagnosis model: used to determine the patient's diagnosed symptoms, medical department, pending treatment time and diagnosed induced diseases according to the patient's symptom characteristics;

[0244] Step S251: define the processing logic of the (random forest model) decision tree; assume that the model (symptom diagnosis model) input is X, where X represents the symptom characteristics of the patient;

[0245] Step S252: (load jieba library, Gensim library and transformers library; use Word2Vec model to determine whether X is in set Bi1~set Bi pn The matching degree in

[0246] Let X and set Bi i The matching degree is mat(X_i);

[0247] The logical expression of mat(X_i): mat(X(left scale)_i(right scale)) = Word2Vec(X, Bi i ), where i ranges from 1 to pn, Bi i represents the set of symptom manifestations corresponding to the ith symptom (i.e. Bi i The range of representation is: set Bi1 to set Bi pn ); Word2Vec represents the Word2Vec model (the Word2Vec model is a model for calculating word vectors, which represents the semantic relationship between words);

[0248] Use logical expressions to calculate the relationship between X and the set Bi1 to the set Bi pn The matching degree is obtained by mat(X_1)~mat(X_pn);

[0249] Step S253: Obtain a judgment threshold, denoted as β; define relation 4: mat(X_i)>β;

[0250] Substitute mat(X_1)~mat(X_pn) into equation 4, and summarize the right index of mat(X_i) that satisfies the matching degree of equation 4 as the search sequence number; read the corresponding disease according to the search sequence number, record it as the pending disease, and count the number of pending diseases, record it as uni;

[0251] It should be noted that the "judgment threshold" in the present invention represents β: the difference between the input X of the disease diagnosis model and the set Bi1 to the set Bi2. pn " similarity, the value of β is 0.75; users or relevant technical personnel can adjust the value of β according to actual needs.

[0252] Step S254: determine the patient's gender and age (step S123), and read the gender incidence rate and age incidence rate corresponding to the pending disease;

[0253] Definition selection formula 5: g(j) = mat(X_j)*(fpb(j)*Pr(j))1 / 2 ; Wherein, the value range of j is: 1~uni; g(j) represents the selection coefficient of the jth undetermined disease, mat(X_j) represents the matching degree of the disease manifestation set corresponding to X and the jth undetermined disease, fpb(j) represents the gender incidence rate of the jth undetermined disease (corresponding to the patient's gender), and Pr(j) represents the age incidence rate of the jth undetermined disease (corresponding to the patient's age);

[0254] According to the selection formula 5, the selection coefficients g(1) to g(uni) of the undetermined diseases from No. 1 to No. uni are calculated;

[0255] Step S255: Compare the sizes of g(1) to g(uni), read the disease with the largest selection coefficient as the determined disease; call SQL statement 1 according to the determined disease to obtain the pathological classification of the determined disease, that is, the medical department; read the comprehensive treatment time corresponding to the determined disease as the pending treatment time;

[0256] Step S256: (based on the inducing relationship between different symptoms determined in step S24) read all the inducing symptoms corresponding to the determined symptoms, which are recorded as pending inducing symptoms; count the number of pending inducing symptoms, which is recorded as unf;

[0257] Step S257: Read the gender incidence rate and age incidence rate corresponding to the pending induced disease according to the patient's gender and age;

[0258] Definition selection formula 6: g`(q) = (fpb`(q)*Pr`(q)) 1 / 2 ; Wherein, the value range of q is: 1~unf; g`(q) represents the alternative coefficient of the qth undetermined induced disease, fpb`(q) represents the gender incidence rate of the qth undetermined induced disease (corresponding to the patient's gender), and Pr`(q) represents the age incidence rate of the qth undetermined induced disease (corresponding to the patient's age);

[0259] According to the selection formula 6, the candidate coefficients g`(1)~g`(unf) of the first to unf undetermined diseases are calculated;

[0260] Step S258: Compare the sizes of g`(1) to g`(unf), and read the induced disease with the largest candidate coefficient as the determined induced disease;

[0261] Step S259: Output the determined symptom, medical department, pending treatment time and determined induced disease corresponding to the patient's symptom feature X.

[0262] Step S3: Obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, and read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model (symptom diagnosis model) as the initial diagnosis report;

[0263] Step S4: Summarize the initial diagnosis report and provide feedback to the physician;

[0264] Embodiment 2

[0265] See also Figure 2 An outpatient auxiliary examination and analysis system based on big data includes: a data acquisition module, a model building module, a patient analysis module, a physician interaction module, a database and a server; wherein the data acquisition module, the model building module, the patient analysis module and the physician interaction module are connected to the database and the server respectively.

[0266] Medical record acquisition module: used to obtain all historical medical records of the target hospital (last year) and the total number of residents in the target area; extract the names of the symptoms in the historical medical records, summarize all the historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents;

[0267] Model building module: used to obtain all physician information of the target hospital; obtain the symptom manifestations of the corresponding disease according to the historical medical record disease database, and judge the pathological classification of different diseases and the induction relationship between different diseases according to the symptoms of all diseases in the historical medical records; build a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease;

[0268] Patient analysis module: used to obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model (symptom diagnosis model) as the initial diagnosis report;

[0269] Physician interaction module: used to summarize the initial diagnosis report and provide feedback to the physician.

[0270] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technical personnel in this field according to actual conditions. For example, if there are weight coefficients and proportional coefficients, their set sizes are to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. Regarding the size of the weight coefficient and the proportional coefficient, it is sufficient as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0271] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for outpatient auxiliary examination and analysis based on big data, characterized in that: Analytical methods include: Step S1: Obtain all historical medical records of the target hospital and the total number of residents in the target area; extract the names of the symptoms in the historical medical records, summarize all historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents; Step S2: Obtain all physician information of the target hospital; obtain the symptom manifestations of the corresponding disease in the disease database according to the historical medical records, and determine the pathological classification of different diseases and the induction relationship between different diseases in combination with the onset symptoms of all diseases in the historical medical records; construct a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease; Step S3: Obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model as the initial diagnosis report; Step S4: Summarize the initial diagnosis report and provide feedback; The specific steps of step S1 are as follows: Step S11: Obtain all historical medical records of the target hospital; extract all disease names from the historical medical records, count the number of all disease names, and record it as dn; obtain the total number of residents in the target area, and record it as pn; Step S12: Summarize all historical medical records of patients with the first disease; calculate the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease, and obtain a comprehensive incidence rate coefficient group of the first disease; Step S13: Repeat the same steps of calculating the comprehensive morbidity coefficient group of the first disease to calculate the comprehensive morbidity coefficient groups of the second to dnth diseases; The specific steps of step S12 are as follows: Step S121: Record the patient suffering from the first disease as the target patient; Count the number of historical medical records of all target patients, recorded as fb; Calculate the overall incidence of the first disease, denoted as Ir; the calculation formula for Ir is as follows: Ir=fb / pn;wherein, Ir≤1; Step S122: Count the number of historical medical records of male target patients, recorded as mb; count the number of historical medical records of female target patients, recorded as wb; Calculate the male incidence of the first disease, recorded as Mr; the calculation formula for Mr is as follows: Mr=mb / pn;wherein, Mr≤1; Calculate the female incidence of the first disease, denoted as Wr; the calculation formula for Wr is as follows: Wr=wb / pn;Wr≤1; Among them, the relationship between mb, fb and fb satisfies: mb+fb=fb; the relationship between Mr, Wr and Ir satisfies: Mr+Wr=Ir; Mr and Wr were used as the sex incidence of the first disease; Step S123: Count the number of historical medical records of infants and young children as target patients, recorded as fpn1; The number of historical medical records of target patients who were adolescents was counted, recorded as fpn2; The number of historical medical records of target patients who were adults was counted, recorded as fpn3; The number of historical medical records of the target patient being middle-aged was counted, recorded as fpn4; The number of historical medical records of the target patient being elderly was counted, recorded as fpn5; Among them, fpn1~fpn5 are all greater than or equal to 0; calculate the sum of fpn1~fpn5, denoted as fpn; the relationship between fpn and fb satisfies: fpn≥fb; Step S124: The infant morbidity rate of the first disease is recorded as Fbb1; the calculation formula of Fbb1 is: Fbb1 = (Tr*fpn1) / fb; The adolescent incidence of the first disease is recorded as Fbb2; the calculation formula of Fbb2 is: Fbb2 = (Tr*fpn2) / fb; The adult incidence of the first disease is recorded as Fbb3; the calculation formula of Fbb3 is: Fbb3 = (Tr*fpn3) / fb; The middle-aged incidence of the first disease is recorded as Fbb4; the calculation formula of Fbb4 is: Fbb4 = (Tr*fpn4) / fb; The elderly incidence of the first disease is recorded as Fbb5; the calculation formula of Fbb5 is: Fbb5 = (Tr*fpn5) / fb; Fbb1 to Fbb5 were used as the age incidence of the first disease; Step S125: Calculate the age group coefficients a1-a5 of the first disease; The calculation formula of a1 is: a1 = fpn1 / fpn; the calculation formula of a2 is: a2 = fpn2 / fpn; the calculation formula of a3 is: a3 = fpn3 / fpn; the calculation formula of a4 is: a4 = fpn4 / fpn; the calculation formula of a5 is: a5 = fpn5 / fpn; The sick time of all infants and young children is recorded as atⅠ1~atⅠ fpn1 ; Recovery time, recorded as atⅡ1~atⅡ fpn1 ; The duration of illness of all adolescents is counted and recorded as btⅠ1~btⅠ fpn2 ; Recovery time, recorded as btⅡ1~btⅡ fpn2 ; The duration of illness of all adults was counted and recorded as ctⅠ1~ctⅠ fpn3 ; Recovery time, recorded as ctⅡ1~ctⅡ fpn3 ; The total duration of illness in middle-aged people is recorded as dtⅠ1~dtⅠ fpn4 ; Recovery time, recorded as dtⅡ1~dtⅡ fpn4 ; The duration of illness of all elderly people is counted and recorded as etⅠ1~etⅠ fpn5 ; Recovery time, recorded as etⅡ1~etⅡ fpn5 ; Step S126: Calculate the comprehensive treatment time of the first disease, recorded as tt; the calculation formula of tt is as follows: ； Step S127: Summarize steps S121 to S126 to obtain the overall incidence rate, gender incidence rate, age incidence rate and comprehensive treatment time corresponding to the first disease as the comprehensive incidence rate coefficient group of the first disease.

2. According to the big data-based outpatient auxiliary examination analysis method of claim 1, it is characterized in that: The specific steps of step S2 are as follows: Step S21: extracting the medical record information of each patient from the historical medical records; The patient's medical record information includes: patient name, disease name, symptoms, illness time and attending physician; Step S22: Obtain the names of doctors in the target hospital and the corresponding departments of each doctor; Link to the disease database to obtain the symptoms corresponding to the disease names in the historical medical records; Step S23: Based on the data obtained in steps S21 to S22, construct a patient table, a disease table, and a physician table in the database; the corresponding SQL statement is as follows: Patient table: Create table Patients( Patient name String, disease name String, disease status String, Illness time String, attending physician String, primary key (patient name), foreign key (disease name) referances disease (disease name) Disease table: Create table Disease( Symptom name String, Symptom String, primary key (disease name) Physician table: Create table Physician( Doctor's name String, Department String, primary key (Doctor's name) Step S24: judging the pathological classification of different symptoms and the inducing relationship between different symptoms according to the patient table, the disease table and the physician; Step S25: constructing a disease diagnosis model based on the random forest model; Symptom diagnosis model: used to determine the patient's diagnosed symptoms, medical department, pending treatment time and diagnosed induced diseases based on the patient's symptom characteristics.

3. The outpatient auxiliary examination and analysis method based on big data according to claim 2 is characterized in that: The specific steps of step S24 are as follows: Step S241: Read the number dn of all disease names; Step S242: Based on the patient table and the physician table, the pathological classification of different diseases is output using SQL statement 2. SQL statement 1 is as follows: select patient.disease name, physician.department count(select patient.attending physician, physician.physician name from patient inner join patient.attending physician=physician.physician name) from physician group by patient.disease name; The output of SQL statement 1 is the pathological classification of the 1st to pnth diseases; Step S243: determining the inducing relationship between the 1st to pnth diseases; Step S2431: Obtain the first illness time corresponding to the 1st to pnth diseases, recorded as ft1, ft2, ft3~ft pn ; Step S2432: Obtain all the symptoms corresponding to the 1st to pnth diseases, and obtain sets Bm1, Bm2, Bm3 to Bm4. pn ; Step S2433: Obtain all symptom manifestations corresponding to the 1st to pnth symptom, and obtain sets Bi1, Bi2, Bi3 to Bi4. pn ; Step S2434: determining the inducing relationship of the first symptom among the first to pn th symptom; Step S2435: Repeat the same steps of determining the induction relationship of the first disease to determine the induction relationship of the second to pnth diseases.

4. The outpatient auxiliary examination and analysis method based on big data according to claim 3 is characterized in that: The specific steps of step S2434 are as follows: Step S24341: determining whether there is an inducing relationship between the first symptom and the second symptom; The first disease is recorded as disease A, and the second disease is recorded as disease B; In the comprehensive incidence coefficient group of the first disease, read the overall incidence of disease A, recorded as P (A); in the comprehensive incidence coefficient group of the second disease, read disease B; Step S24342: Count the number of people who suffer from both disease A and disease B, recorded as pn AB ; Calculate the probability of having both disease A and disease B, recorded as P(AB); the calculation formula of P(AB) is: P(AB) = pn AB / pn; Step S24343: Calculate the probability of first contracting disease A and then contracting disease B, recorded as P(B|A); the calculation formula of P(B|A) is: P(B|A)=P(AB) / P(A); Calculate the probability of contracting disease B before contracting disease A, recorded as P(A|B); the calculation formula of P(A|B): P(A|B) = P(AB) / P(B); Step S24344: defining relationship equations 1 to 2 that the first symptom and the second symptom have an inducing relationship; Relationship 1: ft1≤ftB; where ftB represents ft2; Relation 2: ; Where BmB and BiB represent the set Bm2 and the set Bi2 respectively Relation 3: P(B|A)+P(A|B)<1; Step S24345: Determine whether all of equations 1 to 3 are satisfied; If all of these conditions are met, it means that there is an induction relationship between the first symptom and the second symptom, that is, the first symptom induces the second symptom; If none of them are met or partially met, it means that there is no inductive relationship between the first and second symptoms; Step S24346: Expand the representation range of disease B to the third to pnth diseases, and expand the representation range of ftB to the set ft3 to the set ft pn , the range of BmB and BiB is extended to the set Bm3 to the set Bm pn and set Bi3~set Bi pn ; Repeat steps S24341 to S24345 to determine whether there is an inducing relationship between the first symptom and the third to pnth symptoms, and summarize them.

5. The outpatient auxiliary examination and analysis method based on big data according to claim 4 is characterized in that: The specific steps of step S25 are as follows: Step S251: define the processing logic of the decision tree; assume that the model input is X, where X represents the symptom characteristics of the patient; Step S252: Let X and set Bi i The matching degree is mat(X_i); The logical expression of mat(X_i): mat(X_i) = Word2Vec(X, Bi i ), where i ranges from 1 to pn, Bi i represents the set of symptom manifestations corresponding to the ith symptom; Word2Vec represents the Word2Vec model; Use logical expressions to calculate the relationship between X and the set Bi1 to the set Bi pn The matching degree is obtained by mat(X_1)~mat(X_pn); Step S253: Obtain a judgment threshold, denoted as β; define relation 4: mat(X_i)>β; Substitute mat(X_1)~mat(X_pn) into equation 4, and summarize the right index of mat(X_i) that satisfies the matching degree of equation 4 as the search sequence number; read the corresponding disease according to the search sequence number, record it as the pending disease, and count the number of pending diseases, record it as uni; Step S254: determine the patient's gender and age, and read the gender incidence rate and age incidence rate corresponding to the pending disease; Definition selection formula 5: g(j) = mat(X_j)*(fpb(j)*Pr(j)) 1 / 2 ; Wherein, the value range of j is: 1~uni; g(j) represents the selection coefficient of the jth undetermined disease, mat(X_j) represents the matching degree of the disease manifestation set corresponding to X and the jth undetermined disease, fpb(j) represents the gender incidence rate of the jth undetermined disease, and Pr(j) represents the age incidence rate of the jth undetermined disease; According to the selection formula 5, the selection coefficients g(1) to g(uni) of the undetermined diseases from No. 1 to No. uni are calculated; Step S255: Compare the sizes of g(1) to g(uni), read the disease with the largest selection coefficient as the determined disease; call SQL statement 1 according to the determined disease to obtain the pathological classification of the determined disease, that is, the medical department; read the comprehensive treatment time corresponding to the determined disease as the pending treatment time; Step S256: read all the induced diseases corresponding to the determined disease, and record them as pending induced diseases; count the number of pending induced diseases, and record them as unf; Step S257: Read the gender incidence rate and age incidence rate corresponding to the pending induced disease according to the patient's gender and age; Definition selection formula 6: g`(q) = (fpb`(q)*Pr`(q)) 1 / 2 ; Wherein, the value range of q is: 1~unf; g`(q) represents the alternative coefficient of the qth undetermined induced disease, fpb`(q) represents the gender incidence rate of the qth undetermined induced disease, and Pr`(q) represents the age incidence rate of the qth undetermined induced disease; According to the selection formula 6, the candidate coefficients g`(1)~g`(unf) of the first to unf undetermined diseases are calculated; Step S258: Compare the sizes of g`(1) to g`(unf), and read the induced disease with the largest candidate coefficient as the determined induced disease; Step S259: Output the determined symptom, medical department, pending treatment time and determined induced disease corresponding to the patient's symptom feature X.

6. An outpatient auxiliary examination and analysis system based on big data, applicable to any one of the outpatient auxiliary examination and analysis methods based on big data in claims 1-5, characterized in that: The analysis system comprises: Medical record acquisition module: used to obtain all historical medical records of the target hospital and the total number of residents in the target area; extract the names of the symptoms in the historical medical records, summarize all the historical medical records corresponding to each symptom, and calculate the comprehensive incidence coefficient group of each symptom in combination with the total number of residents; Model building module: used to obtain all the physician information of the target hospital; obtain the symptom manifestations of the corresponding disease in the disease database according to the historical medical records, and judge the pathological classification of different diseases and the induction relationship between different diseases by combining the symptoms of all diseases in the historical medical records; build a disease diagnosis model according to the induction relationship between different diseases and the comprehensive incidence coefficient group of each disease; Patient analysis module: used to obtain the patient's gender, age and symptom characteristics as judgment parameters; input the judgment parameters into the symptom diagnosis model, read the determined symptom, medical department, pending treatment time and determined induced symptom output by the model as the initial diagnosis report; Physician interaction module: used to summarize the initial diagnosis report and provide feedback to the physician.

Images