Digital intelligent robot for health risk assessment based on big data

Abstract

The invention provides a digital intelligent robot for health risk assessment based on big data, and belongs to the field of health assessment. The problem of low health assessment efficiency is solved; the specific evaluation system comprises an information acquisition module for acquiring pre-diagnosis information; the medical record analysis module is used for acquiring pathological data, calculating a benign treatment cycle and malignant treatment of each disease, and constructing a benign lesion equation and a malignant lesion equation of each disease; the patient diagnosis module is used for determining disease names and treatment drugs of the patients, acquiring to-be-analyzed parameters, and analyzing pathological changes and diseased time of the patients in combination with benign lesion equations and malignant lesion equations corresponding to the disease names to obtain a pre-diagnosis report; the physician interaction module is used for feeding back a pre-diagnosis report; by acquiring, processing and analyzing the relevant information of the pathology of the patient, the disease and disease change of the patient are determined, and the treatment of the patient is facilitated.

Description

Technical Field

[0001] The present invention discloses a digital intelligent robot for health risk assessment based on big data, which relates to the field of health assessment. Background Art

[0002] Existing health assessment systems or digital robots have the following shortcomings: Failure to provide complete medical history: Existing systems usually only focus on current health data and potential health risks, and are unable to provide complete medical history information in the patient’s area; this will lead to a lack of comprehensive medical history reference when formulating health management plans, thus affecting the accuracy and effectiveness of the plans.

[0003] The assessment content is redundant and not targeted: Existing health management projects include multiple links such as health information collection, health risk assessment and health hazard prediction; while the existing health assessment system only completes the health risk assessment part; it cannot constitute a complete health management project and needs to be combined with other health management measures.

[0004] The problem of single purpose: Most of the existing (medical) digital robots are engaged in patrol inspections, drug transportation, guard security and other tasks that are irrelevant to patient recovery or have little significance. In addition, some existing health assessment systems can only perform risk assessments and have few auxiliary functions for patient treatment or doctors. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a digital robot for health risk assessment based on big data, aiming to solve the problem of low efficiency of health assessment.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions: A digital robot for health risk assessment based on big data includes: Health risk assessment systems and digital robots; The health risk assessment system includes: information acquisition module: used to obtain the patient's medical examination report, extract the patient's disease pictures and symptoms, and obtain pre-diagnosis information; Medical record analysis module: used to obtain all medical record information of the target hospital as reference medical records; extract the disease name, illness time, recovery time, lesion picture, clinical symptoms, clinical medication and lesion index content in the reference medical records to obtain pathological data; According to the duration of illness, recovery time and lesion index content of each disease, the benign treatment cycle and malignant treatment cycle of each disease are calculated, and the benign lesion equation and malignant lesion equation of each disease are constructed; Patient diagnosis module: used to build a disease diagnosis model and a drug matching model based on the disease name, lesion pictures, clinical symptoms and clinical medications; substitute the pre-diagnosis information into the disease diagnosis model and the drug matching model to determine the patient's disease name and treatment drugs; obtain the patient's lesion index content as the parameter to be analyzed, and combine the benign lesion equation and malignant lesion equation corresponding to the disease name to analyze the patient's pathological changes and the duration of illness to obtain a pre-diagnosis report; Physician interaction module: used to provide feedback on pre-diagnosis reports and control the digital robot based on the pre-diagnosis reports; Digital robot: used to link the health risk assessment system, obtain the patient's medical examination report, extract pre-diagnosis information; register the patient based on the pre-diagnosis report.

[0007] Furthermore, the workflow of the medical record analysis module is as follows: Process A1: Obtain all medical records of the target hospital as reference medical records; The disease names, illness duration, recovery time, lesion pictures, clinical symptoms, clinical medications and lesion index contents of all patients in the reference medical records were extracted to obtain pathological data; The content of pathological indicators includes: the patient's body temperature, immune cell content, pathogen content and endocrine content; Process A2: Count the number of all disease names in the reference medical records, recorded as bn; Count the number of all reference medical records, recorded as ln; Count the number of pathological data corresponding to the first disease, denoted as bl （1） ; Similarly, count the number of pathological data corresponding to the bn-th disease name, denoted as bl （bn） ; Among them, bl （1） ~bl （bn） The sum of is ln; Process A3: Count the illness time of all patients in the reference medical records and obtain tf(1,1)~tf(bn,bl （bn） ); By counting the recovery time of all patients in the reference medical records, we can get te(1,1)~te(bn,bl （bn） ); Among them, tf (1,1) and te (1,1) represent the time when the first patient with the first disease becomes ill and the time when the patient recovers, respectively; By analogy, tf(bn,bl （bn） ) and te(bn,bl （bn） ), respectively representing the bl （bn） The duration of illness of a patient with the bnth disease and the time of recovery; Process A4: Define calculation formula a1: tt（i，j） =te （i，j） -tf （i，j） ; Among them, tt （i，j） 、te （i，j） and tf （i，j） , respectively represent the illness time, recovery time and treatment cycle of the jth patient with the i-th disease; the value range of i is: 1~bn; The value range of j changes with i; if i is 1, the value range of j is: 1~bl （1） ; Similarly, if i is bn, the value range of j is: 1~bl （bn） ; Set tf(1,1)~tf(bn,bl （bn） ) and te (1,1) ~ te (bn,bl （bn） ) into the calculation formula a1, we get tt(1,1)~tt(bn,bl （bn） ); Among them, tt(1,1) represents the treatment cycle of the first patient with the first disease; By analogy, tt (bl （bn） ,bn) indicates the bl （bn） treatment cycles for patients with bn disease; According to tt (1,1) ~ tt (bn,bl （bn） ), calculate the weighted treatment time of the 1st to bnth diseases, and get tz （1） ~tz （bn） .

[0008] Furthermore, the subsequent process of process A4 is as follows: Process A5: Calculate the proportion of benign lesions and the proportion of malignant lesions for the first condition; Define the relationship a2: tt(1,i)≤tz （1） ; Where tt(1,i) represents the treatment cycle of the i-th patient with the first disease, and the value range of i is: 1~bl （1） ; Set tt(1,1)~tt(1,bl （1） ) into relational expression a2, and the patients who satisfy relational expression a2 are summarized as benign patients, and the number of benign patients is recorded as sn1. The patients who do not satisfy relational expression a2 are summarized as malignant patients, and the number of malignant patients is recorded as sn2. Among them, sn1 + sn2 = bl （1） ; Calculate the proportion of benign lesions in the first condition, denoted as yb （1） ;yb （1） The calculation formula is: =sn1 / bl （1） ; Calculate the proportion of benign lesions in the first condition, denoted as nb （1） ; nb （1） The calculation formula is: =sn2 / bl （1） ; Process A6: Repeat the process of calculating the proportion of benign lesions and the proportion of malignant lesions in the first disease, and calculate the proportion of benign lesions in the second to bnth diseases, and obtain: yb （2） ~yb （bn） ; The proportion of malignant lesions in the second to bnth diseases is: nb （2） ~nb （bn） ; Process A7: calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and constructing the benign lesion equation and the malignant lesion equation of the first disease; Process A8: Repeat the same process of calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and calculate the benign treatment cycle of the second to bnth diseases, recorded as ty （2） ~ty （bn） ; Calculate the malignant treatment cycle of the 2nd to bnth diseases, recorded as tn （2） ~tn （bn） ; Repeat the same process of constructing the benign lesion equation and the malignant lesion equation for the first disease, construct the benign lesion equation and the malignant lesion equation for the second to bnth diseases, and enter the patient diagnosis module.

[0009] Furthermore, the specific process of process A4 is as follows: Process A41: According to tt (1,1) ~ tt (1,bl （1） ), calculate the weighted treatment time tz of the first disease （1） ; Extract tt(1,1)~tt(1,bl （1） ) is denoted as ttm, and the minimum value is denoted as ttl; calculate tt(1,1)~tt(1,bl （1） ), denoted as att; Process A42: At tt(1,1)~tt(1,bl （1） ), count the number of cells in the interval [ttl, att), denoted as aa1; count the number of cells in the interval [att, ttm), denoted as aa2; tz （1） The calculation formula is: ; Process A43: Repeated calculation of tz （1） The same process as above is used to calculate the weighted treatment time for the 2nd to bnth diseases, and we get tz（2） ~tz （bn） .

[0010] Furthermore, the specific process of process A7 is as follows: Process A71: benign patients with the first disease are classified as category a patients, and malignant patients with the first disease are classified as category b patients; The treatment cycles of the first to the sn1th patients of category a are recorded as: tyy(1,1)~tyy(1,sn1); the treatment cycles of the first to the sn2th patients of category b are recorded as: tnn(1,1)~tnn(1,sn2); Calculate the average value of tyy(1,1)~tyy(1,sn1) and record it as ty （1） , will ty （1） As a benign treatment cycle for the first condition; Calculate the average value of tnn(1,1)~tnn(1,sn1) and record it as tn （1） , will tn （1） As a vicious treatment cycle of the first disease; Process A72: The first patient of type a is taken as the target patient, and the time of illness of the target patient is recorded as st, and the time of recovery is recorded as ts; where st∈{tf(1,1)~tf(1,bl （1） )}, ts∈{te(1,1)~te(1,bl （1） )}; In the reference medical records, the temperature changes of the target patient in the time period of st to ts are extracted to obtain T （st） , T （st+1） ~T （ts） ; Changes in immune cell content, get Im （st） 、Im （st+1） ~Im （ts） ; Changes in pathogen content, get Pa （st） 、Pa （st+1） ~Pa （ts） ; Changes in endocrine content, obtain Nf （st） 、Nf （st+1） ~Nf （ts） ; Among them, T （st） 、Im （st） 、Pa （st） and Nf （st） , respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at st; T （st+1） 、Im （st+1） 、Pa （st+1） and Nf （st+1）, respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at (st+1); By analogy, T （ts） 、Im （ts） 、Pa （ts） and Nf （ts） , respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at ts; Process A73: Extraction of T （st） ~T （ts） The maximum value is denoted as Tm, and the minimum value is denoted as Tl; Extract Im （st） ~Im （ts） The maximum value is recorded as Imm, and the minimum value is recorded as Iml; Extraction of Pa （st） ~Pa （ts） The maximum value is denoted as Pam, and the minimum value is denoted as Pal; Extraction of Nf （st） ~Nf （ts） The maximum value is denoted as Nfm, and the minimum value is denoted as Nfl; The highest order of body temperature is denoted as tx; the calculation formula for tx is: tx = log Tl (Tm); The highest order of calculating immune cell content is recorded as ix; the calculation formula of ix is: ix = log Iml (Imm); The highest order of pathogen content is calculated, denoted as px; the calculation formula for px is: px = log Pal (Pam); The highest order number for calculating the endocrine content is recorded as nx; the calculation formula for nx is: nx = log Nfl (Nfm); Among them, tx, ix, px and nx are all positive integers, rounded up; Process A74: Assume that a good equation for the change of body temperature with time is equation a1; determine a mathematical expression of equation a1; Process A75: Assume that a good equation for the change of immune cell content over time is equation b1; determine a mathematical expression of equation b1; Process A76: Repeat the same process of calculating equation a1 and equation b1 to calculate the benign equation of pathogen content changing over time and the benign equation of endocrine content changing over time; The benign equations of body temperature changing with time, immune cell content changing with time, pathogen content changing with time, and endocrine content changing with time are summarized as the benign lesion equation of the first disease; Process A77: Repeat the same process of constructing the benign lesion equation for the first condition to calculate the malignant lesion equation for the first condition.

[0011] Furthermore, the specific process of the process A74 is as follows: Process A741: The mathematical expression of equation a1 is: ; Where x represents the number of days of illness, i.e., the independent variable; T(x) represents the body temperature of the target patient on the xth day, i.e., the body temperature when the patient is confirmed to be ill; T(x+1) represents the body temperature of the target patient on the (x+1)th day, i.e., the dependent variable; α0 represents the parameter-free weight of equation a1, and the initial value of α0 is 0; x j represents the j-order term of the independent variable, α j It represents the parameter weight of the j-order term of the independent variable in equation a1, j represents the order of the independent variable, and the value range of j is: 1~tx; ! represents factorial; The expanded form of the equation T(x+1) is: ; Among them, α1~α tx Indicates the parameter weights of the first-order term to the tx-order term of the independent variable in equation a1, α1~α tx The initial value of is 1; Process A742: T （s） As the initial value, T （st+1） ~T （ts） Substituting tx into equation a1, we get equation group a1-1; The mathematical expression of equation group a1-1 is: ; Use MATLAB software to calculate the program group a1-1 and get α0~α tx The value of , denoted as yα0~yα tx ; Set yα0~yα tx Substituting this into the equation T(x+1), we get a good equation for how body temperature changes over time.

[0012] Furthermore, the specific process of the process A75 is as follows: Process A751: The mathematical expression of equation b1 is: ; Among them, y represents the number of days of illness, i.e., the independent variable; Im(y) represents the immune cell content of the target patient on day y, i.e., the immune cell content when the patient is confirmed to be ill; Im(y+1) represents the immune cell content of the target patient on day (y+1), i.e., the dependent variable; β0 represents the parameter-free weight of equation b1, and the initial value of β0 is 0; y jrepresents the i-th order term of the independent variable in equation b1, β i It represents the parameterized weight of the term of the independent variable i, i represents the order of the independent variable, and the value range of i is: 1~tx; ! represents factorial; Process A752: Im （s） As the initial value, Im （st+1） ~Im （ts） Substituting tx into equation b1, we get equation group b1-1; The mathematical expression of equation group b1-1 is: ; Among them, β1~β tx Indicates the parameter weights of the first-order term of the independent variable of equation b1 to the tx-order term, β1~β tx The initial value of is 1; Use MATLAB software to calculate program group b1-1 and get β0~β tx The value of , denoted as yβ0~yβ tx ; Set yβ0~yβ tx Substituting into the equation Im(x+1), we obtain a good equation for the change of immune cell content over time.

[0013] Furthermore, the workflow of the patient diagnosis module is as follows: Process B1: In the reference medical records, the lesion images corresponding to each disease are summarized as sample data A, the clinical symptoms as sample data B, and the clinical medication as sample data C; According to sample data A and sample data B, a disease diagnosis model is constructed; Process B2: Repeat the same process of building a disease diagnosis model, using sample data A and sample data C as combined sample data to build a drug matching model; Substitute the pre-diagnosis information into the disease diagnosis model and drug matching model to determine the patient's disease name and treatment drug; Process B3: Obtain the weighted treatment time tz for the 1st to bnth diseases （1） ~tz （bn） , proportion of benign lesions （1） ~yb （bn） , proportion of malignant lesions nb （1） ~nb （bn） ; The patient's disease name is used as the quasi-disease name, and the benign lesion equation and the malignant lesion equation corresponding to the quasi-disease name are obtained; Obtain the patient's current body temperature as To, immune cells as Imo, pathogen content as Pao, and endocrine content as Nfo as parameters to be analyzed; Process B4: The benign equation of the change of body temperature over time in the benign lesion equation corresponding to the quasi-disease name is recorded as equation yy1; the benign equation of the change of immune cell content over time is recorded as equation yy2; the benign equation of the change of pathogen content over time is recorded as yy3; the benign equation of the change of endocrine content over time is recorded as yy4; The malignant equation of the change of body temperature over time in the malignant lesion equation corresponding to the quasi-disease name is recorded as equation nn1; the malignant equation of the change of immune cell content over time is recorded as equation nn2; the malignant equation of the change of pathogen content over time is recorded as nn3; the malignant equation of the change of endocrine content over time is recorded as nn4; The weighted treatment time corresponding to the quasi-disease name is tzo, the proportion of benign lesions is ybo, and the proportion of malignant lesions is nbo; where tzo∈{tz （1） ~tz （bn）},ybo∈{yb （1） ~yb （bn）}, nbo∈{nb （1） ~nb （bn）}; Process B5: Substitute To, Imo, Pao, and Nfo as dependent variables, and reversely substitute them into equations yy1 to yy4 in turn to calculate the time when the patient develops benign lesions of body temperature, immune cells, pathogen content, and endocrine content, recorded as lt1, lt2, lt3, and lt4; To, Imo, Pao, and Nfo were used as dependent variables and substituted into equations nn1 to nn4 in reverse order to calculate the time when the patient developed malignant changes in body temperature, immune cells, pathogen content, and endocrine content, recorded as et1, et2, et3, and et4; The time the patient has been ill is recorded as yht; the calculation formula of yht is: .

[0014] Furthermore, the subsequent process of process B5 is as follows: Process B6: Calculate the maximum value of equation yy1 in the interval (yht, tzo], denoted as T mm ; Calculate the maximum value of equation yy2 in the interval (yht, tzo], denoted as Im mm ; Calculate the maximum value of equation yy3 in the interval (yht, tzo], denoted as Pa mm ; Calculate the maximum value of equation yy4 in the interval (yht, tzo], denoted as Nf mm ; Process B7: Calculate the maximum value of equation nn1 in the interval (yht, tzo], denoted as T lm ; Calculate the maximum value of equation nn2 in the interval (yht, tzo], denoted as Im lm ; Calculate the maximum value of equation nn3 in the interval (yht, tzo], denoted as Pa lm ; Calculate the maximum value of equation nn4 in the interval (yht, tzo], denoted as Nf lm ; Process B8: Get the current system time, recorded as dtt; analyze the patient's pathological changes based on the data in processes B6 to B7; Process B81: Get T mm The corresponding time in equation yy1 is denoted as aht1; get T lm The corresponding time in equation nn1 is denoted as aht2; Compare (ybo × T mm ) and (nbo×T lm ) to determine the patient's temperature change; If (ybo × T mm )≥(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht1), and the highest body temperature is T mm ; If (ybo × T mm )<(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht2), and the highest body temperature is T lm ; Process B82: Get lm mm The corresponding time in equation yy2 is denoted as bht1; get lm lm The corresponding time in equation nn2 is denoted as bht2; Compare (ybo × lm mm ) and (nbo×lm lm ) to determine the changes in the patient's immune cell content; If (ybo×lm mm )≥(nbo×lm lm ), it means that the patient will have the highest immune cell content at (dtt + bht1), and the highest immune cell content is lm mm ; If (ybo×lm mm )<(nbo×lm lm), it means that the patient will have the highest immune cell content at (dtt + bht2), and the highest immune cell content is lm lm ; Process B83: Repeat the same process of determining the patient's temperature changes and immune cell content changes to determine the patient's pathogen content changes and endocrine content changes; Process B9: Summarize the analysis results of processes B1 to B8 as the patient's pre-diagnosis report.

[0015] Furthermore, the specific process of process B1 is as follows: Process B11: Use the BERT model or the BioBERT model to process the text data in sample data A and convert it into the BERT input format to obtain the text features of each disease; Process B12: Reconstruct the layer structure of the CNN model to obtain the *CNN model; Flatten layer: Use the Flatten() function to flatten the multi-dimensional feature map output by the maximum pooling layer 3 into a one-dimensional array; Fully connected layer: Use the add() function and the Dense() function to add 128 neurons, and use the ReLU function as the activation function to learn high-level features; Dropout layer: Use regularization techniques to randomly drop 50% of neurons to reduce overfitting; Process B13: Using sample data B as input of the *CNN model to obtain image features of each disease; Using the cross-modal attention algorithm, the text features and image features of each disease are fused to obtain the fused features of each disease; Process B14: Use the Softmax function as the activation function and set the output layer of the disease diagnosis model; Output layer: The number of neurons output is equal to the number of different categories, and the output is converted into a probability distribution, outputting the name of the disease with the highest probability of the patient suffering from it; Process B15: Take sample data A and sample data B as combined sample data; back up the sample information in three copies, respectively as training set, test set and validation set; use the cross entropy loss function as the loss function, use the training set and test set to train the model constructed by processes B11 to B14, and use the validation set to test the model until the output value of the loss function is minimized, thus completing the construction of the disease diagnosis model.

[0016] Compared with the prior art, the present invention has the following beneficial effects: Early warning: By continuously monitoring and analyzing health data, the present invention enables the system to promptly detect potential health problems and provide individuals with early warnings, thereby avoiding or delaying the occurrence of diseases. At the same time, the present invention can provide doctors with treatment recommendations based on the evaluation results of health data, helping patients improve their health and improve treatment efficiency.

[0017] Comprehensive assessment: The health risk assessment of the present invention covers multiple factors including physiology, pathology and geography, which can fully assess the patient's physical health and provide a reference for the patient's treatment plan.

[0018] Simple application: The present invention can automatically obtain the medical record information of a certain area and the medical examination information of the patients, conduct health assessment and disease diagnosis on the patients, and obtain analysis results; through the evaluation results, the pathological changes of different patients can be predicted, thereby improving the treatment efficiency of the patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] 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: Figure 1 Schematic diagram of the health risk assessment system and (medical) digital intelligence robot of the present invention; Figure 2 is a schematic diagram of the health risk assessment system of the present invention; Figure 3 This is a schematic diagram of data processing of the present invention. DETAILED DESCRIPTION

[0020] 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.

[0021] See also Figure 1 ,A digital intelligent robot for health risk assessment based on big data includes: a health risk assessment system and a (medical) digital intelligent robot; See also Figure 2 and Figure 3 The health risk assessment system includes: an information acquisition module, a medical record analysis module, a patient diagnosis module, a physician interaction module, a database and a server; wherein the information acquisition module, the medical record analysis module, the patient diagnosis module and the physician interaction module are connected to the database and the server respectively; Among them, the database and server are used for information transmission and data calculation between the information acquisition module, medical record analysis module, patient diagnosis module and physician interaction module; (Medical) Digital Robot: Used to link the health risk assessment system, obtain the patient's medical examination report, extract pre-diagnosis information; register the patient according to the time of illness in the pre-diagnosis report; Information acquisition module: used to obtain the patient's medical examination report, extract the patient's disease pictures and symptoms, and obtain pre-diagnosis information; Medical record analysis module: used to obtain all medical record information of the target hospital as reference medical records; extract the disease name, illness time, recovery time, lesion picture, clinical symptoms, clinical medication and lesion index content in the reference medical records to obtain pathological data; According to the duration of illness, recovery time and lesion index content of each disease, the benign treatment cycle and malignant treatment cycle of each disease are calculated, and the benign lesion equation and malignant lesion equation of each disease are constructed; It should be noted that the “target hospital” in the present invention refers to: city and county hospitals that use the present invention (a digital intelligence robot for health risk assessment based on big data) to conduct health risk assessment; Process A: The workflow of the medical record analysis module is as follows: Process A1: Obtain all medical records of the target hospital as reference medical records; The disease names, illness duration, recovery time, lesion pictures, clinical symptoms, clinical medications and lesion index contents of all patients in the reference medical records were extracted to obtain pathological data; The content of pathological indicators includes: the patient's body temperature, immune cell content (in the blood), pathogen content and endocrine content (in the blood); Endocrines (in the blood) refer to: pathogen antibodies, insulin, thyroid hormones, dopamine, alanine aminotransferase, aspartate aminotransferase and other hormones or peptide substances secreted by human tissues or organs; Process A2: Count the number of all disease names in the reference medical records, recorded as bn; Count the number of all reference medical records, recorded as ln; Count the number of pathological data corresponding to the first disease (i.e. the disease corresponding to the first disease name), denoted as bl （1） ; Count the number of pathological data corresponding to the second disease (i.e. the disease corresponding to the second disease name), recorded as bl （2） ; Similarly, count the number of pathological data corresponding to the name of the bn-th disease (i.e. the disease corresponding to the bn-th disease name), recorded as bl （bn） ; Among them, bl （1） ~bl （bn） The sum of is ln; Process A3: Count the illness time of all patients in the reference medical records and obtain tf(1,1), tf(1,2)~tf(1,bl （1） )~tf(bn,bl （bn） ); (unit: day) The recovery time of all patients in the reference medical records is calculated to obtain te(1,1), te(1,2)~te(1,bl （1） )~te(bn,bl （bn） ); (unit: day) Among them, tf (1,1) and te (1,1) represent the time when the first patient with the first disease becomes ill and the time when the patient recovers, respectively; tf(1,2) and te(1,2) represent the illness time of the second patient suffering from the first disease and the patient recovery time, respectively; By analogy, tf(1,bl （1） ) and te (1,bl （1） ), respectively representing the bl （1） The duration of illness and recovery of patients with the first disease; tf(bn,bl （bn） ) and te(bn,bl （bn） ), respectively representing the bl （bn） The duration of illness of a patient with the bnth disease and the time of recovery; Process A4: Define calculation formula a1: tt （i，j） =te （i，j） -tf （i，j） ; Among them, tt （i，j） 、te （i，j） and tf （i，j） , respectively represent the illness time, recovery time and treatment cycle of the jth patient with the i-th disease; the value range of i is: 1~bn; The value range of j changes with i; if i is 1, the value range of j is: 1~bl （1） ; If i is 2, the value range of j is: 1~bl （2） ; Similarly, if i is bn, the value range of j is: 1~bl （bn） ; Set tf(1,1)~tf(bn,bl （bn） ) and te (1,1) ~ te (bn,bl （bn） ) into the calculation formula a1, we get tt(1,1), tt(1,2)~tt(1,bl （1） )~tt(bn,bl （bn） ); Among them, tt(1,1) represents the treatment cycle of the first patient with the first disease; tt(1,2) indicates the second treatment cycle of the patient with the first condition; By analogy, tt(1,bl （1） ) indicates the bl （1）treatment cycles for patients with condition 1; tt(bl （bn） ,bn) indicates the bl （bn） treatment cycles for patients with bn disease; According to tt (1,1) ~ tt (bn,bl （bn） ), calculate the weighted treatment time of the 1st to bnth diseases, and get tz （1） ~tz （bn） ; Process A41: According to tt (1,1) ~ tt (1,bl （1） ), calculate the weighted treatment time tz of the first disease （1） ; Extract tt(1,1)~tt(1,bl （1） ) is denoted as ttm, and the minimum value is denoted as ttl; calculate tt(1,1)~tt(1,bl （1） ), denoted as att; Process A42: At tt(1,1)~tt(1,bl （1） ), count the number of cells in the interval [ttl, att), denoted as aa1; count the number of cells in the interval [att, ttm), denoted as aa2; tz （1） The calculation formula is: ; Process A43: Repeated calculation of tz （1） The same process as above is used to calculate the weighted treatment time for the 2nd to bnth diseases, and we get tz （2） ~tz （bn） ; Process A5: Calculate the proportion of benign lesions and the proportion of malignant lesions for the first condition; Define the relationship a2: tt(1,i)≤tz （1） ; Where tt(1,i) represents the treatment cycle of the i-th patient with the first disease, and the value range of i is: 1~bl （1） ; Set tt(1,1)~tt(1,bl （1） ) into relational expression a2, and the patients who satisfy relational expression a2 are summarized as benign patients, and the number of benign patients is recorded as sn1. The patients who do not satisfy relational expression a2 are summarized as malignant patients, and the number of malignant patients is recorded as sn2. Among them, sn1 + sn2 = bl （1） ; Calculate the proportion of benign lesions in the first condition, denoted as yb （1） ;yb （1） The calculation formula is: =sn1 / bl（1） ; Calculate the proportion of benign lesions in the first condition, denoted as nb （1） ; nb （1） The calculation formula is: =sn2 / bl （1） ; Process A6: Repeat the process of calculating the proportion of benign lesions and the proportion of malignant lesions in the first disease, and calculate the proportion of benign lesions in the second to bnth diseases, and obtain: yb （2） ~yb （bn） ; The proportion of malignant lesions in the second to bnth diseases is: nb （2） ~nb （bn） ; Process A7: calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and constructing the benign lesion equation and the malignant lesion equation of the first disease; Process A71: benign patients with the first disease are classified as category a patients, and malignant patients with the first disease are classified as category b patients; The treatment cycles of the first to the sn1th patients of category a are recorded as: tyy(1,1)~tyy(1,sn1); the treatment cycles of the first to the sn2th patients of category b are recorded as: tnn(1,1)~tnn(1,sn2); Calculate the average value of tyy(1,1)~tyy(1,sn1) and record it as ty （1） , will ty （1） As a benign treatment cycle for the first condition; Calculate the average value of tnn(1,1)~tnn(1,sn1) and record it as tn （1） , will tn （1） As a vicious treatment cycle of the first disease; Process A72: The first patient of type a is taken as the target patient, and the time of illness of the target patient is recorded as st, and the time of recovery is recorded as ts; where st∈{tf(1,1)~tf(1,bl （1） )}, ts∈{te(1,1)~te(1,bl （1） )}; In the reference medical records, the temperature changes of the target patient in the time period of st to ts are extracted to obtain T （st） , T （st+1） ~T （ts） ; Changes in the content of immune cells (in the blood) to obtain Im （st） 、Im （st+1） ~Im （ts） ; Changes in pathogen content, get Pa （st） 、Pa （st+1） ~Pa （ts）; Changes in endocrine content (in blood) to obtain Nf （st） 、Nf （st+1） ~Nf （ts） ; Among them, T （st） 、Im （st） 、Pa （st） and Nf （st） , respectively represent the target patient's body temperature, immune cell content (in the blood), pathogen content, and endocrine content (in the blood) at st; T （st+1） 、Im （st+1） 、Pa （st+1） and Nf （st+1） , respectively represent the target patient’s body temperature, immune cell content (in the blood), pathogen content, and endocrine content (in the blood) at (st+1); By analogy, T （ts） 、Im （ts） 、Pa （ts） and Nf （ts） , respectively represent the target patient's body temperature, immune cell content (in the blood), pathogen content, and endocrine content (in the blood) at ts; Process A73: Extraction of T （st） ~T （ts） The maximum value is denoted as Tm, and the minimum value is denoted as Tl; Extract Im （st） ~Im （ts） The maximum value is recorded as Imm, and the minimum value is recorded as Iml; Extraction of Pa （st） ~Pa （ts） The maximum value is denoted as Pam, and the minimum value is denoted as Pal; Extraction of Nf （st） ~Nf （ts） The maximum value is denoted as Nfm, and the minimum value is denoted as Nfl; The highest order of body temperature is denoted as tx; the calculation formula for tx is: tx = log Tl (Tm); Calculate the highest order of immune cell content (in the blood), denoted as ix; the calculation formula for ix is: ix = log Iml (Imm); The highest order of pathogen content is calculated, denoted as px; the calculation formula for px is: px = log Pal (Pam); The highest order number for calculating the endocrine content (in the blood) is recorded as nx; the calculation formula for nx is: nx = log Nfl (Nfm); Among them, tx, ix, px and nx are all positive integers, rounded up; Process A74: Assume that a good equation for the change of body temperature with time is equation a1; determine a mathematical expression of equation a1; Process A741: The mathematical expression of equation a1 is: ; Where x represents the number of days of illness, i.e., the independent variable; T(x) represents the body temperature of the target patient on the xth day, i.e., the body temperature when the patient is confirmed to be ill; T(x+1) represents the body temperature of the target patient on the (x+1)th day, i.e., the dependent variable; α0 represents the parameter-free weight of equation a1, and the initial value of α0 is 0; x j represents the j-order term of the independent variable, α j It represents the parameter weight of the j-order term of the independent variable in equation a1, j represents the order of the independent variable, and the value range of j is: 1~tx; ! represents factorial; The expanded form of the equation T(x+1) is: ; Among them, α1~α tx Indicates the parameter weights of the first-order term to the tx-order term of the independent variable in equation a1, α1~α tx The initial value of is 1; Process A742: T （s） As the initial value, T （st+1） ~T （ts） Substituting tx into equation a1, we get equation group a1-1; The mathematical expression of equation group a1-1 is: ; Use MATLAB software to calculate the program group a1-1 and get α0~α tx The (estimated) value of tx ; Set yα0~yα tx Substituting this into the equation T(x+1), we get a good equation for the change of body temperature over time; Process A75: Assume that the good equation for the change of immune cell content (in blood) over time is equation b1; determine the mathematical expression of equation b1; Process A751: The mathematical expression of equation b1 is: ; Among them, y represents the number of days of illness, that is, the independent variable; Im(y) represents the immune cell content (in the blood) of the target patient on day y, that is, the immune cell content (in the blood) when the patient is confirmed to be ill; Im(y+1) represents the immune cell content (in the blood) of the target patient on day (y+1), that is, the dependent variable; β0 represents the parameter-free weight of equation b1, and the initial value of β0 is 0; yj represents the i-th order term of the independent variable in equation b1, β i It represents the parameterized weight of the term of the independent variable i, i represents the order of the independent variable, and the value range of i is: 1~tx; ! represents factorial; Process A752: Im （s） As the initial value, Im （st+1） ~Im （ts） Substituting tx into equation b1, we get equation group b1-1; The mathematical expression of equation group b1-1 is: ; Among them, β1~β tx Indicates the parameter weights of the first-order term of the independent variable of equation b1 to the tx-order term, β1~β tx The initial value of is 1; Use MATLAB software to calculate program group b1-1 and get β0~β tx The (estimated) value of tx ; Set yβ0~yβ tx Substituting into the equation Im(x+1), we get a good equation for the change of immune cell content (in the blood) over time; Process A76: Repeat the same process of calculating equation a1 and equation b1 to calculate the benign equation of pathogen content changing over time and the benign equation of endocrine content changing over time; The benign equations of the changes of body temperature over time, immune cell content (in the blood) over time, pathogen content over time and endocrine content (in the blood) over time are summarized as the benign lesion equation of the first disease; Process A77: Repeat the same process of constructing the benign lesion equation for the first condition to calculate the malignant lesion equation for the first condition; Process A8: Repeat the same process of calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and calculate the benign treatment cycle of the second to bnth diseases, recorded as ty （2） ~ty （bn） ; Calculate the malignant treatment cycle of the 2nd to bnth diseases, recorded as tn （2） ~tn （bn） ; The same process of constructing the benign lesion equation and the malignant lesion equation for the first disease is repeated to construct the benign lesion equation and the malignant lesion equation for the second to bnth diseases.

[0022] Patient diagnosis module: used to build a disease diagnosis model and a drug matching model based on the disease name, lesion pictures, clinical symptoms and clinical medications; substitute the pre-diagnosis information into the disease diagnosis model and the drug matching model to determine the patient's disease name and treatment drugs; obtain the patient's lesion index content as the parameter to be analyzed, and combine the benign lesion equation and malignant lesion equation corresponding to the disease name to analyze the patient's pathological changes and the duration of illness to obtain a pre-diagnosis report; Process B: The workflow of the patient diagnosis module is as follows: Process B1: In the reference medical records, the lesion images corresponding to each disease are summarized as sample data A, the clinical symptoms as sample data B, and the clinical medication as sample data C; According to sample data A and sample data B, a disease diagnosis model is constructed; Process B11: Use the BERT model or the BioBERT model (a variant of BERT in the medical field) to process the text data in sample data A (i.e., the BERT model processes the text content) and convert it into the BERT input format to obtain the text features of each disease; Process B12: Reconstruct the layer structure of the CNN model to obtain the *CNN model; Process B121: Add 3 convolutional layers: Convolutional layer 1: The number of filters is set to 32; the convolution kernel size is set to 3×3; Use ReLU (Rectified Linear Unit) as the activation function; The output shape is set to: 128×128 pixels, 3 color channels (RGB); Convolutional layer 2: Set the number of filters to 64; the rest of the settings are the same as convolutional layer 1; Convolutional layer 3: Set the number of filters to 128; the rest of the settings are the same as convolutional layer 1; Process B122: Add 3 max pooling layers: Max pooling layer 1: The pooling window is set to: 2×2; Max pooling layer 2: The settings are the same as max pooling layer 1; Max pooling layer 3: The settings are the same as max pooling layer 1; Process B123: Adding additional layers: Flatten layer: Use the Flatten() function to flatten the multi-dimensional feature map output by the maximum pooling layer 3 into a one-dimensional array; Fully connected layer: Use the add() function and the Dense() function to add 128 neurons, and use the ReLU function as the activation function to learn high-level features; Dropout layer: Use regularization techniques to randomly discard 50% of neurons to reduce overfitting; Process B13: Using sample data B as the input of the *CNN model (i.e., the *CNN model processes the image content) to obtain the image features of each disease; Use the Cross-Modal Attention algorithm to fuse the text features and image features of each disease to obtain the fused features of each disease; Process B14: Use the Softmax function as the activation function and set the output layer of the disease diagnosis model; Output layer: The number of neurons output is equal to the number of different categories (i.e. predicting the name of the patient's disease), and the output is converted into a probability distribution, outputting the name of the disease with the highest probability of the patient suffering from the disease; Process B15: Take sample data A and sample data B as combined sample data; back up three copies of the sample information, and use them as training set, test set and validation set respectively; use the cross entropy loss function as the loss function, use the training set and test set to train the model constructed by processes B11 to B14, and use the validation set to test the model until the output value of the loss function is minimized, thus completing the construction of the disease diagnosis model (disease name - lesion image - clinical symptoms); Process B2: Repeat the same process of building a disease diagnosis model, using sample data A and sample data C as combined sample data to build a (disease name - clinical symptoms - clinical medication) drug matching model; Substitute the pre-diagnosis information into the disease diagnosis model and drug matching model to determine the patient's disease name and treatment drug; Process B3: Obtain the weighted treatment time tz for the 1st to bnth diseases （1） ~tz （bn） , proportion of benign lesions （1） ~yb （bn） , proportion of malignant lesions nb （1） ~nb （bn） ; The patient's disease name is used as the quasi-disease name, and the benign lesion equation and the malignant lesion equation corresponding to the quasi-disease name are obtained; Obtain the patient's current body temperature as To, immune cells (in the blood) as Imo, pathogen content as Pao, and endocrine content (in the blood) as Nfo as parameters to be analyzed; Process B4: The benign equation of the change of body temperature over time in the benign lesion equation corresponding to the quasi-disease name is recorded as equation yy1; the benign equation of the change of immune cell content over time (in the blood) is recorded as equation yy2; the benign equation of the change of pathogen content over time is recorded as yy3; the benign equation of the change of endocrine content (in the blood) over time is recorded as yy4; The malignant equation of the change of body temperature over time in the malignant lesion equation corresponding to the quasi-disease name is recorded as equation nn1; the malignant equation of the change of immune cell content over time (in the blood) is recorded as equation nn2; the malignant equation of the change of pathogen content over time is recorded as nn3; the malignant equation of the change of endocrine content (in the blood) over time is recorded as nn4; The weighted treatment time corresponding to the quasi-disease name is tzo, the proportion of benign lesions is ybo, and the proportion of malignant lesions is nbo; where tzo∈{tz （1） ~tz （bn）},ybo∈{yb （1） ~yb （bn）}, nbo∈{nb （1） ~nb （bn）}; Process B5: Substitute To, Imo, Pao, and Nfo as dependent variables, and reversely substitute them into equations yy1 to yy4 in turn to calculate the time when the patient develops benign changes in body temperature, immune cells (in the blood), pathogen content, and endocrine content (in the blood), recorded as lt1, lt2, lt3, and lt4; Take To, Imo, Pao, and Nfo as dependent variables, and reversely substitute them into equations nn1 to nn4 in turn to calculate the time when the patient develops malignant changes in body temperature, immune cells (in the blood), pathogen content, and endocrine content (in the blood), which are recorded as et1, et2, et3, and et4; The time the patient has been ill is recorded as yht; the calculation formula of yht is: ; Process B6: Calculate the maximum value of equation yy1 in the interval (yht, tzo], denoted as T mm ; Calculate the maximum value of equation yy2 in the interval (yht, tzo], denoted as Im mm ; Calculate the maximum value of equation yy3 in the interval (yht, tzo], denoted as Pa mm ; Calculate the maximum value of equation yy4 in the interval (yht, tzo], denoted as Nf mm ; Process B7: Calculate the maximum value of equation nn1 in the interval (yht, tzo], denoted as Tlm ; Calculate the maximum value of equation nn2 in the interval (yht, tzo], denoted as Im lm ; Calculate the maximum value of equation nn3 in the interval (yht, tzo], denoted as Pa lm ; Calculate the maximum value of equation nn4 in the interval (yht, tzo], denoted as Nf lm ; Process B8: Get the current system time, recorded as dtt; analyze the patient's pathological changes based on the data in processes B6 to B7; Process B81: Get T mm The corresponding time in equation yy1 is denoted as aht1; get T lm The corresponding time in equation nn1 is denoted as aht2; Compare (ybo × T mm ) and (nbo×T lm ) to determine the patient's temperature change; If (ybo × T mm )≥(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht1), and the highest body temperature is T mm ; If (ybo × T mm )<(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht2), and the highest body temperature is T lm ; Process B82: Get lm mm The corresponding time in equation yy2 is denoted as bht1; get lm lm The corresponding time in equation nn2 is denoted as bht2; Compare (ybo × lm mm ) and (nbo×lm lm ) to determine changes in the patient's immune cell content (in the blood); If (ybo×lm mm )≥(nbo×lm lm ), it means that the patient will have the highest immune cell content (in the blood) at (dtt + bht1), and the highest immune cell content (in the blood) is lm mm ; If (ybo×lm mm )<(nbo×lm lm ), it means that the patient will have the highest immune cell content (in the blood) at (dtt + bht2), and the highest immune cell content (in the blood) is lmlm ; Process B83: Repeat the same process of determining changes in the patient's body temperature and changes in the immune cell content (in the blood) to determine changes in the patient's pathogen content and changes in the endocrine content (in the blood); Process B9: Summarize the analysis results of processes B1 to B8 as the patient's pre-diagnosis report.

[0023] Physician interaction module: used to provide feedback on pre-diagnosis reports and control the (medical) digital robot based on the pre-diagnosis reports.

[0024] It should be noted that due to the excessive number of physiological parameters in the human body, the label "lesion index content" displayed in the present invention is vaguely defined, and the specific definition of the pathological index content needs to be defined by the user or relevant technical personnel in actual use; In the process of using the present invention, users or relevant technical personnel can define or design more labels or indicators on their own to improve the accuracy of the health assessment and patient diagnosis of the present invention.

[0025] 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.

[0026] 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 digital robot for health risk assessment based on big data, characterized in that: The robot includes: a health risk assessment system and a digital intelligence robot; The health risk assessment system includes: information acquisition module: used to obtain the patient's medical examination report, extract the patient's disease pictures and symptoms, and obtain pre-diagnosis information; Medical record analysis module: used to obtain all medical record information of the target hospital as reference medical records; extract the disease name, illness time, recovery time, lesion picture, clinical symptoms, clinical medication and lesion index content in the reference medical records to obtain pathological data; According to the duration of illness, recovery time and lesion index content of each disease, the benign treatment cycle and malignant treatment cycle of each disease are calculated, and the benign lesion equation and malignant lesion equation of each disease are constructed; Patient diagnosis module: used to build a disease diagnosis model and a drug matching model based on the disease name, lesion pictures, clinical symptoms and clinical medications; substitute the pre-diagnosis information into the disease diagnosis model and the drug matching model to determine the patient's disease name and treatment drugs; obtain the patient's lesion index content as the parameter to be analyzed, and combine the benign lesion equation and malignant lesion equation corresponding to the disease name to analyze the patient's pathological changes and the duration of illness to obtain a pre-diagnosis report; Physician interaction module: used to provide feedback on pre-diagnosis reports and control the digital robot based on the pre-diagnosis reports; Digital robot: used to link the health risk assessment system, obtain the patient's medical examination report, extract pre-diagnosis information; register the patient based on the pre-diagnosis report.

2. According to claim 1, a digital robot for health risk assessment based on big data is characterized in that: The workflow of the medical record analysis module is as follows: Process A1: Obtain all medical records of the target hospital as reference medical records; The disease names, illness duration, recovery time, lesion pictures, clinical symptoms, clinical medications and lesion index contents of all patients in the reference medical records were extracted to obtain pathological data; The content of pathological indicators includes: the patient's body temperature, immune cell content, pathogen content and endocrine content; Process A2: Count the number of all disease names in the reference medical records, recorded as bn; Count the number of all reference medical records, recorded as ln; Count the number of pathological data corresponding to the first disease, denoted as bl （1） ; Similarly, count the number of pathological data corresponding to the bn-th disease name, denoted as bl （bn） ; Among them, bl （1） ~bl （bn） The sum of is ln; Process A3: Count the illness time of all patients in the reference medical records and obtain tf(1,1)~tf(bn,bl （bn） ); By counting the recovery time of all patients in the reference medical records, we can get te(1,1)~te(bn,bl （bn） ); Among them, tf (1,1) and te (1,1) represent the time when the first patient with the first disease becomes ill and the time when the patient recovers, respectively; By analogy, tf(bn,bl （bn） ) and te(bn,bl （bn） ), respectively representing the bl （bn） The duration of illness of a patient with the bnth disease and the time of recovery; Process A4: Define calculation formula a1: tt （i，j） =te （i，j） -tf （i，j） ; Among them, tt （i，j） 、te （i，j） and tf （i，j） , respectively represent the illness time, recovery time and treatment cycle of the jth patient with the i-th disease; the value range of i is: 1~bn; The value range of j changes with i; if i is 1, the value range of j is: 1~bl （1） ; Similarly, if i is bn, the value range of j is: 1~bl （bn） ; Set tf(1,1)~tf(bn,bl （bn） ) and te (1,1) ~ te (bn,bl （bn） ) into the calculation formula a1, we get tt(1,1)~tt(bn,bl （bn） ); Among them, tt(1,1) represents the treatment cycle of the first patient with the first disease; By analogy, tt (bl （bn） ,bn) indicates the bl （bn） treatment cycles for patients with bn disease; According to tt (1,1) ~ tt (bn,bl （bn） ), calculate the weighted treatment time of the 1st to bnth diseases, and get tz （1） ~tz （bn） .

3. A digital robot for health risk assessment based on big data according to claim 2, characterized in that: The subsequent process of process A4 is as follows: Process A5: Calculate the proportion of benign lesions and the proportion of malignant lesions for the first condition; Define the relationship a2: tt(1,i)≤tz （1） ; Where tt(1,i) represents the treatment cycle of the i-th patient with the first disease, and the value range of i is: 1~bl （1） ; Set tt(1,1)~tt(1,bl （1） ) into relational expression a2, and the patients who satisfy relational expression a2 are summarized as benign patients, and the number of benign patients is recorded as sn1. The patients who do not satisfy relational expression a2 are summarized as malignant patients, and the number of malignant patients is recorded as sn2. Among them, sn1 + sn2 = bl （1） ; Calculate the proportion of benign lesions in the first condition, denoted as yb （1） ;yb （1） The calculation formula is: =sn1 / bl （1） ; Calculate the proportion of benign lesions in the first condition, denoted as nb （1） ; nb （1） The calculation formula is: =sn2 / bl （1） ; Process A6: Repeat the process of calculating the proportion of benign lesions and the proportion of malignant lesions in the first disease, and calculate the proportion of benign lesions in the second to bnth diseases, and obtain: yb （2） ~yb （bn） ; The proportion of malignant lesions in the second to bnth diseases is: nb （2） ~nb （bn） ; Process A7: calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and constructing the benign lesion equation and the malignant lesion equation of the first disease; Process A8: Repeat the same process of calculating the benign treatment cycle and the malignant treatment cycle of the first disease, and calculate the benign treatment cycle of the second to bnth diseases, recorded as ty （2） ~ty （bn） ; Calculate the malignant treatment cycle of the 2nd to bnth diseases, recorded as tn （2） ~tn （bn） ; Repeat the same process of constructing the benign lesion equation and the malignant lesion equation for the first disease, construct the benign lesion equation and the malignant lesion equation for the second to bnth diseases, and enter the patient diagnosis module.

4. The digital robot for health risk assessment based on big data according to claim 2, characterized in that: The specific process of process A4 is as follows: Process A41: According to tt (1,1) ~ tt (1,bl （1） ), calculate the weighted treatment time tz of the first disease （1） ; Extract tt(1,1)~tt(1,bl （1） ) is denoted as ttm, and the minimum value is denoted as ttl; calculate tt(1,1)~tt(1,bl （1） ), denoted as att; Process A42: At tt(1,1)~tt(1,bl （1） ), count the number of cells in the interval [ttl, att), denoted as aa1; count the number of cells in the interval [att, ttm), denoted as aa2; tz （1） The calculation formula is: ； Process A43: Repeated calculation of tz （1） The same process as above is used to calculate the weighted treatment time for the 2nd to bnth diseases, and we get tz （2） ~tz （bn） .

5. The digital robot for health risk assessment based on big data according to claim 3 is characterized in that: The specific process of process A7 is as follows: Process A71: benign patients with the first disease are classified as category a patients, and malignant patients with the first disease are classified as category b patients; The treatment cycles of the first to the sn1th patients of category a are recorded as: tyy(1,1)~tyy(1,sn1); the treatment cycles of the first to the sn2th patients of category b are recorded as: tnn(1,1)~tnn(1,sn2); Calculate the average value of tyy(1,1)~tyy(1,sn1) and record it as ty （1） , will be （1） As a benign treatment cycle for the first condition; Calculate the average value of tnn(1,1)~tnn(1,sn1) and record it as tn （1） , will tn （1） As a vicious treatment cycle of the first disease; Process A72: The first patient of type a is taken as the target patient, and the time of illness of the target patient is recorded as st, and the time of recovery is recorded as ts; where st∈{tf(1,1)~tf(1,bl （1） )}, ts∈{te(1,1)~te(1,bl （1） )}; In the reference medical records, the temperature changes of the target patient in the time period of st to ts are extracted to obtain T （st） , T （st+1） ~T （ts） ; Changes in immune cell content, get Im （st） 、Im （st+1） ~Im （ts） ; Changes in pathogen content, get Pa （st） 、Pa （st+1） ~Pa （ts） ; Changes in endocrine content, obtain Nf （st） 、Nf （st+1） ~Nf （ts） ; Among them, T （st） 、Im （st） 、Pa （st） and Nf （st） , respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at st; T （st+1） 、Im （st+1） 、Pa （st+1） and Nf （st+1） , respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at (st+1); By analogy, T （ts） 、Im （ts） 、Pa （ts） and Nf （ts） , respectively represent the target patient’s body temperature, immune cell content, pathogen content, and endocrine content at ts; Process A73: Extraction of T （st） ~T （ts） The maximum value is denoted as Tm, and the minimum value is denoted as Tl; Extract Im （st） ~Im （ts） The maximum value is recorded as Imm, and the minimum value is recorded as Iml; Extraction of Pa （st） ~Pa （ts） The maximum value is denoted as Pam, and the minimum value is denoted as Pal; Extraction of Nf （st） ~Nf （ts） The maximum value is denoted as Nfm, and the minimum value is denoted as Nfl; The highest order of body temperature is denoted as tx; the calculation formula for tx is: tx = log Tl (Tm); The highest order of calculating immune cell content is recorded as ix; the calculation formula of ix is: ix = log Iml (Imm); The highest order of pathogen content is calculated, denoted as px; the calculation formula for px is: px = log Pal (Pam); The highest order number for calculating the endocrine content is recorded as nx; the calculation formula for nx is: nx = log Nfl (Nfm); Among them, tx, ix, px and nx are all positive integers, rounded up; Process A74: Assume that a good equation for the change of body temperature with time is equation a1; determine a mathematical expression of equation a1; Process A75: Assume that a good equation for the change of immune cell content over time is equation b1; determine a mathematical expression of equation b1; Process A76: Repeat the same process of calculating equation a1 and equation b1 to calculate the benign equation of pathogen content changing over time and the benign equation of endocrine content changing over time; The benign equations of body temperature changing with time, immune cell content changing with time, pathogen content changing with time, and endocrine content changing with time are summarized as the benign lesion equation of the first disease; Process A77: Repeat the same process of constructing the benign lesion equation for the first condition to calculate the malignant lesion equation for the first condition.

6. The digital robot for health risk assessment based on big data according to claim 5, characterized in that: The specific process of process A74 is as follows: Process A741: The mathematical expression of equation a1 is: ; Where x represents the number of days of illness, i.e., the independent variable; T(x) represents the body temperature of the target patient on the xth day, i.e., the body temperature when the patient is confirmed to be ill; T(x+1) represents the body temperature of the target patient on the (x+1)th day, i.e., the dependent variable; α0 represents the parameter-free weight of equation a1, and the initial value of α0 is 0; x j represents the j-order term of the independent variable, α j It represents the parameter weight of the j-order term of the independent variable in equation a1, j represents the order of the independent variable, and the value range of j is: 1~tx; ! represents factorial; The expanded form of the equation T(x+1) is: ; Among them, α1~α tx Indicates the parameter weights of the first-order term to the tx-order term of the independent variable in equation a1, α1~α tx The initial value of is 1; Process A742: T （s） As the initial value, T （st+1） ~T （ts） Substituting tx into equation a1, we get equation group a1-1; The mathematical expression of equation group a1-1 is: ； Use MATLAB software to calculate the program group a1-1 and get α0~α tx The value of , denoted as yα0~yα tx ; Set yα0~yα tx Substituting this into the equation T(x+1), we get a good equation for how body temperature changes over time.

7. The digital intelligent robot for health risk assessment based on big data according to claim 5, characterized in that: The specific process of process A75 is as follows: Process A751: The mathematical expression of equation b1 is: ; Among them, y represents the number of days of illness, that is, the independent variable; Im(y) represents the immune cell content of the target patient on day y, that is, the immune cell content when the patient is confirmed to be ill; Im(y+1) represents the immune cell content of the target patient on day (y+1), that is, the dependent variable; β0 represents the parameter-free weight of equation b1, and the initial value of β0 is 0; y j represents the i-th order term of the independent variable in equation b1, β i It represents the parameterized weight of the term of the independent variable i, i represents the order of the independent variable, and the value range of i is: 1~tx; ! represents factorial; Process A752: Im （s） As the initial value, Im （st+1） ~Im （ts） Substituting tx into equation b1, we get equation group b1-1; The mathematical expression of equation group b1-1 is: ； Among them, β1~β tx Indicates the parameter weights of the first-order term of the independent variable of equation b1 to the tx-order term, β1~β tx The initial value of is 1; Use MATLAB software to calculate program group b1-1 and get β0~β tx The value of , denoted as yβ0~yβ tx ; Set yβ0~yβ tx Substituting into the equation Im(x+1), we obtain a good equation for the change of immune cell content over time.

8. The digital robot for health risk assessment based on big data according to claim 3, characterized in that: The workflow of the patient diagnosis module is as follows: Process B1: In the reference medical records, the lesion images corresponding to each disease are summarized as sample data A, the clinical symptoms as sample data B, and the clinical medication as sample data C; According to sample data A and sample data B, a disease diagnosis model is constructed; Process B2: Repeat the same process of building a disease diagnosis model, using sample data A and sample data C as combined sample data to build a drug matching model; Substitute the pre-diagnosis information into the disease diagnosis model and drug matching model to determine the patient's disease name and treatment drug; Process B3: Obtain the weighted treatment time tz for the 1st to bnth diseases （1） ~tz （bn） , proportion of benign lesions （1） ~yb （bn） , proportion of malignant lesions nb （1） ~nb （bn） ; The patient's disease name is used as the quasi-disease name, and the benign lesion equation and the malignant lesion equation corresponding to the quasi-disease name are obtained; Obtain the patient's current body temperature as To, immune cells as Imo, pathogen content as Pao, and endocrine content as Nfo as parameters to be analyzed; Process B4: The benign equation of the change of body temperature over time in the benign lesion equation corresponding to the quasi-disease name is recorded as equation yy1; the benign equation of the change of immune cell content over time is recorded as equation yy2; the benign equation of the change of pathogen content over time is recorded as yy3; the benign equation of the change of endocrine content over time is recorded as yy4; The malignant equation of the change of body temperature over time in the malignant lesion equation corresponding to the quasi-disease name is recorded as equation nn1; the malignant equation of the change of immune cell content over time is recorded as equation nn2; the malignant equation of the change of pathogen content over time is recorded as nn3; the malignant equation of the change of endocrine content over time is recorded as nn4; The weighted treatment time corresponding to the quasi-disease name is tzo, the proportion of benign lesions is ybo, and the proportion of malignant lesions is nbo; where tzo∈{tz （1） ~tz （bn） },ybo∈{yb （1） ~yb （bn） }, nbo∈{nb （1） ~nb （bn） }; Process B5: Substitute To, Imo, Pao, and Nfo as dependent variables, and reversely substitute them into equations yy1 to yy4 in turn to calculate the time when the patient develops benign lesions of body temperature, immune cells, pathogen content, and endocrine content, recorded as lt1, lt2, lt3, and lt4; To, Imo, Pao, and Nfo were used as dependent variables and substituted into equations nn1 to nn4 in reverse order to calculate the time when the patient developed malignant changes in body temperature, immune cells, pathogen content, and endocrine content, recorded as et1, et2, et3, and et4; The time the patient has been ill is recorded as yht; the calculation formula of yht is: 。 9. The digital robot for health risk assessment based on big data according to claim 8, characterized in that: The subsequent process of process B5 is as follows: Process B6: Calculate the maximum value of equation yy1 in the interval (yht, tzo], denoted as T mm ; Calculate the maximum value of equation yy2 in the interval (yht, tzo], denoted as Im mm ; Calculate the maximum value of equation yy3 in the interval (yht, tzo], denoted as Pa mm ; Calculate the maximum value of equation yy4 in the interval (yht, tzo], denoted as Nf mm ; Process B7: Calculate the maximum value of equation nn1 in the interval (yht, tzo], denoted as T lm ; Calculate the maximum value of equation nn2 in the interval (yht, tzo], denoted as Im lm ; Calculate the maximum value of equation nn3 in the interval (yht, tzo], denoted as Pa lm ; Calculate the maximum value of equation nn4 in the interval (yht, tzo], denoted as Nf lm ; Process B8: Get the current system time, recorded as dtt; analyze the patient's pathological changes based on the data in processes B6 to B7; Process B81: Get T mm The corresponding time in equation yy1 is denoted as aht1; get T lm The corresponding time in equation nn1 is denoted as aht2; Compare (ybo × T mm ) and (nbo×T lm ) to determine the patient's temperature change; If (ybo × T mm )≥(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht1), and the highest body temperature is T mm ; If (ybo × T mm )<(nbo×T lm ), it means that the patient's body temperature will be the highest at (dtt + aht2), and the highest body temperature is T lm ; Process B82: Get lm mm The corresponding time in equation yy2 is denoted as bht1; get lm lm The corresponding time in equation nn2 is denoted as bht2; Compare (ybo × lm mm ) and (nbo×lm lm ) to determine the changes in the patient's immune cell content; If (ybo×lm mm )≥(nbo×lm lm ), it means that the patient will have the highest immune cell content at (dtt + bht1), and the highest immune cell content is lm mm ; If (ybo×lm mm )<(nbo×lm lm ), it means that the patient will have the highest immune cell content at (dtt + bht2), and the highest immune cell content is lm lm ; Process B83: Repeat the same process of determining the patient's temperature changes and immune cell content changes to determine the patient's pathogen content changes and endocrine content changes; Process B9: Summarize the analysis results of processes B1 to B8 as the patient's pre-diagnosis report.

10. The digital robot for health risk assessment based on big data according to claim 8, characterized in that: The specific process of process B1 is as follows: Process B11: Use the BERT model or the BioBERT model to process the text data in sample data A and convert it into the BERT input format to obtain the text features of each disease; Process B12: Reconstruct the layer structure of the CNN model to obtain the *CNN model; Flatten layer: Use the Flatten() function to flatten the multi-dimensional feature map output by the maximum pooling layer 3 into a one-dimensional array; Fully connected layer: Use the add() function and the Dense() function to add 128 neurons, and use the ReLU function as the activation function to learn high-level features; Dropout layer: Use regularization techniques to randomly drop 50% of neurons to reduce overfitting; Process B13: Using sample data B as input of the *CNN model to obtain image features of each disease; Using the cross-modal attention algorithm, the text features and image features of each disease are fused to obtain the fused features of each disease; Process B14: Use the Softmax function as the activation function and set the output layer of the disease diagnosis model; Output layer: The number of neurons output is equal to the number of different categories, and the output is converted into a probability distribution, outputting the name of the disease with the highest probability of the patient suffering from it; Process B15: Take sample data A and sample data B as combined sample data; back up the sample information in three copies, respectively as training set, test set and validation set; use the cross entropy loss function as the loss function, use the training set and test set to train the model constructed by processes B11 to B14, and use the validation set to test the model until the output value of the loss function is minimized, thus completing the construction of the disease diagnosis model.

Images