Uka index-based uncertainty measurement method in traditional Chinese medicine diagnosis and treatment process

Abstract

The invention provides a method for measuring uncertainty in a traditional Chinese medicine diagnosis and treatment process based on a Wuka index, and the method comprises the following steps: constructing a Wuka auxiliary diagnosis and treatment model, in the formula, wV + wU + wC + wA = 1, V is a diagnosis and treatment mutability index value, U is a diagnosis and treatment information uncertainty index value, C is a diagnosis and treatment complexity index value, and A is a diagnosis and treatment fuzziness index value; obtaining diagnosis information, a treatment scheme and daily monitoring data of the patient; acquiring a diagnosis and treatment variability index value, a diagnosis and treatment information uncertainty index value, a diagnosis and treatment complexity index value and a diagnosis and treatment fuzziness index value according to the diagnosis information, the treatment scheme and the daily monitoring data of the patient; and obtaining a comprehensive Wuka index of the patient according to the Wuka auxiliary diagnosis and treatment model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traditional Chinese medicine auxiliary diagnosis and treatment, and in particular relates to a method for measuring uncertainty in traditional Chinese medicine diagnosis and treatment based on the Uka index. BACKGROUND

[0002] In the field of modern medicine, traditional Chinese medicine diagnosis and treatment faces many challenges and opportunities. The modernization of traditional Chinese medicine theory, as a key prerequisite for the modernization of traditional Chinese medicine, is crucial. However, the environment in which traditional Chinese medicine is practiced is becoming increasingly complex and variable, exhibiting significant VUCA characteristics, namely variability, uncertainty, complexity, and ambiguity, which presents many difficulties for traditional Chinese medicine diagnosis and treatment.

[0003] 1. Variability

[0004] The variability of traditional Chinese medicine is mainly reflected in its core treatment principle, syndrome differentiation. The occurrence and development of diseases have dynamic change characteristics, and the disease condition may have both acute onset and chronic development trends. In the process of syndrome differentiation, doctors need to consider the severity of the disease and adjust the treatment strategy flexibly according to the specific symptoms of the patient and their changes. However, this dynamic change makes it difficult to accurately grasp the disease condition and develop a precise treatment plan.

[0005] 2. Uncertainty

[0006] Uncertainty in traditional Chinese medicine is reflected in many aspects. From the perspective of the object of knowledge, the physiological and pathological phenomena of the human body are complex and diverse, making it difficult to completely and accurately grasp them. In terms of diagnostic methods, although the traditional four diagnostic methods of observation, listening, questioning, and palpation are rich in experience, the information obtained is mostly qualitative description and is highly subjective. In terms of conceptual language, traditional Chinese medicine terminology is relatively abstract and vague, lacking precise quantitative standards. Although traditional Chinese medicine uses "pattern" reasoning and syndrome differentiation to grasp the rules, when faced with a large number of uncertain symptoms, more precise methods are needed to improve the certainty of diagnosis and treatment.

[0007] 3. Complexity

[0008] Complexity is another significant feature of traditional Chinese medicine. Traditional Chinese medicine believes that the human body is an organic whole, with organs and functions interacting and influencing each other, forming a complex network relationship. Any local lesion can trigger a systemic reaction, and various factors need to be considered when treating. At the same time, the internal elements of the core diagnosis and treatment framework of traditional Chinese medicine (principle, method, prescription, and medicine) are closely connected and interact with each other, making the development and optimization of treatment plans extremely complex.

[0009] 4. Ambiguity

[0010] Fuzziness is obvious in TCM diagnosis. Traditional Chinese medicine relies on looking, listening, asking and cutting to collect information, and these information are mostly qualitative description, such as "pale" and "red" of the face, "string" and "smooth" of the pulse, etc., which are difficult to quantify accurately. Although modern fuzzy mathematics has certain application in TCM diagnosis, it still cannot completely solve the problem of fuzziness. This fuzziness may lead to differences in doctors' judgment of the disease, and then affect the accuracy and consistency of treatment.

[0011] In view of these problems, current research mainly focuses on the separate analysis of each link of TCM diagnosis and treatment, and there is lack of method for comprehensively considering these VUCA characteristics from the whole. VUCA theory provides a new idea for analyzing TCM environment, but the four characteristics of VUCA have not been fully studied.

[0012] In order to solve the above problems, people have been seeking an ideal technical solution. SUMMARY

[0013] The purpose of the present application is to introduce VUCA theory into TCM diagnosis and treatment research, construct an uncertainty measurement model based on VUCA index, and quantify the uncertainty in TCM diagnosis and treatment process, so as to provide a scientific evaluation basis for TCM diagnosis and treatment, and fill this gap, assist doctors in decision-making, optimize resource allocation, and promote the modernization development of TCM.

[0014] In order to achieve the above purpose, the technical scheme adopted by the present application is to provide a VUCA index-based uncertainty measurement method in TCM diagnosis and treatment process, comprising the following steps:

[0015] A VUCA auxiliary diagnosis and treatment model is constructed, wherein the VUCA auxiliary diagnosis and treatment model comprises:

[0016] VUCA = w V ·V + w U ·U + w C ·C + w A ·A

[0017] In the formula, w V , w U , w C , w A are weights determined by machine learning, and w V +w U +w C +w A = 1, V is the value of diagnosis and treatment variability index, U is the value of diagnosis and treatment information uncertainty index, C is the value of diagnosis and treatment complexity index, and A is the value of diagnosis and treatment fuzziness index.

[0018] obtain diagnosis information, treatment scheme and daily monitoring data of the patient, obtain diagnosis and treatment variability index value, diagnosis and treatment information uncertainty index value, diagnosis and treatment complexity index value and diagnosis and treatment ambiguity index value according to the diagnosis information, treatment scheme and daily monitoring data of the patient;

[0019] obtain the comprehensive UCA index of the patient according to the UCA auxiliary diagnosis and treatment model, and the comprehensive UCA index is the uncertainty value.

[0020] The present application has outstanding substantial characteristics and significant progress compared with the prior art, specifically, the whole scheme quantifies the diagnosis information, treatment scheme and daily monitoring data of the patient through variability, uncertainty, complexity and ambiguity four dimensions, provides a systematic and scientific quantitative evaluation method for traditional Chinese medicine diagnosis and treatment, makes up for the lack of accuracy of experience judgment, opens up a new perspective and method system for traditional Chinese medicine theory research, further deepens the understanding of complex phenomena in the process of traditional Chinese medicine diagnosis and treatment at the theoretical level, and enriches the connotation of traditional Chinese medicine theory; in practice, it provides a more operational and scientific method for traditional Chinese medicine diagnosis and treatment, which helps to improve the clinical treatment effect of traditional Chinese medicine and promote the modernization development process of traditional Chinese medicine. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a flowchart of the present application. DETAILED DESCRIPTION

[0022] The technical scheme of the present application will be further described in detail through specific embodiments.

[0023] The present embodiment provides a UCA index-based uncertainty measurement method in the process of traditional Chinese medicine diagnosis and treatment, as shown in Figure 1 , comprising the following steps:

[0024] constructing a UCA auxiliary diagnosis and treatment model, wherein the UCA auxiliary diagnosis and treatment model:

[0025] VUCA=w V ·V+w U ·U+w C ·C+w A ·A

[0026] In the formula, w V , w U , w C , w A are weights determined through machine learning, and w V +w U +w C +w A =1, V is the diagnosis and treatment variability index value, U is the diagnosis and treatment information uncertainty index value, C is the diagnosis and treatment complexity index value, and A is the diagnosis and treatment ambiguity index value.

[0027] Obtain the patient's diagnostic information, treatment plan, and daily monitoring data, and obtain the diagnosis and treatment variability index value, diagnosis and treatment information uncertainty index value, diagnosis and treatment complexity index value, and diagnosis and treatment ambiguity index value based on the patient's diagnostic information, treatment plan, and daily monitoring data;

[0028] The patient's comprehensive VUCA index is obtained according to the VUCA-assisted diagnosis and treatment model, and the comprehensive VUCA index is an uncertain value.

[0029] Specifically, the patient's daily monitoring data, as well as each diagnostic information and treatment plan, will be uploaded and updated to the patient case database. This application needs to extract data from the patient case database for calculation.

[0030] Daily monitoring data includes: general patient information and specific physiological parameters. General information includes basic information such as patient gender, age, blood pressure, and heart rate; specific physiological parameters include laboratory indicators such as blood routine, blood sugar, glycosylated hemoglobin, and liver and kidney function. Daily monitoring data also includes the patient's medication records and the daily disease severity score determined according to the CRF scale developed by professional doctors.

[0031] Diagnostic information includes the four diagnostic information, such as observation, auscultation, questioning and palpation, and also includes the diagnosis results of multiple doctors on the same four diagnostic information of the patient.

[0032] The treatment plan includes the overall drug treatment plan, including the types and quantity of drugs and the frequency of use; it is understandable that there are multiple treatment plans, and each treatment plan is scored by multiple doctors, but ultimately only one optimal treatment plan is selected to treat the patient.

[0033] Specifically, the variability index values ​​are obtained based on the patient's diagnostic information, treatment plan, and daily monitoring data, including:

[0034] 1.1 Based on daily monitoring data, obtain the patient's condition change rate index.

[0035] Specifically, the case record form (CRF) prepared by a professional doctor records the patient's daily monitoring data. The doctor gives a severity score of the disease based on the patient's daily monitoring data and uploads it to the patient's case database.

[0036] Calculate the disease change rate index based on the patient's daily disease severity score. By comparing it with the preset benchmark value, the stability of the disease is judged. If it is lower than the benchmark value, the disease is stable; if it is higher than the benchmark value, the disease fluctuates greatly and there is a risk of deterioration. Suppose the patient is divided into time t1, t2, ..., t n测 The severity scores of the disease are S(t1), S(t2), ..., S(t n测 ), calculate the patient's disease change rate index RMSD.

[0037] The following is announced:

[0038]

[0039] In the formula, is the average value of the disease score; S(t i is the disease score at time point t i ; n 测 is the number of measurements; i = 1, 2, …, n.

[0040] 1.2 According to the daily monitoring data and historical monitoring data, the treatment response change index of the patient is obtained.

[0041] Specifically, the relative change rate is introduced to measure the treatment effect. By comparing the disease scores before and after treatment, the treatment response change index is calculated. A negative value indicates improvement in symptoms, and the larger the absolute value, the better the treatment effect; a positive value indicates disease deterioration, and the larger the value, the more serious the deterioration.

[0042] The treatment response change index RCR of the patient is calculated.

[0043] The following is announced:

[0044]

[0045] In the formula, S 开始 is the initial disease score at the beginning of treatment; S 结束 is the final disease score after a treatment period T.

[0046] 1.3 According to the daily monitoring data and historical monitoring data, the laboratory index of the patient is obtained.

[0047] Specifically, the patient's daily specific physiological parameters are recorded, and the coefficient of variation and the change in the coefficient of variation before and after treatment are calculated. If the coefficient of variation decreases after treatment, it indicates that the stability of the physiological parameter has improved and the patient's health condition is good; otherwise, the stability decreases, which may indicate a health problem.

[0048] The average value μ L of the patient's laboratory index is calculated as follows:

[0049]

[0050] The standard deviation σ L of the laboratory index is calculated as follows:

[0051]

[0052] The coefficient of variation CV is:

[0053]

[0054] The coefficient of variation change ΔCV:

[0055] ΔCV = CV 治疗前 -CV 治疗后 (7)

[0056] In the formula, L i is the laboratory index value of the i-th measurement; n 测 is the number of measurements; i = 1, 2,..., n 测 ; CV 治疗前 is the coefficient of variation before treatment; and CV 治疗后 is the coefficient of variation after treatment.

[0057] 1.4 Obtain the diagnosis and treatment variability index value according to the illness change rate index, the treatment response change index, and the laboratory index.

[0058] Specifically, the comprehensive variability index is obtained by weighted averaging each sub-variability index, and the value thereof reflects the comprehensive variability of the illness, the treatment response, and the laboratory index, thereby providing a basis for evaluating the health status of the patient.

[0059] The comprehensive variability index V can be obtained by weighted averaging each sub-variability index.

[0060]

[0061] In the formula

[0062] Specifically, the diagnosis and treatment information uncertainty index value is obtained according to the diagnosis information and the treatment plan of the patient, including:

[0063] 2.1 Obtain the diagnosis information integrity index according to the completeness of the four diagnostic information of the patient.

[0064] Specifically, the completeness of the four diagnostic information indicators of the patient is checked, the diagnosis information integrity index is calculated, and the accuracy of diagnosis is affected.

[0065]

[0066] In the formula, U1 is the diagnosis information integrity index, n 实 is the number of four diagnostic information items actually filled in by the doctor according to the illness; and n 总 is the total number of information items in the medical record.

[0067] 2.2 Determine the symptom description keywords according to the symptom description document of the patient, process the symptom description keywords using the TF-IDF algorithm, and obtain the symptom description fuzziness index.

[0068] Specifically, we extract symptom description keywords, calculate keyword values ​​using the TF-IDF algorithm, and take the average value as an indicator of description clarity to evaluate uncertainty.

[0069]

[0070] Where U2 is the clarity index of the patient's symptom description, TFIDF j is the TFIDF value of the jth keyword, j = 1, 2, ..., n 总 , n j is the number of times keyword j appears in this document; n 词 The total number of words in this document; n 档 is the total number of documents; df j is the number of documents containing keyword j; n 总 is the number of keywords.

[0071] 2.3 Based on the number of treatment options, a logarithmic function is used to calculate the treatment option index. As you can understand, the more options there are, the higher the uncertainty. The following is the public statement:

[0072] U3=ln n 案 (12)

[0073] Where U3 is the selectivity index of the treatment plan, n 案 is the number of treatment options.

[0074] It should be noted that all treatment plans are provided by professional doctors.

[0075] 2.4. Convert the symptom frequency into a probability problem and calculate the entropy value of each symptom indicator. According to the entropy value of each symptom indicator, the uncertainty index of the diagnosis process is obtained.

[0076] Specifically, information entropy is used to represent the uncertainty of the diagnostic process. The frequency of symptoms is converted into a probability problem. The entropy value of each indicator is calculated, and the weighted average is used to obtain the uncertainty index of the diagnostic process.

[0077] Probability distribution p={p1,p2,...,p n症}, where p f represents the probability of disease f occurring, n 症 is the number of symptoms. At this time, the uncertainty of the diagnosis process is expressed by information entropy, which is as follows:

[0078]

[0079] Where U 4f is the uncertainty index of different indicators in the diagnosis process of a single patient, p f is the probability of disease f occurring, f=1,2,...,n 症Specifically, the probability of each disease occurrence is determined by the frequency of a large number of cases.

[0080] The entropy values of various aspects are integrated to obtain the uncertainty index of the diagnosis process of a single patient. A weighted average is used,

[0081]

[0082] In the formula, U4 refers to the uncertainty index of the diagnosis process of a single patient; w 4f is the importance weight of the fth index, U 4f is the entropy value of the fth index, f = 1, 2,..., n 症 .

[0083] 2.5 Obtain the comprehensive uncertainty index value according to the diagnosis information integrity index, the symptom description fuzziness index, the treatment scheme selectability index, and the diagnosis process uncertainty index.

[0084] Specifically, the relative importance weights of various aspects are determined, and the uncertainty index is obtained by weighted summation to evaluate the information quality and assist the doctor in decision-making. The relative importance of the four aspects is set as the weight w U1 , w U2 , w U3 , w U4 .

[0085] The above formula is integrated to obtain the uncertainty index U as follows:

[0086]

[0087] In the formula, w U1 +w U2 +w U3 +w U4 = 1.

[0088] Specifically, the diagnosis and treatment fuzziness index value is obtained according to the diagnosis information and treatment scheme of the patient, including:

[0089] 3.1 Quantitatively evaluate the consistency of diagnosis results of multiple doctors on the same case using Fleiss' Kappa coefficient, obtain the doctor diagnosis consistency index, eliminate the influence of random consistency, and judge the degree of diagnosis fuzziness.

[0090] The calculation formula of Fleiss' Kappa coefficient K is as follows,

[0091]

[0092] Observe the overall consistency That is, the actual proportion of consistent diagnosis results of doctors, which is shown as follows:

[0093]

[0094] Expected consistency That is, the consistency ratio when assuming that doctors make diagnoses randomly is as follows:

[0095]

[0096] The proportion p of each diagnostic category q for,

[0097]

[0098] Where n 病 is the number of cases with consistent diagnosis among all doctors, that is, the number of cases in which all doctors have the same diagnosis for a disease; p q is the proportion of each diagnostic category, that is, the frequency with which each diagnostic category is selected in the entire sample; n 诊 is the total number of diagnostic categories; n q is the total number of cases diagnosed as category q; q = 1, 2, ..., n 诊 ; N is the total number of cases; N 医 is the total number of physicians for whom the case was evaluated.

[0099] 3.2 Calculate the compliance coefficient based on the patient's medication and follow-up records.

[0100] Specifically, based on the patient's medication and follow-up records, the compliance coefficient is calculated to evaluate the patient's compliance with the treatment plan and affect the treatment effect.

[0101] The calculation formula of patient compliance coefficient M is as follows:

[0102]

[0103] Where m 治疗 The number of times the patient actually took the medication as prescribed during treatment; n 服药 The number of times the medicine should be taken as prescribed by the doctor; m 复诊 The number of follow-up visits for patients on time; n 复诊 The number of follow-up visits prescribed by the doctor.

[0104] 3.3 Based on the doctors' scores on the treatment plans, the treatment plan disagreement coefficient is obtained.

[0105] Specifically, based on the doctors' scores on the treatment plans, the mean square error formula is used to calculate the degree of disagreement and quantify the disagreement on the treatment plans. In practice, one doctor can provide a treatment plan, and other doctors can give corresponding scores to the treatment plan.

[0106] The formula for the divergence coefficient B of the treatment plan is as follows:

[0107]

[0108] where n 建 is the total number of optional treatment plans; n 医 is the total number of doctors participating in scoring; d rs is the score of the s-th doctor on the r-th treatment plan; is the average score of all doctors on the r-th treatment plan; r = 1, 2,..., n 建 ; s = 1, 2,..., n 医 .

[0109] 3.4 Obtain the diagnosis-treatment ambiguity index value according to the doctor diagnosis consistency index, the compliance coefficient, and the treatment plan divergence coefficient.

[0110] Sum the above indexes by weight to obtain the comprehensive ambiguity index. Assuming that the weights of the indexes are w a1 , w a2 , and w a3 respectively, the comprehensive ambiguity score A can be expressed as:

[0111]

[0112] where w a1 + w a2 + w a3 = 1.

[0113] Specifically, obtain the diagnosis-treatment complexity index value according to the patient's diagnosis information and treatment plan, including:

[0114] 4.1 Obtain the symptom complexity index according to the mutual information between symptoms.

[0115] Specifically, calculate the mutual information between symptoms and sum and normalize to obtain the symptom complexity index, which reflects the syndrome feature and its associated complexity. The greater the mutual information value, the closer the symptom association. According to the deviation degree of probability, the association strength between two symptoms X and Y is measured, and the mutual information formula is as follows:

[0116]

[0117] where X and Y are two random variables; x and y are specific values of random variables; p(x, y) is the joint probability distribution function of x and y, which is obtained by dividing the number of cases with both x and y symptoms by the total number of cases; p(x) and p(y) are the marginal probability distribution functions of x and y respectively. Assuming that X1, X2,..., X n症 represent n different symptoms, and n represents the number of symptoms.

[0118] The complexity index C1 between symptoms is obtained by summing and normalizing the mutual information between all possible pairs of symptoms.

[0119]

[0120] where I(X c ,X d ) represents the mutual information between the cth and dth symptoms; c < d.

[0121] 4.2 Complexity of treatment regimen

[0122] It is noted that the optimal treatment regimen is determined by obtaining the treatment regimen divergence coefficient in 3.3, and the complexity of the treatment regimen is calculated.

[0123] 4.2.1 Calculate the average dose intensity of all drugs.

[0124] The number of drug types: the total number of drugs used is counted, reflecting the number of drugs involved in the treatment regimen. The more the number, the higher the complexity and risk.

[0125] n drug = total number of drug types.

[0126] Average dose intensity: calculate the average dose intensity of all drugs to assess the impact of drug dosage on the complexity of the treatment regimen.

[0127] Calculate the average dose intensity D of all drugs, as follows:

[0128]

[0129] where D m is the dose of the mth drug; m = 1, 2,..., n 药 .

[0130] 3.2.2 Calculate the average drug administration frequency of drugs.

[0131] Drug average administration frequency reflects the impact of drug administration frequency on the complexity of the regimen. Calculate the average administration frequency O of all drugs, as follows:

[0132]

[0133] where O t is the number of daily drug administrations of the tth drug; n 天 is the number of drug administration days; t = 1, 2,..., n 天 .

[0134] 3.3 Evaluate the risk level of each drug combination based on the drug interaction database, and calculate the average risk value.

[0135] Specifically, the risk level of each drug combination is evaluated based on the drug interaction database, the average risk value is calculated, and the risk degree of drug interaction is quantified. Each pair of drug combination is assigned a risk value (1-3), where 1 represents no risk, 2 represents low risk, and 3 represents high risk.

[0136] To accurately quantify the risk degree of drug interaction in the overall drug treatment plan, the average risk value of all drug combinations is calculated The calculation formula is as follows,

[0137]

[0138] In the formula, r gh is the risk value of the interaction between drug g and drug h; g≠h; is the number of combinations, indicating the number of combinations of 2 drugs selected from n drugs.

[0139] 3.4 Assign scores according to the special medication guidance of the drug, and calculate the average special requirement score.

[0140] Specifically, the average special requirement score is calculated according to the special medication guidance of the drug, and the special medication requirement situation is comprehensively evaluated. The special medication requirement of each drug is assigned a score (1-3), where 1 represents no special requirement, 2 represents general strictness, and 3 represents very strict requirement.

[0141] To comprehensively evaluate the overall situation of special medication requirement in the treatment plan, the average special requirement score Q of all drugs is calculated, and the formula is as follows:

[0142]

[0143] In the formula, Q l is the special medication requirement score of the lth drug, n 药 refers to the number of types of drugs included in the treatment plan.

[0144] 3.5 Obtain the diagnosis and treatment complexity index value according to the complexity index between symptoms, the average risk value of drug average dose intensity and drug average administration frequency, and the average special requirement score.

[0145] Specifically, the comprehensive complexity score is obtained by weighted sum of each index, and the complexity of TCM treatment strategy is comprehensively evaluated.

[0146] The above indexes are weighted and summed to obtain the comprehensive complexity score. Assuming that the weights of each index are w c1 , w c2 , w c3 , w c4 , w c5 , w c6The comprehensive complexity score can be represented as,

[0147]

[0148] where C is the complexity index, w c1 +w c2 +w c3 +w c4 +w c5 +w c6 = 1.

[0149] Specifically, the weighted sum of each indicator is obtained to obtain a comprehensive fuzziness index, which reflects the degree of fuzziness in diagnosis and treatment and affects medical decision-making.

[0150] Step five: constructing a comprehensive UCA index

[0151] The weights of variability, uncertainty, complexity, and fuzziness indices are set, and the weighted sum is obtained to obtain a comprehensive UCA index, which provides quantitative evaluation of various characteristics in the process of traditional Chinese medicine diagnosis and treatment, and assists doctors in formulating precise diagnosis and treatment plans.

[0152] Suppose the weights of variability, uncertainty, complexity, and fuzziness indices are w V , w U , w C , and w A .

[0153] Finally, the comprehensive UCA index is obtained, and the formula is as follows,

[0154] VUCA = w V ·V + w U ·U + w C ·C + w A ·A (30)

[0155] where w V +w U +w C +w A = 1.

[0156] It can be understood that in practice, a large amount of patient data is collected, and machine learning is used to determine the specific values of the weights w V , w U , w C , and w A . Specifically, subjective weighting method and / or objective weighting method can be used to calculate the weights.

[0157] The present application will be described in further detail below in conjunction with a specific implementation of diagnosis and treatment analysis of a patient with a complex disease.

[0158] (1) Diagnosis information and symptom description processing

[0159] 1) A patient with diabetes combined with cardiovascular disease comes to see a doctor. The doctor records 20 pieces of information in detail, and the total information in the medical record is 25 pieces.

[0160] According to formula (9), the diagnostic information integrity index is 0.8.

[0161] 2) The patient's symptoms are described as "often feel heart palpitations, chest tightness, thirst but not much water, sometimes numb hands and feet, urine foam, blurred vision, sleep poorly, easy to fatigue".

[0162] Extract the keywords "heart palpitations", "chest tightness", "thirst", "numb hands and feet", "urine foam", "blurred vision", "poor sleep", "fatigue", etc. Calculate the value of each keyword using the TF-IDF algorithm. Assume that the average value of all keyword values after calculation is 0.35, which is the degree of clarity of the symptom description index.

[0163] (2) Treatment plan related calculation

[0164] The patient's treatment plan involves 5 drugs, and according to formula (12), the selectability index of the treatment plan is 1.609.

[0165] Calculate the drug-related complexity index. The doses of the 5 drugs are 10mg, 20mg, 15mg, 5mg, and 8mg, respectively. According to formula (25), the average dose intensity is 11.6mg; the number of times of administration per day for the 5 drugs is 2, 3, 1, 2, and 1, respectively, and the administration days are 10 days. According to formula (26), the average administration frequency is 1.7 times / day.

[0166] Assume that the drug combination risk level is evaluated through the drug interaction database, and the average risk value of drug interaction is calculated to be 1.5. The scores of special drug requirements for the 5 drugs are 1, 2, 1, 1, and 2, respectively. According to formula (28), the average special requirement score is 1.4.

[0167] (3) Uncertainty and complexity index calculation

[0168] According to the information entropy, the diagnostic process uncertainty index is calculated. Assume that the entropy values of inspection, auscultation, interrogation, and palpation are 0.5, 0.3, 0.4, and 0.4, respectively, and the body type is 0.3. The weight of each indicator is 0.2. According to formula (14), the diagnostic process uncertainty index is 0.38.

[0169] The complexity index between symptoms is calculated, assuming that the patient has 8 symptoms, the mutual information between each two symptoms is calculated and summed and normalized (the specific calculation process is omitted), and the complexity index between symptoms is obtained. The weight of each index is determined, assuming that the weight of the complexity index between symptoms is 0.3, and the weights of each index of the complexity of the treatment scheme are 0.2 for the number of drug varieties, 0.15 for the average dose intensity, 0.15 for the frequency of administration, 0.2 for the risk of drug interaction, and 0.2 for the special drug requirement. The comprehensive complexity index is calculated according to formula (29).

[0170] It can be understood that the calculation of variability and ambiguity can refer to the above, which will not be described in detail here.

[0171] After obtaining the uncertainty index, the complexity index, the variability index and the ambiguity index, the UCI is calculated by formula (30).

[0172] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application and not to limit them; although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or some technical features can be replaced by equivalent ones; without departing from the spirit of the technical solutions of the present application, they should be covered in the technical solution range of the present application.

Claims

1. A method for measuring uncertainty in TCM diagnosis and treatment based on the Wuka index, characterized in that: The following steps are involved: Construct a Wuka assisted diagnosis and treatment model, wherein the Wuka assisted diagnosis and treatment model: VUCA=w V ·V+w U ·U+w C ·C+w A A Where w V , w U , w C , w A is the weight determined by machine learning, and w V +w U +w C +w A =1, V is the index value of diagnosis and treatment variability, U is the index value of diagnosis and treatment information uncertainty, C is the index value of diagnosis and treatment complexity, and A is the index value of diagnosis and treatment ambiguity; Obtain the patient's diagnostic information, treatment plan, and daily monitoring data, and obtain the diagnosis and treatment variability index value, diagnosis and treatment information uncertainty index value, diagnosis and treatment complexity index value, and diagnosis and treatment ambiguity index value based on the patient's diagnostic information, treatment plan, and daily monitoring data; The patient's comprehensive VUCA index is obtained according to the VUCA-assisted diagnosis and treatment model, and the comprehensive VUCA index is an uncertain value.

2. The uncertainty measurement method in the TCM diagnosis and treatment process based on the Vuka index according to claim 1 is characterized in that: Obtain variability index values ​​based on the patient's diagnostic information, treatment plan, and daily monitoring data, including: Obtain the patient's condition change rate index based on daily monitoring data and historical monitoring data; Obtain the patient's treatment response change index based on daily monitoring data and historical monitoring data; Obtain the patient's laboratory index based on daily monitoring data and historical monitoring data; The diagnostic and treatment variability index value was obtained based on the disease change rate index, treatment response change index and laboratory indicators.

3. The uncertainty measurement method in the TCM diagnosis and treatment process based on the Vuka index according to claim 1 or 2, characterized in that: Obtain the uncertainty index value of diagnosis and treatment information based on the patient's diagnosis information and treatment plan, including: The diagnostic information completeness index is obtained based on the completeness of the patient's four diagnostic information; Determine symptom description keywords based on the patient's symptom description document, use the TF-IDF algorithm to process the symptom description keywords, and obtain the symptom description fuzziness index; The treatment option selectivity index was calculated using a logarithmic function based on the number of treatment options; Convert the symptom frequency into a probability problem to calculate the entropy value of each symptom indicator, and obtain the uncertainty index of the diagnosis process based on the entropy value of each symptom indicator; The comprehensive uncertainty index value was obtained based on the diagnostic information completeness index, symptom description ambiguity index, treatment option index, and diagnostic process uncertainty index.

4. The uncertainty measurement method in the TCM diagnosis and treatment process based on the Vuka index according to claim 3 is characterized in that: Obtain the diagnosis and treatment ambiguity index value based on the patient's diagnosis information and treatment plan, including: Fleiss' Kappa coefficient was used to quantitatively evaluate the consistency of multiple doctors' diagnostic results on the same case and obtain the doctor's diagnostic consistency index; The compliance coefficient was calculated based on the patient's medication and follow-up records; Based on the doctors' scores on the treatment plans, the treatment plan disagreement coefficient is obtained; The diagnosis and treatment ambiguity index value was obtained based on the doctor's diagnosis consistency index, compliance coefficient and treatment plan disagreement coefficient.

5. The uncertainty measurement method in the TCM diagnosis and treatment process based on the Vuka index according to claim 4 is characterized in that: Obtain the diagnosis and treatment complexity index value based on the patient's diagnosis information and treatment plan, including: According to the mutual information between symptoms, the complexity index between symptoms is obtained; The average dosage intensity and average frequency of drug administration for all drugs were calculated; Evaluate the risk level of each drug combination based on the drug interaction database and calculate the average risk value; Points were assigned based on the special medication instructions, and the average special requirement score was calculated; The diagnosis and treatment complexity index value was obtained based on the complexity index between symptoms, average drug dosage intensity, average drug administration frequency, average risk value, and average special requirement score.

Images