CN105825064A - Daily traditional Chinese medicine therapy intelligent consultancy system - Google Patents

Abstract

The invention discloses a daily traditional Chinese medicine therapy intelligent consultancy system for western medicine or public self-service application. The daily traditional Chinese medicine therapy intelligent consultancy system adopts a technical scheme that the technical scheme comprises the following steps of checking and determining three states of health, sub-health and disease by western medicine, and respectively designing to-be-selected syndrome groups according to related factors of the discrimination; distributing the weights of the syndrome groups by a deviation weight method; defining membership functions for recognizing a pattern of the syndrome, and recognizing and sequencing the pattern of the syndrome by a multi-feature fuzzy pattern recognizing method; outputting the preferable sequence of traditional Chinese medicine therapy prescriptions by the pattern of the syndrome-prescription conversion; according to the effect feedback, optimizing the sequence. The daily traditional Chinese medicine therapy intelligent consultancy system has the advantages that by establishing a user information database, a diagnosis-selection syndrome database, a diagnosis-syndrome group database, a therapy prescription database and the like, and designing the mutual contact and support between modules of syndrome collection, discrimination, prescription outputting, feedback correcting and the like, the intelligent process of syndrome, discrimination, prescription output and feedback correcting is realized, so that the rapid and accurate query of the daily traditional Chinese medicine therapy is realized, and the problems of inconvenience in query, lack of discrimination and the like in the western medicine or public self-service application are solved.

Description

Routine tcm therapy Intelligence Consulting System

[technical field]

The present invention relates to the technical field of medical information that traditional Chinese medical science intelligence is dialectical, be specifically related to doctor trained in Western medicine or popular self-service The routine tcm therapy Intelligence Consulting System of application.

[background technology]

Routine tcm therapy universal and use and embody simple doctor trained in Western medicine be beyond one's reach curative effect be the traditional Chinese medical science and , there is positive meaning on the basis of integrated Chinese-western medical science to common people's rehabilitation, common such as dietetic therapy, folk prescription Proved recipe, OTC (over-the-counter) Chinese patent medicine etc., this be class need not the common Chinese traditional treatment mode of TCM Major physician guidance, Class huge instructing far beyond medical practitioner has the medical treatment letter of individuation feature outside ability and energy in other words conj.or perhaps Breath demand, and be difficult to meet in reality.These therapies are for non-TCM Major doctor and without in system For the people of theory of medicine study and practice, when needing, there is many difficult points such as inquiry, dialectical, the side of selecting, Cause its usefulness not fully demonstrate and misuse situation, have impact on traditional Chinese medical science reputation and popular popularization.Use interconnection The implementations such as net set up the routine tcm therapy intelligent consulting intercommunion platform of doctor trained in Western medicine or popular self-service application then Being efficient solution, set up the technological core that dialectical model is scheme, its effect will make consultation process accord with Close mode and the feature of Chinese medical discrimination thinking, improve consulting precision greatly.

Relevant area of computer aided Chinese medical discrimination carries out existing more than 30 years at home, is used for describing traditional Chinese medicine expert system The Chinese medical discrimination mathematical model of core and Knowledge Representation Method mainly have: based on statistical method such as polynary system The discriminant function and regression equation, method of maximum likelihood etc. set up analyzed by meter；Based on data digging method such as shellfish This model of leaf, rough set, artificial neural network；Other as weighted sum model, weighting mathematics logical model, Case-based reasioning, Grey Relational Model based on entropy weight, Group Decision Theory model, Fisher and Bayes sentence Not Fen Xi, support vector machine, principal component analysis and factorial analysis, correlation rule, analytic hierarchy process (AHP) etc..Though The feature so having taken into full account all kinds of process challenge mathematical method and the successful warp realized at other field Test, and carry out independent or connected applications trial, but for research this nonlinear complication system of dialectical process All Shortcomings, summary problem has: 1. most scholars attempt to be found to express from the Essential study of syndrome With differentiate the index of various syndromes or data, reach the purpose that differential diagnosis in tcm quantifies, and Chinese medical discrimination logic master Will be with as counting method approach application the most qualitatively is in dialectical, be with this think of with distinct fuzzy thinking feature Dimension process carries out fusion and has difficulties.2. as the syndrome information design of dialectical Analyses Main Basis, its intension Conceptualization is the most wide in range, causes syndrome to be subordinate to frequency poor with the dependency of card, the judgement of impact card；Its The secondary weight that can not reflect the state of an illness, i.e. the cause of disease, sick position, characteristic of disease, the degree of priority of patient's condition, cause group Side is based on insufficient grounds with dosage.3. the diagnostic criteria that there is card is difficult to formulate.4. syndrome elements differentiation body is included System, syndrome verification element, card element quantify all have great subjectivity to pattern of syndrome.

The determination for the treatment of based on pathogenesis obtained through differentiation of symptoms and signs process of the traditional Chinese medical science is the complication system engineering with non-linear behavior, in terms of dialectical key element, Non-linear behavior show syndrome as card diagnosis basis be fuzzy, non-quantized is uncertain, mutually it Between and and sick and card between be non-linear relation；In terms of thinking characteristic, Chinese medical discrimination logic is mainly with as number Method approach application the most qualitatively, in dialectical process, defines the traditional Chinese medical science such as Qu Xiang Bi Lei, illumination of the heart cognition traditional Mode of thinking, is subjective and experimental.Can express in all kinds of algorithms of above-mentioned fuzzy thinking feature with mould Stick with paste the theoretical multiple features Fuzzy Pattern Recognition Method formed that is combined with pattern recognition theory to agree with the most, be applied to card Type identification process can embody the system flexibility disposal ability that it is good.Difficult point is the definition of membership function.

Each syndrome weighted during determination for the treatment of based on pathogenesis obtained through differentiation of symptoms and signs, has influence on the accuracy of system, shows as pattern of syndrome Or card element diagnosis in same pattern of syndrome the weighted of each syndrome；Same syndrome is for the power of different pattern of syndrome Heavily there is also difference.Therefore, objective distribution syndrome weight is also problem demanding prompt solution.

The application of multiple features Fuzzy Pattern Recognition is divided into two aspects, and one is disease common pattern of syndrome identification, simultaneously Also serve as therapy data input layer；Two is the therapy prescription identification after staining effect adjusts indication, this One aspect essence is to seek a kind of segmentation pattern of syndrome adapted for every kind of therapy prescription.Can be understood as how many Therapy prescription just has how many pattern of syndrome, because different therapy prescription exists the difference of action intensity and mechanism.

Should using doctor trained in Western medicine check result as point of penetration, 1. at present people have got used to first carrying out doctor trained in Western medicine inspection and have examined Disconnected, select doctor trained in Western medicine, the traditional Chinese medical science or therapy of combining Chinese and Western medicine, doctor trained in Western medicine inspection timely not only to avoid diagnosis and treatment subsequently and prolong By mistake, also evade relating to evidence-based medicine EBM as foundation and cure laws and regulations risk.2. Western medicine diagnose has clear and definite Diagnostic criteria, to embody physiopathologic function and laboratory checking index as Main Basis, it is simple to carry out Association study between doctor trained in Western medicine index and Chinese traditional treatment；3. Chinese and western medicine disease combines (Western medicine diagnose Chinese medical discrimination) Common TCM Syndrome Typing under Western medicine diagnose premise and relevant main syndromes for many years, are presented with enough by research Understanding and experience accumulation, beneficially system design application.Western medicine diagnose result is roughly divided into health, Asia is good for Health, three kinds of states of disease.

The tcm therapies such as dietetic therapy, folk prescription proved recipe, OTC (over-the-counter) Chinese patent medicine totally can be divided into the ill, to sick, verification Three classes also carry out recording application accordingly, and verification situation occupies the majority；And abundant in content through long term accumulation, quantity Far beyond human brain memory ability, inquiry inconvenience.

Consulting system for amateur person designs the problem also needing to solve: the syndrome information 1. inputted reaches many Number people's culture level persons institute energy perception and the level of description, it is not necessary to have medical knowledge background, and need to examine The disappearance of worry information and Fault-Tolerant Problems；2. the INTERFACE DESIGN of popular operation level it is suitable for；3. the initial of therapy should With setting up on existing experiential basis, it is modified by user feedback information, to improving the standard of consulting Really property.

[summary of the invention]

It is contemplated that help popular or doctor of Western medicine's safety, application routine tcm therapy reasonable, efficient (dietetic therapy side, folk prescription proved recipe, OTC (over-the-counter) Chinese patent medicine etc.).For reaching conveniently to inquire about and higher inquiry precision Purpose, set up a kind of doctor trained in Western medicine or popular self-service application based on Chinese medical discrimination and the routine of fuzzy mathematics theory Tcm therapy Intelligence Consulting System, system solution can be real by modes such as app, dynamic web page and stand-alone program Existing.

User applies front preparation: suggestion user carries out doctor trained in Western medicine before using native system and checks the most healthy or sub-strong Health or morbid state；Healthy and sub-health state person be proposed with according to Chinese medicine association of China issue " in Doctor's classification of TCM constitution and judgement " tcm constitution designed by standard ZYYXH/T157-2009 measures software or problem Answering scoring (the most within the scope of the present invention) and carry out nine kinds of tcm constitutions judgements, result is stored in user profile number According to storehouse, as ignored, all treat by gentle body constitution.

User profile applying step: 1, first log into registration for identifying the cell-phone number of identity, password, select Property the input privacy information relevant to medical treatment such as sex, age, body weight height；2, healthy or subhealth state is selected Or morbid state enter respective interface, wherein there are symptom and two kinds of selections of diagnosis in morbid state interface；3, it is System lists selective relevant syndrome group according to different conditions, and user is according to the perception of syndrome group and reason Solution carries out single choice or multiselect；4, through system, extraction and the computing of correlated inputs syndrome are provided medical treatment in routine Method sequence and related efficacy explanation etc.；5, user selects therapy and gives recruitment evaluation after application；6 systems According to content in feedback information correction Relational database so as to get Continuous optimization.

To achieve these goals, it is provided that a kind of tcm therapy Intelligence Consulting System, it is characterised in that include,

A. storage element, stores data below storehouse:

User Information Database, including user basic information table: be used for recording user mobile phone number, password, Sex, age, body weight, height, occupation, custom hobby etc.；Body constitution table: record measurement of physical fitness and health result； Feedback information table: the information such as record description comment, gradation of effects comment and User reliability assessment；Make With log: record the syndrome and therapy prescription application record once inputted,

Diagnosis-select syndrome data base, including, inside and outside, woman, youngster, each section diagnosis-select of face classification Syndrome table: have in storage system name, Western medicine diagnose name, diagnosis be called for short or be commonly called as, selective dialectical phase Closing syndrome group, such as, internal medicine diagnoses-selects syndrome table record chronic bronchitis content: respiratory system- Chronic bronchitis-chronic bronchitis-syndrome group: generate heat, cough, expectoration ...；Symptom-selection syndrome table: record Common sympton, selective dialectical relevant syndrome group；Subhealth state selects syndrome table: record selective Asia Healthy dialectical relevant syndrome group,

Diagnosis-syndrome group data storehouse, including, subhealth state syndrome group data table: record subhealth state pattern of syndrome, Syndrome group, point syndrome, a point syndrome are subordinate to frequency etc.；Diagnosis-syndrome group data table: record Western medicine diagnose Name, pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to the contents such as frequency；Symptom-syndrome group data table: Record symptom, common pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to the contents such as frequency, and point syndrome is subordinate to Belong to frequency values=point syndrome positive number/total case load * 100%,

Therapy prescription database, including, dietetic therapy prescription table, folk prescription proved recipe table, OTC (over-the-counter) Chinese patent medicine table: point Do not record dietetic therapy, folk prescription proved recipe, the prescription name of OTC (over-the-counter) Chinese patent medicine, classification (to the ill, to sick, verification), Form, cure mainly, effect, the content such as effect explanation；Prescription is subordinate to frequency rule table: records prescription name, fit With diagnosis or symptom name, be suitable for pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to frequency,

B. syndrome acquisition module, carries out information alternately with User Information Database, diagnosis-selection syndrome data base, Selective syndrome proposes respectively with regard to three kinds of different conditions of user, and then it is dialectical to gather the syndrome that selected etc. Relevant information sends to dialectical module.Such as health status, extract and transmit obtained by the Internet temperature, The public informations such as humidity, air pressure, region, and obtained by User Information Database age, sex, body The medical science relevant informations such as high body weight, body constitution；The Asia that sub-health state is then selected syndrome table to extract by subhealth state is good for The dialectical relevant syndrome group of health is selective, and the information such as the body constitution obtained by User Information Database；Disease Status has symptom and two kinds of selections of diagnosis, is diagnosed-select syndrome table and symptom-selection syndrome by each section respectively Table provide select syndrome, above each selection syndrome can single choice or multiselect,

The most dialectical module, stores multiple features Fuzzy Pattern Recognition function and deviation weighting function,

Dialectical module reads the syndrome collected, and extracts in the correlation table of diagnosis-syndrome group data storehouse according to syndrome In corresponding syndrome group, a point syndrome is subordinate to frequency, and then obtains standard card according to the pattern of syndrome membership function of definition Type and pattern of syndrome to be identified be subordinate to angle value,

Multiple features fuzzy diagnosis function is used to obtain the approach degree of pattern of syndrome to be identified and standard pattern of syndrome storehouse,

The frequency that is subordinate to collected is calculated corresponding each pattern of syndrome weight by deviation weighting function,

Weighting approach degree is calculated according to the weight obtained and approach degree,

Pattern of syndrome sequence is obtained according to Similarity Principle,

D. the side's of sending module, is connected with dialectical module, reads in therapy prescription database correlation table according to pattern of syndrome and associates Therapy prescription, and through matching output therapy prescription and sequence, it should be noted that prescription apply for the first time be through Pattern of syndrome-prescription conversion, prescription obtains a point syndrome identical with pattern of syndrome and is subordinate to frequency values, and adjusts through application feedback After whole prescription is subordinate to frequency, the most directly carry out prescription sequence, similar each prescription is processed as a pattern of syndrome.

E. feedback module, obtains feedback information and includes descriptive comment and the assessment of curative effect grade, comprehensively analyze Rear employing is graded corresponding by height numerical value on earth on earth by height, is subordinate in order to adjust prescription in therapy prescription database Belong to and be subordinate to frequency values in frequency rule table.

F. display interface, includes for display: for syndrome, the side's of sending module of user's selection in syndrome acquisition module The therapy prescription of output and sequence and the feedback information etc. of feedback module.

Above-mentioned consulting system also has a following prioritization scheme:

In multiple features Fuzzy Pattern Recognition function, define standard pattern of syndrome storehouse A_iMembership function be:

$A_i\left(x\right)=(A_{i1},A_{i2},A_{i3}\mathrm{...}{A}_{ij})=\frac{\underset{r}{\max }\left\{x_{1r}\right\}}{A_{i1}}+\frac{\underset{r}{\max }\left\{x_{2r}\right\}}{A_{i2}}+\frac{\underset{r}{\max }\left\{x_{3r}\right\}}{A_{i3}}+\mathrm{...}+\frac{\underset{r}{\max }\left\{x_{jr}\right\}}{A_{ij}}$

Define pattern of syndrome B to be identified_iMembership function be:

$B_i\left(x\right)=(B_{i1},B_{i2},B_{i3}\mathrm{...}{B}_{ij})=\frac{x_{1q}}{A_{i1}}+\frac{x_{2t}}{A_{i2}}+\frac{x_{3s}}{A_{i3}}+\mathrm{...}+\frac{x_{jo}}{A_{ij}}$

Wherein i is pattern of syndrome number, and j is syndrome group number, and r is point symptom number in syndrome group, and q, t, s, c are for waiting to know Other person divides syndrome sequence number in each syndrome group.

In the dialectical module of described syndrome, during application multiple features Fuzzy Pattern Recognition function, approach degree between two syndrome groups Computational methods employing fuzzy nearness computing formula is

$N(A,B)=\frac{1}{2}\left(Hgt\right(A\∩B)+(1-Dpn\left(A\∪B\right)\left)\right)$

Wherein, HgtA=sup{A (x) x ∈ X} is the height of fuzzy set A,

DpnA=inf{A (x) x ∈ X} is the degree of depth of fuzzy set A.

In described deviation weighting function:

The standard deviation S of feature_iComputing formula is,

$S_i=\sqrt{\frac{1}{n-1}\underset{j=1}{\overset{n}{\Σ}}{(x_{ij}-\stackrel{\‾}{x_i})}^2}$

WhereinIt is characterized the average of i,

Weight distribution formula is,

$w_j=\frac{S_i}{m}$

Weight vectors is defined as W_j={ w₁,w₂,…,w_n, meet normalizationM is syndrome group mark Quasi-difference summation.

Described object to be identified B and mode standard A_iBetween weight approach degree P_iComputing formula

$P_i=\underset{j=1}{\overset{n}{\Σ}}{w}_{j}N(A_{ij},B_j).$

Described Similarity Principle is: set A_i, B ∈ F (X), i=1,2 ..., n, if N is (A_ij, B) and=max_i{N(A_i, B) }, then judge B With A_ijFor similar, wherein B is object to be identified, A_ijFor mode standard.

In described feedback module, feedback information includes descriptive comment and the assessment of curative effect grade, descriptive comment Being set to f after comprehensively analyzing, the assessment of curative effect grade uses the corresponding numerical value from high to low of grading from high to low to set For x, point card marquis in prescription is subordinate to frequency rule table in therapy prescription database is subordinate to frequency adjustment=former numerical value + xf*100%.

Beneficial effects of the present invention: for a long time, effective row that tcm field have accumulated countless use is convenient Dietetic therapy, folk prescription proved recipe, the routine tcm therapy such as OTC (over-the-counter) Chinese patent medicine, use modern information technology means Can improve its application efficiency greatly, native system is for doctor of Western medicine, it is possible to help them to apply easily These routine tcm therapies, in clinic, make up the limitation of western medical treatment, promote the development of the combination of Chinese and Western medicine； For masses, it is not necessary to possess medical speciality theoretical knowledge and experience, can seek advice from the most easily Exchange, contributes to solving people's puzzlement to health-care recovery method in reality, it is important that the dialectical mould of the present invention The design of type will improve the accuracy of inquiry greatly.System on humans keep healthy at ordinary times, alleviate disease card shape, Reduce medication, fast track rehabilitation has positive meaning, will promote the exploitation of pools of traditional Chinese medicine.

[accompanying drawing explanation]

Fig. 1 is the flow chart of steps of the present invention.

[detailed description of the invention]

The present invention is illustrated below in conjunction with the accompanying drawings with specific embodiment.These embodiments are interpreted as being merely to illustrate The present invention rather than limit the scope of the invention.The present invention can be made various change by those skilled in the art Or amendment, these equivalence changes and modification fall into the scope of the claims in the present invention equally.

Fig. 1 be the present invention system composition structural representation, the present embodiment initially set up User Information Database, Diagnosis-select syndrome data base, diagnosis-syndrome group data storehouse, therapy prescription database etc., stores system institute Need data, connect support each other by intermodules such as syndrome collection, dialectical, the side of sending, feedbacks, complete syndrome- Dialectical-the side of sending-feedback modifiers application process.

User Information Database in the present embodiment, is made up of following each table, user basic information table: be used for remembering Record user mobile phone number, password, sex, age, body weight, height, occupation, custom etc.；Body constitution table: note Record measurement of physical fitness and health result；Feedback information table: record description comment, gradation of effects comment and user are credible The information such as degree assessment；Use log: record the syndrome and therapy prescription application record once inputted.Each table by The same file-name field of corresponding contents is constituted, and cell-phone number is unique index labelling.

The present embodiment diagnoses-selects syndrome data base, including: inside and outside, woman, youngster, face classification each Section diagnoses-selects syndrome table: each disease order record included according to each system and each system of lower point of each section, Inside have system name, Western medicine diagnose name, diagnosis be called for short or be commonly called as, selective dialectical relevant syndrome group, Wherein Western medicine diagnose is to be diagnosed as reference in doctor trained in Western medicine undergraduate course serial textbooks, dialectical relevant syndrome group under Western medicine diagnose Group design Primary Reference higher medical education combination of Chinese and Western medicine serial textbooks；Symptom-selection syndrome table: record is often See symptom, selective dialectical relevant syndrome group, Primary Reference whole nation higher learning of Chinese medicine " tcm diagnosis Learn " etc. teaching material；Subhealth state selects syndrome table: record the dialectical relevant syndrome group of selective subhealth state.

Diagnosis-syndrome group data storehouse in the present embodiment, including: subhealth state syndrome group data table: record Asia Healthy pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to frequency etc., and content is with reference to " Sub-health State in TCM faces Bed guide " ZYYXH/T2-2006；Diagnosis-syndrome group data table: record Western medicine diagnose name, pattern of syndrome, syndrome Group, point syndrome, a point syndrome are subordinate to the contents such as frequency；Symptom-syndrome group data table: record symptom, often Witness type, syndrome group, point syndrome, a point syndrome are subordinate to the contents such as frequency.It should be noted that syndrome group And the design of point syndrome: can express the dialectical main points such as the cause of disease, sick position, characteristic of disease, patient's condition, evil positive status and Card or the card element same class scope of light and heavy degree and character (Symptoms is either-or), degree of priority Syndrome, summary and induction forms multiple syndrome group as dialectical fundamental, point syndrome table in its composition Reach minimizing intension concept, such as: the dialectical main points etc. of tinea manus and pedis are carried out feature selection (i.e. structure Feature extraction), combing goes out 3 characteristic syndrome groups: the course of disease, pain of itching, skin lesion situation, each group is respectively Formed by minimizing intension concept point syndrome: acute or chronic do not heal, without or slightly or lighter pruritus or Acutely pruritus or companion's pain, papulovesicle be main or erosion to ooze out be the main or coarse desquamation that thickens, each syndrome group In group, a point syndrome has either-or feature.The distributed data that above-mentioned point of syndrome is subordinate to frequency is certain through statistics Quantity clinical case produces, and a point syndrome is subordinate to frequency values=point syndrome positive number/total case load * 100%.

Therapy prescription database in the present embodiment, including: in dietetic therapy prescription table, folk prescription proved recipe table, OTC (over-the-counter) Patent medicine table: record dietetic therapy, folk prescription proved recipe, the prescription name of OTC (over-the-counter) Chinese patent medicine, classification respectively (to the ill, right Sick, verification), form, cure mainly, effect, the content such as effect explanation；Prescription is subordinate to frequency rule table: record Prescription name, applicable diagnosis or symptom name, applicable pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to frequency, It may be noted that a point card marquis for each prescription is subordinate to frequency initial value and divides card marquis to be subordinate to frequency, instead with applicable pattern of syndrome Feedback is the most identical after adjusting.

Syndrome acquisition module in the present embodiment: carry out with User Information Database, diagnosis-selection syndrome data base Information is mutual, proposes selective syndrome respectively with regard to user health, subhealth state, three kinds of different conditions of disease, And then gather the dialectical relevant informations such as the syndrome that selected and send to dialectical module.As: health status, extract And transmit obtained by the Internet gas epidemic disaster, air pressure, the public information such as region, and by user profile The medical science relevant informations such as age that data base obtains, sex, height and weight, body constitution；Sub-health state then by The dialectical relevant syndrome group of subhealth state that subhealth state selects syndrome table to extract is selective, and by user profile number According to information such as the body constitution that storehouse obtains；Morbid state state has symptom and two kinds of selections of diagnosis, is examined by each section respectively Disconnected-to select syndrome table and symptom-selection syndrome table to provide selection syndrome.

Syndrome acquisition module in the present embodiment: input syndrome is designed as can single choice or multiselect.

Dialectical module in the present embodiment: be connected and read in data with syndrome acquisition module, uses the deviation method of weighting Being weighted input syndrome, definition pattern of syndrome judges membership function, uses multiple features Fuzzy Pattern Identification to push away Derive the sequence of TCM Syndrome Type probability.Step is as follows:

1. syndrome domain is determined；

2. by the feature selection of object of study, Criterion model library and master pattern and object to be identified Membership function；

If master pattern (pattern of syndrome) storehouse A_iMembership function be:

$A_i\left(x\right)=(A_{i1},A_{i2},A_{i3}\mathrm{...}{A}_{ij})=\frac{\underset{r}{\max }\left\{x_{1r}\right\}}{A_{i1}}+\frac{\underset{r}{\max }\left\{x_{2r}\right\}}{A_{i2}}+\frac{\underset{r}{\max }\left\{x_{3r}\right\}}{A_{i3}}+\mathrm{...}+\frac{\underset{r}{\max }\left\{x_{jr}\right\}}{A_{ij}}$

If pattern of syndrome B to be identified_iMembership function be:

$B_i\left(x\right)=(B_{i1},B_{i2},B_{i3}\mathrm{...}{B}_{ij})=\frac{x_{1q}}{A_{i1}}+\frac{x_{2t}}{A_{i2}}+\frac{x_{3s}}{A_{i3}}+\mathrm{...}+\frac{x_{jo}}{A_{ij}}$

Wherein i is pattern of syndrome number, and j is syndrome group number, and r is point card shape number in syndrome group, and q, t, s, c are for treating Identification person divides syndrome sequence number in each syndrome group.

3. the application deviation method of weighting determines the weight distribution of characteristics index；

The standard deviation S of feature_iComputing formula

$S_i=\sqrt{\frac{1}{n-1}\underset{j=1}{\overset{n}{\Σ}}{(x_{ij}-\stackrel{\‾}{x_i})}^2}$

Wherein, the average of i it is characterized

Weight distribution formula:

Weight vectors is defined as W_j={ w₁,w₂,…,w_n, meet normalizationM is syndrome group mark Quasi-difference summation.

4. multiple features Fuzzy Pattern Recognition algorithm:

Proximity computation method between two fuzzy subsets: discrete type membership function uses fuzzy nearness to calculate public affairs Formula

$N(A,B)=\frac{1}{2}\left(Hgt\right(A\∩B)+(1-Dpn\left(A\∪B\right)\left)\right)$

In formula

HgtA=sup{A (x) x ∈ X} is the height (peak value) of fuzzy set A

DpnA=inf{A (x) x ∈ X} is the degree of depth (valley) of fuzzy set A

Two fuzzy vector set race proximity computation method: object to be identified B and mode standard A_iBetween paste Recency P_iApplication weighting approach degree computing formula

$P_i=\underset{j=1}{\overset{n}{\Σ}}{w}_{j}N(A_{ij},B_j)$

5. it is identified sequence according to the Similarity Principle of multiple characteristics；

Similarity Principle: set A_i, B ∈ F (X), i=1,2 ..., n, if N is (A_ij, B) and=max_i{N(A_i, B) }

Then judge B and A_ijFor similar, wherein B is object to be identified, A_ijFor mode standard.

Feedback module in the present embodiment: descriptive comment is the cognition narration during user uses, system User trust evaluation is set up according to the reasonability commented on, number of words, content, user's liveness etc. are multifactor F (wherein with reference to certainty theory), and this evaluation is similar to weight；Curative effect grade assessment x is user pair The final comment of proved recipe, letter is divided into " effectively ", " typically ", engineering noise three grades, and each grade is demarcated to be had accordingly 3,1 ,-2 numerical applications point card in prescription is subordinate to frequency rule table in calculating, therapy prescription database Marquis is subordinate to frequency adjustment=former numerical value+xf*100%, for avoiding malice or the random infringement evaluated system, is System is designed with filtration and arranges.Therapy prescription rule of elimination be according to always apply number of cases, effective percentage (efficiently individual quantity/ Total number of users), the many index such as untoward reaction determine, reaching certain standard system will delete automatically.

The present embodiment further illustrates as a example by the simple tinea manus and pedis of Symptoms.

The common pattern of syndrome of tinea manus and pedis is A₁(damp-heat accumulation), A₂(blood-deficiency and wind-dry), for pattern of syndrome A₁Proved recipe be R₁(wintersweet leaf), R₂(Liquor Cortex Mori), A₂For R₃(acetic acid Moschus liquid), R₄(liquid invaded by Radix Et Rhizoma Rhei vinegar).

If tinea manus and pedis patient point syndrome is object of study x, all composition domain U, 3 syndrome groups are on U 3 fuzzy subsets, are designated as C_i=(C_i1,C_i2,C_i3), and constitute a fuzzy vector set race, standard pattern of syndrome storehouse A_i={ A₁, A₂(i=1,2), its judgement is divided syndrome to determine by characteristic, is designated asIf disease to be identified The pattern of syndrome B of people_i=(B_i1,B_i2,B_i3), B_iAlso it is fuzzy vector set race.

The frequency that is subordinate to adding up each point of syndrome according to tinea manus and pedis clinical case obtains table 1 data

1 point of syndrome of table is subordinate to frequency (%)

Each standard pattern of syndrome A_iDegree of membership be decided by each syndrome group that the frequency that is subordinate to of occurred frequently point of syndrome is i.e. subordinate to Frequency, establishing criteria model (pattern of syndrome) storehouse A_iMembership function definition and table 1 know standard pattern of syndrome A₁=(x₁₁,x₂₃,x₃₁)=(0.64,0.62,0.54), A₂=(x₁₂,x₂₄,x₃₃)=(0.72,0.49,0.76).

The weight distribution of each syndrome group is to distribute meter according to table 1 market demand standard deviation s computing formula and weight Calculation formula calculates, and is shown in Table 2.

Table 2 each syndrome group weight allocation table

A₁Premise Xia Ge syndrome group weight W_j=(W₁,W₂,W₃)=(0.28,0.39,0.33), A₂For W_j=(W₁,W₂,W₃)=(0.35,0.23,0.42).

, there is referring to a many places papulovesicle in one pattern of syndrome tinea manus and pedis patient to be identified before February, rear ulceration acute pruritus, Existing pruritus alleviates, and at skin lesion, substrate pale red still has and oozes out.

The degree of membership of object to be identified pattern of syndrome depends on that the lower point syndrome of each standard pattern of syndrome is subordinate to frequency, if card to be identified Type B_iMembership function and corresponding point of syndrome of table 1 be (x₁₁,x₂₂,x₃₂), its degree of membership is at A₁B under premise₁= (x₁₁,x₂₂,x₃₂)=(0.64,0.28,0.37), A₂For B₂=(0.28,0.12,0.16).

Calculated by fuzzy nearness computing formula and weighting approach degree computing formula: P₁=0.586, P₂=0.744.

Know that patient's pattern of syndrome is ordered as A according to Similarity Principle₂(blood-deficiency and wind-dry) > A₁(damp-heat accumulation), replaces proved recipe A_iMaster pattern storehouse { the R that rear formation is new₁,R₂,R₃,R₄},A₂>A₁Sequence becomes R₃、R₄>R₁、R₂, system is pressed 50% output prescription sequence: R₃(acetic acid Moschus liquid), R₄(liquid invaded by Radix Et Rhizoma Rhei vinegar).

This user applies R₃Descriptive comment and the assessment of curative effect grade is given: effectively, system after (acetic acid Moschus liquid) Grading descriptive comment and combine previously reliability and carry out aggregative weighted weights 0.5, curative effect grade adds score value 3, Both long-pending 1.5*100%, be subordinate to the corresponding positive point syndrome in frequency rule table to prescription is subordinate to frequency values phase Add.

Initial R₁With R₂、R₃With R₄The identical sequence of degree of membership is random, and obtains and A₂、A₁Identical Point syndrome is subordinate to frequency values, and through apply a point card marquis for feedback adjustment prescription be subordinate to frequency again apply time, then Skip pattern of syndrome identification and directly carry out prescription identification sequence, and along with increment case correction, final prescription Divide card marquis to be subordinate to frequency values accounting in corresponding syndrome group will progressively tend towards stability, each syndrome group medium-high frequency Rate point syndrome set is the optimal adaptation card of prescription, and sort position, end and many index are beyond setting boundary Prescription will be eliminated automatically.

Claims (7)

1. a routine tcm therapy Intelligence Consulting System, it is characterised in that include,

A. storage element, stores data below storehouse:

User Information Database, including, user basic information table: be used for recording user mobile phone number, password, property Not, age, body weight, height, occupation, custom hobby；Body constitution table: record measurement of physical fitness and health result；Instead Feedforward information table: record description comment, gradation of effects comment and User reliability assessment information；Use Log: record the syndrome and therapy prescription application record once inputted,

Diagnosis-select syndrome data base, including, inside and outside, woman, youngster, each section diagnosis-select of face classification Syndrome table: have system name in storage, Western medicine diagnose name, diagnosis are called for short or are commonly called as, selective dialectical Relevant syndrome group,

Diagnosis-syndrome group data storehouse, including, subhealth state syndrome group data table: record subhealth state pattern of syndrome, card Time group, point syndrome, a point syndrome are subordinate to frequency；Diagnosis-syndrome group data table: record Western medicine diagnose name, Pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to frequency content；Symptom-syndrome group data table: note Record symptom, common pattern of syndrome, syndrome group, point syndrome, a point syndrome are subordinate to frequency content, and a point syndrome is subordinate to Frequency values=point syndrome positive number/total case load * 100%,

Therapy prescription database, including, dietetic therapy prescription table, folk prescription proved recipe table, OTC (over-the-counter) Chinese patent medicine table: respectively Record dietetic therapy, folk prescription proved recipe, prescription name, the classification of OTC (over-the-counter) Chinese patent medicine, form, cure mainly, effect, Effect description；Prescription is subordinate to frequency rule table: records prescription name, applicable diagnosis or symptom name, fit It is subordinate to frequency by pattern of syndrome, syndrome group, point syndrome, a point syndrome,

B. syndrome acquisition module, carries out information alternately with User Information Database, diagnosis-selection syndrome data base, Selective syndrome is proposed respectively with regard to three kinds of different conditions of user, each select syndrome can single choice or multiselect, And then gather the syndrome selected and non-selection dialectical relevant information transmission extremely dialectical module,

The most dialectical module, stores multiple features Fuzzy Pattern Recognition function and deviation weighting function,

Dialectical module reads the syndrome collected, and extracts in the correlation table of diagnosis-syndrome group data storehouse according to syndrome In corresponding syndrome group, a point syndrome is subordinate to frequency, and then obtains standard according to the pattern of syndrome membership function of definition Pattern of syndrome and pattern of syndrome to be identified be subordinate to angle value,

Multiple features fuzzy diagnosis function is used to obtain the approach degree of pattern of syndrome to be identified and standard pattern of syndrome storehouse,

The frequency that is subordinate to collected is calculated corresponding each pattern of syndrome weight by deviation weighting function,

Weighting approach degree is calculated according to the weight obtained and approach degree,

Pattern of syndrome sequence is obtained according to Similarity Principle,

D. the side's of sending module, is connected with dialectical module, reads in therapy prescription database correlation table according to pattern of syndrome and associates Therapy prescription, and through matching output therapy prescription and sequence,

E. feedback module, obtains and is subordinate to frequency rule according to prescription in feedback information adjustment therapy prescription database Table be subordinate to frequency values,

F. display interface, includes for display: for syndrome, the side's of sending module of user's selection in syndrome acquisition module The therapy prescription of output and sequence and the feedback information of feedback module.

2. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that described is the most special Levy in Fuzzy Pattern Recognition function, define standard pattern of syndrome storehouse A_iMembership function be:

$A_i\left(x\right)=\left(A_{i1},A_{i2},A_{i3}...A_{ij}\right)=\frac{\underset{r}{max}\left\{x_{1r}\right\}}{A_{i1}}+\frac{\underset{r}{max}\left\{x_{2r}\right\}}{A_{i2}}+\frac{\underset{r}{max}\left\{x_{3r}\right\}}{A_{i3}}+...+\frac{\underset{r}{max}\left\{x_{jr}\right\}}{A_{ij}}$

Define pattern of syndrome B to be identified_iMembership function be:

$B_i\left(x\right)=(B_{i1},B_{i2},B_{i3}...B_{ij})=\frac{x_{1q}}{A_{i1}}+\frac{x_{2t}}{A_{i2}}+\frac{x_{3s}}{A_{i3}}+...+\frac{x_{jc}}{A_{ij}}$

Wherein i is pattern of syndrome number, and j is syndrome group number, and r is point symptom number in syndrome group, and q, t, s, c are for waiting to know Other person divides syndrome sequence number in each syndrome group.

3. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that described syndrome is demonstrate,proved Wait in module, during application multiple features Fuzzy Pattern Recognition function, proximity computation method between two syndrome groups Employing fuzzy nearness computing formula is

$N\left(A,B\right)=\frac{1}{2}\left(Hgt\right(A\∩B)+(1-Dpn\left(A\∪B\right)\left)\right)$

Wherein, HgtA=sup{A (x) x ∈ X} is the height of fuzzy set A,

DpnA=inf{A (x) x ∈ X} is the degree of depth of fuzzy set A.

4. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that described deviation In weighting function:

The standard deviation S of feature_iComputing formula is,

$S_i=\sqrt{\frac{1}{n-1}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\left(x_{ij}-\stackrel{\‾}{x_i}\right)}^2}$

WhereinIt is characterized the average of i,

Weight distribution formula is,

$w_j=\frac{S_i}{m}$

Weight vectors is defined as W_j={ w₁,w₂,…,w_n, meet normalizationM is syndrome group standard Difference summation.

5. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that object to be identified B With mode standard A_iBetween weight approach degree P_iComputing formula

$P_i=\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{w}_{j}N\left(A_{ij},B_j\right).$

6. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that described in select the most former It is then: set A_i, B ∈ F (X), i=1,2 ..., n, if N is (A_ij, B) and=max_i{N(A_i, B) }, Then judge B and A_ijFor similar, wherein B is object to be identified, A_ijFor mode standard.

7. routine tcm therapy Intelligence Consulting System as claimed in claim 1, it is characterised in that in feedback module, Feedback information includes descriptive comment and the assessment of curative effect grade, and descriptive comment is set to after comprehensively analyzing F, the assessment of curative effect grade uses the corresponding numerical value from high to low of grading to be from high to low set to x, number formulary at therapy Point card marquis being subordinate in frequency rule table according to prescription in storehouse is subordinate to frequency adjustment=former numerical value+xf*100%.