A Digital Differentiation Method and Device for Treatise on Cold Damage

Abstract

This application relates to a digital syndrome differentiation method and device for Treatise on Cold Damage Disorders. The method includes: analyzing and mining the original text of Treatise on Cold Damage Disorders to obtain the conceptual hierarchical tree of diseases, syndromes, symptoms, treatment methods, and prescriptions for the six-channel diseases and their complications; constructing a rule base from four aspects: the outline syndromes of the six-channel diseases, treatment with prescriptions, misuse of treatment methods, and common knowledge; using the obtained user input symptom set and its credibility set as the evidence set; performing pattern matching between the evidence set and the rule base using the Rete algorithm to obtain a set of matching rules; calculating the conclusion credibility of each rule in the set of matching rules, sorting them, and selecting the rules with credibility values greater than a predetermined threshold to form an agenda queue; performing production hybrid reasoning on each matching rule in the agenda queue according to the type of rule conclusion. This method combines production reasoning and credibility reasoning and applies them to the syndrome differentiation process of Treatise on Cold Damage Disorders, providing a new idea and method for the digital syndrome differentiation of Treatise on Cold Damage Disorders.

Description

Technical Field

[0001] The present application relates to the fields of traditional Chinese medicine and computer information technology, and particularly to a digital syndrome differentiation method and device for Treatise on Cold Damage. Background Art

[0002] Treatise on Cold Damage is one of the four classic works of traditional Chinese medicine in China, which summarizes the medical achievements before the Han Dynasty and establishes a theoretical system of syndrome differentiation and treatment - six-meridian syndrome differentiation. Using information science, especially artificial intelligence technology, to realize the digitization and intelligentization of six-meridian syndrome differentiation is a current research hotspot. At present, there are already some digital reasoning methods for Treatise on Cold Damage, which can reason out the disease syndrome based on symptoms and then prescribe prescriptions. However, the existing methods do not handle well the problems such as the ambiguity and uncertainty of data in the knowledge system of Treatise on Cold Damage, and both the reasoning efficiency and the accuracy of the reasoning results need to be improved. Summary of the Invention

[0003] Based on this, it is necessary to provide a digital syndrome differentiation method and device for Treatise on Cold Damage in view of the above technical problems.

[0004] A digital syndrome differentiation method for Treatise on Cold Damage, the method comprising:

[0005] Analyzing and mining the original text of Treatise on Cold Damage to obtain a conceptual hierarchical structure tree of diseases, syndromes, symptoms, treatment methods, and prescriptions for six-meridian diseases and their complications.

[0006] Performing knowledge extraction and construction of production reasoning rules according to the outline syndromes or syndrome definitions of six-meridian diseases, prescription treatments, articles on prohibited and misused treatment methods in the original text of Treatise on Cold Damage, and the conceptual hierarchical structure tree of diseases, syndromes, symptoms, treatment methods, and prescriptions for six-meridian diseases and their complications, to obtain a digital syndrome differentiation rule library for Treatise on Cold Damage; the rule library includes six-meridian disease outline syndrome reasoning rules, formula-syndrome reasoning rules, prohibited and misused treatment method reasoning rules, and common knowledge reasoning rules.

[0007] Taking the obtained symptom set and its credibility set input by the user as an evidence set.

[0008] Using the Rete algorithm to perform pattern matching between the evidence set and the rule library to obtain a set of matching rules.

[0009] Calculating the conclusion credibility of each rule in the set of matching rules, sorting the matching rules in a decreasing order of the conclusion credibility, and selecting the rules with credibility values greater than a predetermined threshold to form an agenda queue.

[0010] Performing production hybrid reasoning on each matching rule in the agenda queue according to the type of rule conclusion until a reasoning result is obtained or a reasoning end condition is satisfied.

[0011] A digital syndrome differentiation device for Treatise on Cold Damage, the device comprising:

[0012] The six-channel disease concept hierarchy tree determination module is used to analyze and mine the original text of Treatise on Cold Damage Disorders to obtain the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions for the six-channel diseases and their complications.

[0013] The Treatise on Cold Damage Disorders digital syndrome differentiation rule base determination module is used to perform knowledge extraction and production inference rule construction based on the outline syndromes of the six-channel diseases or the disease syndrome definitions, prescription treatments, prohibited and misused treatment methods in the original text of Treatise on Cold Damage Disorders, as well as the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions for the six-channel diseases and their complications, to obtain the Treatise on Cold Damage Disorders digital syndrome differentiation rule base; the rule base includes six-channel disease outline syndrome inference rules, formula-syndrome inference rules, prohibited and misused treatment method inference rules, and common knowledge inference rules.

[0014] The Treatise on Cold Damage Disorders digital syndrome differentiation inference module is used to take the symptom set and its credibility set input by the user as the evidence set; perform pattern matching between the evidence set and the rule base using the Rete algorithm to obtain a set of matching rules; calculate the conclusion credibility of each rule in the set of matching rules, sort the matching rules in descending order of the conclusion credibility, select the rules with credibility values greater than a predetermined threshold to form an agenda queue; perform production hybrid inference on each matching rule in the agenda queue according to the type of rule conclusion until an inference result is obtained or the inference end condition is met.

[0015] The above-mentioned method and device for digital syndrome differentiation of Treatise on Cold Damage Disorders, the method includes: analyzing and mining the original text of Treatise on Cold Damage Disorders to obtain the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions for the six-channel diseases and their complications; and constructing rules from four aspects: six-channel disease outline syndromes (or disease syndrome definitions), prescription treatments, prohibited and misused treatment methods, and common knowledge, to obtain the Treatise on Cold Damage Disorders digital syndrome differentiation rule base; taking the symptom set and its credibility set input by the user as the evidence set; performing pattern matching between the evidence set and the rule base using the Rete algorithm to obtain a set of matching rules; calculating the conclusion credibility of each rule in the set of matching rules, sorting the matching rules in descending order of the conclusion credibility, selecting the rules with credibility values greater than a predetermined threshold to form an agenda queue; performing production hybrid inference on each matching rule in the agenda queue according to the type of rule conclusion until an inference result is obtained or the inference end condition is met. This method combines production inference and credibility inference and applies them to the syndrome differentiation process of Treatise on Cold Damage Disorders, providing a new idea and method for the digital syndrome differentiation of Treatise on Cold Damage Disorders. Description of the Drawings

[0016] Figure 1 It is a schematic flowchart of the digital syndrome differentiation method of Treatise on Cold Damage Disorders in an embodiment;

[0017] Figure 2 It is the six-channel disease syndrome concept hierarchy in an embodiment;

[0018] Figure 3 The concept hierarchy of the six-meridian disease syndromes in one embodiment;

[0019] Figure 4 The process schematic diagram of the first stage of reasoning in another embodiment;

[0020] Figure 5 The process schematic diagram of the second stage of reasoning in another embodiment;

[0021] Figure 6 The structural block diagram of the digital syndrome differentiation device for Treatise on Cold Damage in one embodiment. Detailed implementation manners

[0022] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0023] In one embodiment, as Figure 1 shown, a method for digital syndrome differentiation of Treatise on Cold Damage is provided, and the method includes the following steps:

[0024] Step 100: Analyze and mine the original text of Treatise on Cold Damage to obtain the concept hierarchy tree of diseases, syndromes, symptoms, treatment methods, and prescriptions of the six-meridian diseases and their complications.

[0025] Specifically, from the perspective of facilitating digital processing, the basic and concurrent syndromes of the six-meridian diseases and the complicated syndromes of the six-meridian diseases belong to different concept classification systems, and the concept classification method of the planning textbook "Selected Readings of Treatise on Cold Damage (4th Edition in the New Century)" is selected and discarded. Thus, the basic syndromes and concurrent syndromes of the six-meridian diseases are classified into the systematic syndrome differentiation category, and the complicated syndromes of the six-meridian disease syndromes, cholera disease syndromes, yin-yang exchange syndromes, relapse syndromes after recovery, similar syndromes, etc. are classified into the non-systematic syndrome category. In addition, since modern clinical practice is based on syndrome differentiation, it has little clinical significance to judge the combined diseases, concurrent diseases or dual-sensory diseases in the six-meridian disease syndromes. Therefore, the present invention collectively refers to these three situations as "complications of the six-meridian diseases", and the concept hierarchy tree of diseases, syndromes, etc. of the six-meridian diseases and their complications is as Figure 2 、 Figure 3 shown.

[0026] Step 102: Extract knowledge and construct production inference rules according to the outline syndromes or syndrome definitions, prescription treatments, prohibited and misused treatment methods in the original text of Treatise on Cold Damage and the concept hierarchy tree of diseases, syndromes, symptoms, treatment methods, and prescriptions of the six-meridian diseases and their complications to obtain the digital syndrome differentiation rule base of Treatise on Cold Damage. The rule base includes the inference rules for the outline syndromes of the six-meridian diseases, the inference rules for the prescription syndromes, the inference rules for prohibited and misused treatment methods, and the inference rules for common knowledge.

[0027] Specifically, from the five aspects of "disease-syndrome-symptom-treatment-method-prescription", knowledge extraction and inference rule construction are carried out for the articles in the 398 original texts of Treatise on Febrile Diseases that discuss the outline syndromes of the six-channel diseases (or the definitions of disease syndromes), prescription treatments, prohibited and misused treatment methods. In addition, rules need to be set for the common knowledge of the six-channel diseases such as the transitive property of the superordinate and subordinate relationships (i.e., the concept hierarchy relationship) and the necessary conditions for syndrome differentiation.

[0028] Step 104: Use the obtained set of symptoms input by the user and its credibility set as the evidence set; perform pattern matching between the evidence set and the rule base using the Rete algorithm to obtain the set of matching rules.

[0029] Specifically, the construction idea of this syndrome differentiation method is to divide the entire reasoning process into two aspects: forward reasoning and backward reasoning. The reasoning process first uses the forward reasoning method of production rules, takes the set of symptoms input by the user and its credibility set as the initial evidence set, and completes pattern matching according to the Rete algorithm to obtain the set of matching rules.

[0030] Step 106: Calculate the conclusion credibility of each rule in the set of matching rules, sort the matching rules in the order of decreasing conclusion credibility, and select the rules with credibility values greater than the predetermined threshold to form an agenda queue.

[0031] Preferably, the predetermined threshold is set to 0.6. By excluding the rules with low conclusion credibility, the scale of subsequent fact and rule matching can be significantly reduced, and the pattern matching process can be accelerated.

[0032] Step 108: Perform production hybrid reasoning on each matching rule in the agenda queue according to the type of rule conclusion until the reasoning result is obtained or the reasoning end condition is satisfied.

[0033] Specifically, in the second stage of the reasoning process, corresponding processing is performed for the conclusion types of each rule in the agenda queue. In the knowledge base system constructed by this method, the deduced conclusions can be classified into five categories: disease, syndrome, symptom, treatment method, and prescription. Therefore, corresponding processing needs to be done according to the different conclusion types.

[0034] To select targeted hypotheses during reverse reasoning, this method, guided by experts in classical prescriptions and based on the characteristics of the Six-Channel syndrome differentiation in the Treatise on Febrile Diseases, constructs a table of essential symptoms for the Six-Channel syndromes. To strike a balance between predictive accuracy and matching speed, at most one symptom from each syndrome is selected from both pulse symptoms and primary symptoms to form the necessary symptom set for that syndrome. This table is then used to construct a table of essential symptoms for the Six-Channel syndromes. Then, a subsumption operation is performed on each necessary symptom set in the table of essential symptoms for the Six-Channel syndromes. If the existing evidence set contains a syndrome's necessary symptom set, that syndrome is selected as a reverse reasoning hypothesis and added to the hypothesis set. This process is repeated to obtain a hypothesis set containing several hypotheses. These hypotheses are then verified using the aforementioned reverse reasoning method. Because valuable hypotheses are set for conclusions classified as disease and syndrome (non-leaf nodes) in the second stage of the reasoning process, irrelevant disease and syndrome matching processes can be effectively eliminated, improving reasoning efficiency. Table 1 shows the table of essential symptoms for the Six-Channel syndromes.

[0035] Table 1 Comparison table of necessary symptoms for six meridian diseases

[0036]

[0037] The aforementioned digital syndrome differentiation method for Treatise on Febrile Diseases includes: analyzing and mining the original text of Treatise on Febrile Diseases to obtain a conceptual hierarchical structure of diseases, syndromes, symptoms, methods, and prescriptions for six meridian diseases and their variant syndromes; constructing rules based on four aspects, namely, the outline syndromes (or disease and syndrome definitions) of six meridian diseases, prescription treatments, contraindications and misuses of treatment methods, and common knowledge, to obtain a digital syndrome differentiation rule base for Treatise on Febrile Diseases; using a user-entered symptom set and its credibility set as an evidence set; performing pattern matching between the evidence set and the rule base using the Rete algorithm to obtain a matching rule set; calculating the conclusion credibility of each rule in the matching rule set, sorting the matching rules in descending order of conclusion credibility, and selecting rules with credibility values greater than a predetermined threshold to form an agenda queue; and performing generative hybrid reasoning on each matching rule in the agenda queue based on the type of the rule conclusion until an inference result is obtained or an inference termination condition is met. This method combines generative reasoning and credibility reasoning and applies them to the syndrome differentiation process of Treatise on Febrile Diseases, providing a new approach and method for digital syndrome differentiation in Treatise on Febrile Diseases.

[0038] In one embodiment, step 102 includes: constructing a syndrome differentiation outline of the six meridian diseases expressed in a production-type form based on the definitions of the syndrome differentiation outline of the six meridian diseases and their variant syndromes, and the combined diseases, concurrent diseases and transmission of the six meridian diseases mentioned in the original text of the Treatise on Febrile Diseases.

[0039] Specifically, among the 398 original texts in the middle 10 chapters of Treatise on Cold Damage Disorders, it involves the syndrome differentiation outlines of the six-channel diseases such as Taiyang disease and Yangming disease, the definitions of syndromes such as Taiyang wind stroke and Taiyang cold damage, as well as the combined diseases, concurrent diseases and transmission of the six-channel diseases among each other. The following takes the outline syndrome of Taiyang disease, Taiyang wind stroke syndrome, Taiyang cold damage syndrome, and Taiyang-Yangming combined disease as examples to elaborate on the construction process of the rules.

[0040] rule 1:

[0041] (type x disease), (manifestationOf floating pulse x), (manifestationOf stiff neck and headache x), (manifestationOf aversion to cold x) => (isDisease x Taiyang disease)

[0042] rule 2:

[0043] (type x syndrome), (cooccursWith x Taiyang disease), (manifestationOf fever x), (manifestationOf sweating x), (manifestationOf aversion to wind x), (manifestationOf slow pulse x) => (isPattern x Taiyang wind stroke syndrome)

[0044] rule 3:

[0045] (type x syndrome), (cooccursWith x Taiyang disease), (manifestationOf aversion to cold x), (manifestationOf body pain x), (manifestationOf absence of sweating x), (manifestationOf vomiting and nausea x), (manifestationOf floating and tight pulse x) => (isPattern x Taiyang cold damage syndrome)

[0046] rule 4:

[0047] (isDisease x Taiyang-Yangming combined disease) => (isDisease x Taiyang disease), (isDisease x Yangming disease)

[0048] Based on the syndromes corresponding to the leaf nodes in the concept hierarchy tree of disease, syndrome, symptom, treatment method, and formula of the six-channel diseases and their complications, extract the common knowledge and laws of the original texts they involve to construct the formula-syndrome reasoning rules represented in the form of production rules.

[0049] Specifically, most of the 398 original texts in Treatise on Cold Damage Disorders focus on which formula to specifically adopt for the current disease syndrome to carry out syndrome differentiation and treatment. Therefore, the reasoning rules in this part are the core of the entire rule base. According toFigure 2 and Figure 3 Regarding the knowledge structure of the six-meridian syndromes and their variant syndromes, the present invention takes the syndromes corresponding to the leaf nodes in the knowledge structure tree as the basic units, and extracts the common knowledge and laws involved in the original texts corresponding to them to construct inference rules. The following takes the syndrome of wind-strike of Taiyang meridian as an example to illustrate the construction process of this part of the rules.

[0050] The articles related to the treatment of the syndrome of wind-strike of Taiyang meridian (including concurrent syndromes) are Article 12, Article 13, Article 95, Article 24, Article 42, Article 44, Article 45, Article 15, Article 57, Article 14, Article 43, Article 20, Article 21, Article 22, and Article 62. The commonalities of the syndromes described in these articles are usually symptoms such as headache, sweating, aversion to wind, and slow pulse, but at the same time, they also show some unique symptoms. Therefore, the following rules are set for the knowledge in each original text by using the original text knowledge extraction method. The original text knowledge extraction method specifically includes: extracting knowledge from aspects such as the original syndrome, treatment history, current syndrome, etiology and pathogenesis, syndrome differentiation (or disease differentiation) results, treatment plan, etc.; on the other hand, if a certain article involves multiple diagnostic conclusions, then this original text needs to be split into multiple articles so that each of the split articles contains only one diagnostic conclusion.

[0051] For example, in Article 12 of Treatise on Cold Damage Disorders, "In wind-strike of Taiyang meridian, the yang is floating and the yin is weak. When the yang is floating, heat spontaneously appears; when the yin is weak, sweating spontaneously occurs. Being chilly and shivering with aversion to cold, feeling shivery with aversion to wind, having mild fever, nasal congestion and retching, Cinnamon Twig Decoction is the principal prescription." The rules for Article 12 of the original text can be constructed as follows:

[0052] rule 12-1:

[0053] (type x syndrome), (manifestationOf sweating x), (manifestationOf fever x), (manifestationOf aversion to cold x), (manifestationOf aversion to wind x), (manifestationOf headache x), (manifestationOf floating and slow pulse x) => (isPattern x wind-strike syndrome of Taiyang meridian + Cinnamon Twig Decoction syndrome)

[0054] rule 12-2:

[0055] (isPattern x wind-strike syndrome of Taiyang meridian + Cinnamon Twig Decoction syndrome) => (treats Cinnamon Twig Decoction x)

[0056] rule 12-3:

[0057] (type x syndrome), (manifestationOf floating and slow pulse x) => (manifestationOf fever x), (manifestationOf sweating x)

[0058] For another example, the original text of Article 57 in Treatise on Cold Damage Disorders states that "After sweating in the treatment of cold damage, if there is restlessness again about half a day later and the pulse is floating and rapid, sweating can be induced again, and Cinnamon Twig Decoction is appropriate." This original text not only describes the current symptoms, but also elaborates on the original symptoms and treatment history. Therefore, its meaning should be represented in the corresponding reasoning rules. The following are the reasoning rules corresponding to Article 57 of the original text:

[0059] rule 57:

[0060] (type x syndrome), (previousPattern Taiyang Exogenous Cold syndrome x), (previousTreats diaphoresis x), (manifestationOf fever x), (manifestationOf aversion to cold x), (manifestationOf aversion to wind x), (manifestationOf floating and rapid pulse x) => (isPattern x Taiyang Wind-Stroke syndrome)

[0061] Based on the prohibition of using a certain treatment method or formula under specific disease conditions in the original text of Treatise on Cold Damage Disorders and the relationship between misuse and prohibition, reasoning rules for the prohibition and misuse of treatment methods are constructed in the form of production representation.

[0062] Specifically, there are multiple original texts in Treatise on Cold Damage Disorders indicating the prohibition of using a certain treatment method or formula under specific disease conditions. Therefore, in this invention, "┐" is used in the rules to represent the negative meaning. At the same time, there are also many articles elaborating on the changes in the disease condition in case of misuse, resulting in the inability of the original treatment method for the disease to adapt to the current disease condition, that is, misuse and prohibition are closely related. For example, the original text of Article 16 in Treatise on Cold Damage Disorders states that "On the third day of Taiyang disease, after sweating, if vomiting, purging, or warming needling is applied and the condition still persists, this is a deteriorated disease, and Cinnamon Twig Decoction is not applicable. Observe the pulse and symptoms to know what deviation has occurred, and treat according to the syndrome. Cinnamon Twig Decoction is originally for relieving the muscles. If the patient has a floating and tense pulse, fever, and no sweating, it cannot be given. One must always bear this in mind and avoid mistakes." Since this original text elaborates on four different situations (or clinical key points), four rules are used to describe it.

[0063] rule 16-1:

[0064] (type x syndrome), (previousDisease Taiyang disease x), (previousTreats diaphoresis x), (previousTreats emesis x) => ┐(treats Cinnamon Twig Decoction x)

[0065] rule 16-2:

[0066] (Syndrome type x), (Previous disease: Taiyang disease x), (Previous treatment: diaphoresis x), (Previous treatment: purgation x) => ¬(Treatment: Guizhi Decoction x)

[0067] Rule 16-3:

[0068] (Syndrome type x), (Previous disease: Taiyang disease x), (Previous treatment: diaphoresis x), (Previous treatment: fire therapy x) => ¬(Treatment: Guizhi Decoction x)

[0069] Rule 16-4:

[0070] (Syndrome type x), (Manifestation: floating and tense pulse x), (Manifestation: fever x), (Manifestation: absence of sweating x) => ¬(Treatment: Guizhi Decoction x)

[0071] Based on the hierarchical relationships of concepts in terms of disease, syndrome, symptoms, treatment methods, and prescriptions in the conceptual hierarchical structure tree of the six-channel diseases and their complications in Treatise on Cold Damage Disorders, as well as the clauses on the necessary conditions for syndrome differentiation in Treatise on Cold Damage Disorders, common knowledge reasoning rules represented in the form of production rules are constructed.

[0072] Specifically, in the conceptual classification system of Treatise on Cold Damage Disorders, there are hierarchical relationships among concepts such as disease, syndrome, symptoms, and treatment methods. In particular, there may be multi-level hierarchical relationships between syndromes. Therefore, transitive relationship rules are established:

[0073] (Subclass x y), (Subclass y z) => (Subclass x z)

[0074] (Type x y), (Subclass y z) => (Type x z)

[0075] In the original text of Treatise on Cold Damage Disorders, there are also clauses that specify the necessary conditions for syndrome differentiation. For example, in Article 15 of the original text, "For Taiyang disease, after purgation, if there is qi surging upward, Guizhi Decoction can be administered using the previous method. If there is no qi surging upward, it cannot be given." In this original text, the symptom "feeling qi surging upward" is the key to whether Guizhi Decoction can be used in treatment. In other words, after treating Taiyang disease with purgation, if Guizhi Decoction is still used for treatment, the current disease syndrome must have the symptom "feeling qi surging upward". Therefore, the following rule is used to describe this knowledge:

[0076] (Syndrome type x), (Previous disease: Taiyang disease x), (Previous treatment: Purgation method x), (Treatment: Cinnamon Twig Decoction x) => (Manifestation: Conscious sensation of qi ascending rebelliously x)

[0077] In one embodiment, as Figure 4 shown, step 106 specifically includes:

[0078] Step 400: Set the credibility of the rule according to a preset principle, and set the credibility of the evidence according to the type of evidence.

[0079] Step 402: Calculate the credibility of the conclusion based on the credibility of the evidence and the credibility of the rule.

[0080] Step 404: Sort the matching rules in descending order of the conclusion credibility, and select the rules with credibility values greater than a predetermined threshold to form an agenda queue.

[0081] In one embodiment, the evidence is divided into initial evidence and intermediate evidence; step 400 includes: Set the credibility of the Six Meridians disease outline syndrome inference rule and the common knowledge inference rule to 1; Set the credibility of the treatment method prohibited and misused inference rule to -1; Divide the credibility of the formula syndrome inference rule into four levels, including recommended, appropriate, permitted, and prohibited, and set the credibility to 1, 0.8, 0.6, -1 respectively; When the evidence is initial evidence, set the credibility of the evidence to the value given by the user providing the evidence; When the evidence is intermediate evidence, set the credibility of the evidence to the credibility of the inference conclusion obtained in the previous stage; Set the credibility of the negative form of the evidence to the opposite of the credibility of the evidence; When the evidence E is combined evidence, calculate the credibility of the conjunctive evidence set and the disjunctive evidence respectively; The formula for calculating the credibility of the conjunctive evidence is:

[0082] CF(E 和 ) = min{CF(E1), CF(E2), …, CF(E n )} (1)

[0083] where, CF(E 和 ) is the credibility of the conjunctive evidence, E is the conjunctive evidence, E = E1 AND E2 AND … AND E n , E1, E2 …… E n are the evidence 1, evidence 2, …… evidence n respectively, n is the number of evidence, and is an integer greater than 1; CF(E1), CF(E2), …, CF(E n ) are the credibility of the evidence 1, evidence 2, …… evidence n respectively.

[0084] The calculation formula for the credibility of disjunctive evidence is as follows:

[0085] CF(E 析 ) = max{CF(E1), CF(E2), …, CF(E n )} (2)

[0086] Among them, CF(E 析 ) is the credibility of the disjunctive evidence, and E 析 is the disjunctive evidence, and E 析 = E1 OR E2 OR … OR E n .

[0087] In one embodiment, the calculation of the credibility of the conclusion in step 402 includes two cases: when the conclusion is the conclusion deduced from one rule, calculate the conclusion credibility of the rule according to the credibility of the evidence and the rule; the calculation method formula for the credibility of the conclusion is:

[0088] CF(H) = CF(H, E) * max{0, CF(E)} (3)

[0089] Among them, CF(E) is the credibility of evidence E, and calculate the conclusion credibility of each rule according to the credibility of the evidence and the rule of each rule.

[0090] When the same conclusion is deduced from multiple rules, calculate the conclusion credibility of each rule according to the credibility of the evidence and the rule of each rule; and calculate the comprehensive credibility according to the conclusion credibility of multiple rules to obtain the credibility of the conclusion.

[0091] In one embodiment, when the same conclusion is deduced from multiple rules, calculate the conclusion credibility of each rule according to the credibility of the evidence and the rule of each rule; and calculate the comprehensive credibility according to the conclusion credibility of multiple rules to obtain the credibility of the conclusion, including:

[0092] When the same conclusion is deduced from k rules, calculate the conclusion credibility of each rule according to the credibility of the evidence and the rule of each rule; calculate the comprehensive credibility of the conclusion credibility of the first rule and the conclusion credibility of the second rule according to the conclusion credibility of the first rule and the conclusion credibility of the second rule; the calculation formula for the comprehensive credibility is:

[0093]

[0094] Among them, CF1(H) is the conclusion credibility of the first rule, and CF2(H) is the conclusion credibility of the second rule; CF 1-2 (H) is the comprehensive credibility of the conclusion credibility of the first rule and the conclusion credibility of the second rule; P(H) is the prior probability of conclusion H.

[0095] Calculate the combined credibility of the conclusions of the first three rules based on the combined credibility of the conclusion credibility of the first rule, the conclusion credibility of the second rule, and the conclusion credibility of the third rule; continue in this way until the aforementioned k rules are all synthesized to obtain the combined credibility of the k rules, and use the combined credibility of the k rules as the credibility of the conclusion.

[0096] Specifically, in the process of medical diagnosis, there is uncertainty in the existing basis for disease diagnosis (i.e., evidence), that is, there are situations where it is impossible to judge the authenticity of a piece of evidence and only partial belief is possible. Therefore, the medical diagnosis process is a process of uncertain reasoning. Credibility reasoning is a very effective method of uncertain reasoning and has been successfully applied in a large number of medical diagnosis expert systems. Therefore, this embodiment will also implement production reasoning based on the credibility reasoning model. The following explains the calculation process (or setting method) of the credibility CF(E) of evidence, the credibility CF(H) of the conclusion, and the credibility CF(H, E) of the rule in the reasoning system constructed by this method.

[0097] (1) Rule credibility setting

[0098] In the original texts in the middle 10 chapters of Treatise on Febrile and Miscellaneous Diseases that involve specific prescriptions for treatment, words such as "main treatment", "appropriate", "give", "can give", "cannot", "must not" appear frequently. These words represent the suitability of using a certain prescription for treating a certain disease syndrome. Therefore, the present invention sets the rule credibility CF(H, E) corresponding to these relevant articles to four levels, namely recommended (credibility value is 1), appropriate (credibility value is 0.8), permitted (credibility value is 0.6), and prohibited (credibility value is -1).

[0099] The rules involved in the original texts of Treatise on Febrile and Miscellaneous Diseases are divided into four categories. For the reasoning rules of the outline syndromes of the six channels diseases and the reasoning rules of common knowledge, the present invention sets their credibility CF(H, E) to 1; for the prescription-syndrome reasoning rules, the rule credibility CF(H, E) is set to four levels according to the aforementioned method; for the reasoning rules of prohibited and misused treatment methods, their credibility CF(H, E) is set to -1.

[0100] (2) Evidence credibility calculation

[0101] In the process of production reasoning based on the credibility reasoning model, since intermediate conclusions can be deduced from the initial evidence and these intermediate conclusions will participate in the reasoning process as evidence for the next reasoning, it is necessary to continuously calculate the credibility CF(E) of each piece of evidence and the credibility CF(H) of the conclusion during the reasoning process. The following explains the calculation processes of these two credibilities respectively.

[0102] The meaning of the value of the credibility CF(E) of evidence is shown in Table 2. Since evidence can be divided into two types: initial evidence and intermediate evidence, it is necessary to calculate their credibility separately. In addition, the calculation of the negative credibility of evidence and the calculation of the credibility of combined evidence composed of disjunctive and conjunctive operations also need to be considered together.

[0103] Table 2 The meaning of the value of the credibility CF(E)

[0104]

[0105] ① If the evidence E is initial evidence, the value of CF(E) is given by the user providing the evidence;

[0106] ② If the evidence E is intermediate evidence, then: CF(E) = CF(H’). H’ and E are actually the same object, which is the reasoning conclusion obtained in the previous stage and appears as intermediate evidence in this reasoning. The value of CF(H’) is the credibility of this intermediate conclusion (see the calculation method of the conclusion credibility below).

[0107] ③ Denote the negative form of the evidence E as ┐E, and CF(┐E) = -CF(E).

[0108] ④ For combined evidence, deal with conjunctive evidence and disjunctive evidence separately.

[0109] When the credibilities of the sub-items of the conjunctive evidence are CF(E1), CF(E2), …, CF(E n ), for the conjunctive evidence in the form of formula (5), its credibility calculation method is formula (6).

[0110] E = E1 AND E2 AND…AND E n (5)

[0111] CF(E) = min{CF(E1), CF(E2), …, CF(E n )} (6)

[0112] For the disjunctive evidence in the form of formula (7), its credibility calculation method is formula (8).

[0113] E = E1 OR E2 OR…OR E n (7)

[0114] CF(E) = max{CF(E1), CF(E2), …, CF(E n )} (8)

[0115] (3) Calculation of conclusion credibility

[0116] In the process of uncertainty reasoning based on the credibility model, every time uncertain knowledge is applied, the uncertainty of the conclusion (including intermediate conclusions) needs to be calculated according to the uncertainty of the evidence and rules (i.e., knowledge), and its calculation method is shown in Equation (9).

[0117] CF(H) = CF(H, E) * max{0, CF(E)} (9)

[0118] When CF(E) < 0, according to Equation (9), the value of CF(H) can be obtained as 0, indicating that when the evidence is false to a certain extent, the truth or falsehood of conclusion H cannot be judged, so the value of CF(H) is 0. When CF(E) = 1, according to Equation (9), the value of CF(H) is equal to the value of CF(H, E), indicating that when the evidence is true, the credibility of the conclusion is the credibility of the knowledge.

[0119] When the same conclusion is deduced from multiple pieces of knowledge, uncertainty synthesis calculation is required to obtain the comprehensive credibility of the conclusion. Now assume there are rules r1, r2,..., r k , in the following form:

[0120] r1: IF E1 THEN H CF(H, E1)

[0121] r2: IF E2 THEN H CF(H, E2)

[0122] ...

[0123] r k : IF E k THEN H CF(H, E k )

[0124] Then the calculation method of CF(H) is divided into two stages:

[0125] ① According to Equation (9), calculate the conclusion credibility in each rule in turn, denoted as CF1(H), CF2(H),..., CF k (H).

[0126] ② Calculate the comprehensive credibility CF 1-2 (H) of CF1(H) and CF2(H) according to Equation (4), and then calculate the comprehensive credibility CF 1-3 (H) of this value and CF3(H), and so on until the above k rules are all synthesized, and the obtained CF 1-k (H) is the comprehensive credibility of CF(H).

[0127] In one embodiment, as Figure 5 shown, step 108 specifically includes the following steps:

[0128] Step 500: Scan the agenda queue.

[0129] Step 502: When the conclusion set included in the current matching rule includes clinical syndromes and prescriptions, a threshold value may be preset according to the credibility of the clinical syndromes and prescriptions to determine the processing method of the current matching rule.

[0130] Step 504: When the type of the rule conclusion is a symptom and method rule conclusion, the current matching rule is added as a new evidence to the existing evidence set.

[0131] Step 506: When the type of rule conclusion is a rule conclusion of disease and syndrome, a syndrome hypothesis set is constructed according to the table of necessary symptoms for the six meridian diseases and syndromes in the Treatise on Febrile Diseases, and reverse reasoning is performed based on each syndrome hypothesis until the reasoning result is obtained or the reasoning end condition is met; the table of necessary symptoms for the six meridian diseases and syndromes in the Treatise on Febrile Diseases is constructed according to the characteristics of the six meridian syndrome differentiation in the Treatise on Febrile Diseases under the guidance of classical prescription experts.

[0132] Specifically, ① If the current conclusion set contains Figure 2 、 Figure 3 The leaf nodes (i.e., clinical syndromes from the Treatise on Febrile Diseases) and prescriptions in the six-channel disease and its variant syndrome knowledge structure are processed separately based on their credibility. If their credibility exceeds a preset threshold (preferably 0.8), the syndrome and prescription are returned as inference results, terminating the inference process. Otherwise, the syndrome and prescription are saved as inference results, removed from the agenda, and the next stage of knowledge inference is continued.

[0133] ② If the current conclusion is of the type of symptoms and methods, add it as new evidence to the existing evidence set.

[0134] ③ If the current conclusion type is disease and syndrome (non-leaf node), then reasoning is performed according to the production-type reverse reasoning method. Since the disease (or syndrome category) to which the current condition belongs has been obtained, the hypothesis target can be set in a targeted manner ( Figure 2-3 The correctness of the hypothesis is demonstrated using reverse reasoning based on the current set of evidence (see the leaf nodes of the knowledge structure of the Six Meridian Diseases and Their Variations shown in ). If the hypothesis is found to be true, the hypothesis and its prescription are returned as the reasoning result, and the reasoning ends. Otherwise, the conclusion and its rule are deleted from the agenda queue. Then, similar conclusions in the agenda queue are selected and processed in the same manner. If no reasoning result is obtained after processing all such conclusions, the reasoning process returns to the first stage and a new round of mixed reasoning is initiated using the existing evidence set.

[0135] In one embodiment, step 506 includes: when the type of rule conclusion is a rule conclusion of disease and syndrome, constructing a syndrome hypothesis set according to the necessary symptom comparison table of the six meridian diseases and syndromes in the Treatise on Febrile Diseases; processing each syndrome hypothesis in the syndrome hypothesis set: when the syndrome hypothesis is established, adding the syndrome hypothesis and its prescription to the reasoning result set; when the syndrome hypothesis is not established, deleting the conclusion and the current matching rule from the agenda queue; after all syndrome hypotheses are processed, if the obtained reasoning result set is not empty, returning the result set and ending the reasoning; otherwise, continuing the next round of generative mixed reasoning using the existing evidence set until the reasoning end condition is met.

[0136] In one embodiment, step 502 includes: when the conclusion set contained in the current matching rule includes clinical syndromes and prescriptions, when the credibility of the clinical syndromes and prescriptions is not less than a preset threshold, the syndromes and prescriptions are returned as reasoning results and the reasoning process is terminated; when the credibility of the clinical syndromes and prescriptions is less than the preset threshold, the syndromes and prescriptions are taken as a reasoning result and added to the reasoning result set, and the current matching rule is removed from the agenda queue.

[0137] It should be understood that although Figure 1 、 Figure 4 and Figure 5 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in the present invention, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Figure 1 、 Figure 4 and Figure 5 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0138] In one embodiment, Figure 6 As shown, a digital syndrome differentiation device for Treatise on Febrile Diseases is provided, comprising: a module for determining a conceptual hierarchy structure of six meridian diseases, a module for determining a digital syndrome differentiation rule base for Treatise on Febrile Diseases, and a module for determining a digital syndrome differentiation reasoning for Treatise on Febrile Diseases, wherein:

[0139] The module for determining the conceptual hierarchical structure of the Six-Channel Diseases is used to analyze and mine the original text of the Treatise on Febrile Diseases to obtain a conceptual hierarchical structure of diseases, syndromes, symptoms, methods, and prescriptions for the Six-Channel Diseases and their variant syndromes.

[0140] The Digital Syndrome Differentiation Rule Base Determination Module of Treatise on Cold Damage is used to extract knowledge and construct production inference rules based on the outline syndromes of the six-channel diseases, the definitions of disease syndromes, the treatment with prescriptions, the articles on the prohibited and misused treatment methods in the original text of Treatise on Cold Damage, and the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions of the six-channel diseases and their complications, so as to obtain the digital syndrome differentiation rule base of Treatise on Cold Damage. The rule base includes the inference rules for the outline syndromes of the six-channel diseases, the inference rules for formula syndromes, the inference rules for prohibited and misused treatment methods, and the inference rules for common knowledge.

[0141] The Digital Syndrome Differentiation Inference Module of Treatise on Cold Damage is used to take the set of symptoms and their credibility sets input by the user as the evidence set. The Rete algorithm is used to perform pattern matching between the evidence set and the rule base to obtain the set of matching rules. Calculate the conclusion credibility of each rule in the set of matching rules, sort the matching rules in descending order of conclusion credibility, and select the rules with credibility values greater than the predetermined threshold to form an agenda queue. According to the types of rule conclusions, perform production hybrid reasoning on each matching rule in the agenda queue until the reasoning result is obtained or the reasoning end condition is met.

[0142] In one embodiment, the Digital Syndrome Differentiation Rule Base Determination Module of Treatise on Cold Damage is further used to construct the inference rules for the outline syndromes of the six-channel diseases represented in the form of production based on the syndrome differentiation outlines of the six-channel diseases, the definitions of the disease syndromes of the six-channel diseases and their complications, and the combined diseases, concurrent diseases, and transmission of the six-channel diseases mentioned in the original text of Treatise on Cold Damage. Taking the syndromes corresponding to the leaf nodes in the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions of the six-channel diseases and their complications as the basic units, extract the common knowledge and laws of the original text they involve to construct the inference rules for formula syndromes represented in the form of production. According to the prohibition of using a certain treatment method or prescription under specific disease conditions and the relationship between misuse and prohibition mentioned in the original text of Treatise on Cold Damage, construct the inference rules for prohibited and misused treatment methods represented in the form of production. According to the concept hierarchy relationships of diseases, syndromes, symptoms, and treatment methods in the concept hierarchy trees of diseases, syndromes, symptoms, treatment methods, and prescriptions of the six-channel diseases and their complications, and the articles on the necessary conditions for syndrome differentiation in Treatise on Cold Damage, construct the inference rules for common knowledge represented in the form of production.

[0143] In one embodiment, the Digital Syndrome Differentiation Inference Module of Treatise on Cold Damage is further used to set the credibility of the rules according to the preset principles, and set the credibility of the evidence according to the type of evidence. Calculate the conclusion credibility based on the credibility of the evidence and the credibility of the rules. Sort the matching rules in descending order of conclusion credibility, and select the rules with credibility values greater than the predetermined threshold to form an agenda queue.

[0144] In one of the embodiments, the evidence is divided into initial evidence and intermediate evidence; the digital syndrome differentiation reasoning module of Treatise on Febrile and Miscellaneous Diseases is further configured to set the credibility of the outline syndrome reasoning rules of the six-meridian diseases and the common knowledge reasoning rules to 1; set the credibility of the reasoning rules for the prohibited and misused treatment methods to -1; divide the credibility of the formula syndrome reasoning rules into four levels, including recommendation, suitability, permission, and prohibition, and set the credibility to 1, 0.8, 0.6, and -1 respectively; when the evidence is initial evidence, set the credibility of the evidence to the value given by the user providing the evidence; when the evidence is intermediate evidence, set the credibility of the evidence to the credibility of the reasoning conclusion obtained in the previous stage; set the credibility of the negative form of the evidence to the opposite of the credibility of the evidence; when the evidence E is combined evidence, calculate the credibility of the conjunctive evidence set and the credibility of the disjunctive evidence respectively; the calculation formula for the credibility of the conjunctive evidence is:

[0145] CF(E 和 ) = min{CF(E1), CF(E2), …, CF(E n )}

[0146] where CF(E 和 ) is the credibility of the conjunctive evidence, E is the conjunctive evidence, E = E1 AND E2 AND … AND E n , E1, E2 …… E n are the evidence 1, evidence 2, …… evidence n respectively, n is the number of evidence, and is an integer greater than 1; CF(E1), CF(E2), ……, CF(E n ) are the credibility of the evidence 1, evidence 2, …… evidence n respectively.

[0147] The calculation formula for the credibility of the disjunctive evidence is:

[0148] CF(E 析 ) = max{CF(E1), CF(E2), …, CF(E n )}

[0149] where CF(E 析 ) is the credibility of the disjunctive evidence, E 析 is the disjunctive evidence, E 析 = E1 OR E2 OR … OR E n .

[0150] In one of the embodiments, the calculation of the credibility of the conclusion in the digital syndrome differentiation reasoning module of Treatise on Febrile and Miscellaneous Diseases includes two cases: when the conclusion is the conclusion deduced from one rule, calculate the conclusion credibility of the rule according to the credibility of the evidence and the rule; the calculation method formula for the conclusion credibility is:

[0151] CF(H) = CF(H, E) * max{0, CF(E)}

[0152] Among them, CF(E) is the credibility of evidence E, and the credibility of the conclusion of each rule is calculated based on the evidence of each rule and the credibility of the rule.

[0153] When the same conclusion is deduced from multiple rules, the credibility of the conclusion of each rule is calculated based on the evidence of each rule and the credibility of the rule; and the comprehensive credibility is calculated based on the credibility of the conclusions of multiple rules to obtain the credibility of the conclusion.

[0154] In one embodiment, the digital syndrome differentiation reasoning module of Treatise on Febrile Diseases is further configured to, when the same conclusion is deduced from k rules, calculate the credibility of the conclusion of each rule based on the evidence of each rule and the credibility of the rule; calculate the comprehensive credibility of the credibility of the conclusion of the first rule and the credibility of the conclusion of the second rule based on the credibility of the conclusion of the first rule and the credibility of the conclusion of the second rule; the calculation formula of the comprehensive credibility is as shown in formula (4).

[0155] Calculate the comprehensive credibility of the credibility of the conclusions of the first 3 rules based on the comprehensive credibility of the credibility of the conclusion of the first rule and the credibility of the conclusion of the second rule and the credibility of the conclusion of the third rule; and so on until the aforementioned k rules are all synthesized to obtain the comprehensive credibility of the k rules, and use the comprehensive credibility of the k rules as the credibility of the conclusion.

[0156] In one embodiment, the digital syndrome differentiation reasoning module of Treatise on Febrile Diseases further includes:

[0157] A processing method determination module for matching rules, which is used to scan the agenda queue; when the conclusion set included in the currently matched rule contains clinical syndromes and prescriptions, determine the processing method of the currently matched rule according to the preset threshold of the credibility of the clinical syndromes and prescriptions.

[0158] A rule conclusion processing module for symptoms and treatment methods, which is used to add the currently matched rule as new evidence to the existing evidence set when the type of the rule conclusion is a rule conclusion for symptoms and treatment methods.

[0159] A rule conclusion processing module for diseases and syndromes, which is used to construct a syndrome hypothesis set according to the necessary symptom comparison table of the six-channel syndromes of Treatise on Febrile Diseases when the type of the rule conclusion is a rule conclusion for diseases and syndromes, and perform backward reasoning according to each syndrome hypothesis until a reasoning result is obtained or the reasoning end condition is met; the necessary symptom comparison table of the six-channel syndromes of Treatise on Febrile Diseases is constructed under the guidance of experts in classical formulas according to the characteristics of the six-channel syndrome differentiation of Treatise on Febrile Diseases.

[0160] In one embodiment, the disease and syndrome rule conclusion processing module is also used to construct a syndrome hypothesis set according to the necessary symptom comparison table of the six meridian diseases and syndromes in the Treatise on Febrile Diseases when the type of the rule conclusion is the disease and syndrome rule conclusion; process each syndrome hypothesis in the syndrome hypothesis set: when the syndrome hypothesis is established, the syndrome hypothesis and its prescription are added to the reasoning result set; when the syndrome hypothesis is not established, the conclusion and the current matching rule are deleted from the agenda queue; after all syndrome hypotheses are processed, if the obtained reasoning result set is not empty, the result set is returned and the reasoning ends; otherwise, the existing evidence set is used to continue the next round of generative mixed reasoning until the reasoning end condition is met.

[0161] In one embodiment, the matching rule processing method determination module is also used to, when the conclusion set contained in the current matching rule includes clinical symptoms and prescriptions, when the credibility of the clinical symptoms and prescriptions is not less than a preset threshold, return the symptoms and prescriptions as reasoning results and end the reasoning process; when the credibility of the clinical symptoms and prescriptions is less than the preset threshold, add the symptoms and prescriptions as a reasoning result and add them to the reasoning result set, and remove the current matching rule from the agenda queue.

[0162] The specific limitations of the digital syndrome differentiation device for Treatise on Febrile Diseases can be found in the limitations of the digital syndrome differentiation method for Treatise on Febrile Diseases above, and will not be repeated here. The various modules in the digital syndrome differentiation device for Treatise on Febrile Diseases can be implemented in whole or in part through software, hardware, or a combination thereof. The modules can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so that the processor can call and execute the operations corresponding to the modules above.

[0163] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0164] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A digital syndrome differentiation method for Treatise on Cold Damage Disorders, characterized in that, The method includes: Analyzing and mining the original text of Treatise on Cold Damage Disorders to obtain the concept hierarchy trees of diseases, syndromes, symptoms, therapies, and prescriptions for the six-channel diseases and their complications; Based on the outline syndromes of the six-channel diseases or the definitions of disease syndromes, prescription treatments, articles on prohibited and misused therapies in the original text of Treatise on Cold Damage Disorders, and the concept hierarchy trees of diseases, syndromes, symptoms, therapies, and prescriptions for the six-channel diseases and their complications, performing knowledge extraction and construction of production inference rules to obtain the digital dialectical rule base of Treatise on Cold Damage Disorders; the rule base includes inference rules for the outline syndromes of the six-channel diseases, prescription-syndrome inference rules, inference rules for prohibited and misused therapies, and common knowledge inference rules; Taking the symptom set and its credibility set of the user input obtained as the evidence set; Using the Rete algorithm to perform pattern matching between the evidence set and the rule base to obtain a set of matching rules; Calculating the conclusion credibility of each rule in the set of matching rules, sorting the matching rules in the order of decreasing conclusion credibility, and selecting the rules with credibility values greater than a predetermined threshold to form an agenda queue; Performing production hybrid inference on each matching rule in the agenda queue according to the type of rule conclusion until an inference result is obtained or the inference end condition is satisfied; Among them, calculating the conclusion credibility of each rule in the set of matching rules, sorting the matching rules in the order of decreasing conclusion credibility, and selecting the rules with credibility values greater than a predetermined threshold to form an agenda queue includes: Step 400: Setting the credibility of the rules according to a preset principle, and setting the credibility of the evidence according to the type of evidence; the evidence is divided into initial evidence and intermediate evidence; the specific steps include: Setting the credibility of the inference rules for the outline syndromes of the six-channel diseases and the common knowledge inference rules to 1; setting the credibility of the inference rules for prohibited and misused therapies to -1; Dividing the credibility of the prescription-syndrome inference rules into four levels, including recommended, suitable, permitted, and prohibited, and setting the credibility to 1, 0.8, 0.6, and -1 respectively; When the evidence is initial evidence, setting the credibility of the evidence to the value given by the user providing the evidence; When the evidence is intermediate evidence, setting the credibility of the evidence to the credibility of the inference conclusion obtained in the previous stage; Setting the credibility of the negative form of the evidence to the opposite of the credibility of the evidence; When the evidence E is combined evidence, calculating the credibility of the conjunctive evidence set and the credibility of the disjunctive evidence respectively; the calculation formula for the credibility of the conjunctive evidence is: CF(E 和 ) = min{CF(E1), CF(E2), …, CF(E n )} Among them, CF(E 和 ) is the credibility of the conjunctive evidence, E is the conjunctive evidence, E = E1 AND E2 AND…AND E n , E1, E2……E n are the evidence 1, evidence 2, ……, evidence n respectively, n is the number of evidences, and is an integer greater than 1; are CF(E1), CF(E2), ……, CF(E n ) are the credibility of the evidence 1, evidence 2, ……, evidence n respectively; The calculation formula for the credibility of the disjunctive evidence is: CF(E 析 ) = max{CF(E1), CF(E2), …, CF(E n )} Among them, CF(E 析 ) is the credibility of the disjunctive evidence, E 析 is the disjunctive evidence, E 析 = E1 OR E2 OR…OR E n ; Step 402: Calculating the conclusion credibility according to the credibility of the evidence and the credibility of the rule; among them, the calculation of the conclusion credibility includes two cases: When the conclusion is the conclusion deduced by one rule, calculating the conclusion credibility of the rule according to the credibility of the evidence and the rule; the calculation method formula for the conclusion credibility is: CF(H) = CF(H, E) * max{0, CF(E)} Where CF(E) is the credibility of the evidence E, and calculating the conclusion credibility of each rule according to the evidence and the credibility of each rule; When the same conclusion is deduced by multiple rules, calculate the confidence of the conclusion of each rule according to the evidence of each rule and the confidence of the rule; and calculate the comprehensive confidence according to the confidence of the conclusions of multiple rules to obtain the confidence of the conclusion; the specific steps include: When the same conclusion is deduced by k rules, calculate the confidence of the conclusion of each rule according to the evidence of each rule and the confidence of the rule; According to the confidence of the conclusion of the first rule and the confidence of the conclusion of the second rule, calculate the comprehensive confidence of the confidence of the conclusion of the first rule and the confidence of the conclusion of the second rule; the formula for the comprehensive confidence is: Among them, CF1(H) is the confidence of the conclusion of the first rule, and CF2(H) is the confidence of the conclusion of the second rule; CF 1-2 (H) is the combined confidence of the conclusion of the first rule and the conclusion of the second rule; P(H) is the prior probability of the conclusion H; According to the comprehensive confidence of the confidence of the conclusion of the first rule and the confidence of the conclusion of the second rule and the confidence of the conclusion of the third rule, calculate the comprehensive confidence of the confidence of the conclusions of the first 3 rules; and so on until the aforementioned k rules are all synthesized to obtain the comprehensive confidence of the k rules, and use the comprehensive confidence of the k rules as the confidence of the conclusion; Step 404: Sort the matching rules in the order of decreasing confidence of the conclusion, and select the rules with confidence values greater than the predetermined threshold to form an agenda queue.

2. The method according to claim 1, characterized in that, According to the syndrome differentiation outline of the six-meridian diseases or the definition of disease syndromes, formula treatment, prohibited and misused treatment methods in the original text of Treatise on Cold Damage Disorders, and the concept hierarchy tree of diseases, syndromes, symptoms, treatment methods, and formulas of the six-meridian diseases and their complications, perform knowledge extraction and construction of production inference rules to obtain the digital syndrome differentiation rule base of Treatise on Cold Damage Disorders, including: According to the syndrome differentiation outline of the six-meridian diseases, the definition of disease syndromes of the six-meridian diseases and their complications, and the combination diseases, concurrent diseases, and transmission of the six-meridian diseases in the original text of Treatise on Cold Damage Disorders, construct the inference rules of the syndrome differentiation outline of the six-meridian diseases represented in the form of production; Taking the syndromes corresponding to the leaf nodes in the concept hierarchy tree of diseases, syndromes, symptoms, treatment methods, and formulas of the six-meridian diseases and their complications as the basic units, extract the common knowledge and laws of the original texts involved to construct the prescription-syndrome inference rules represented in the form of production; According to the prohibition of using a certain treatment method or formula under specific conditions in the original text of Treatise on Cold Damage Disorders and the relationship between misuse and prohibition, construct the inference rules of prohibited and misused treatment methods represented in the form of production; According to the conceptual hierarchical relationships of diseases, syndromes, symptoms, and treatment methods in the concept hierarchy tree of diseases, syndromes, symptoms, treatment methods, and formulas of the six-meridian diseases and their complications, and the original texts of the necessary conditions for syndrome differentiation in Treatise on Cold Damage Disorders, construct the common knowledge inference rules represented in the form of production.

3. The method according to claim 1, wherein According to the types of rule conclusions, perform production hybrid reasoning on each matching rule in the agenda queue until the reasoning result is obtained or the reasoning end condition is met, including: Scan the agenda queue; When the conclusion set included in the current matching rule contains clinical syndromes and formulas, determine the processing method of the current matching rule according to the confidence of the clinical syndromes and formulas and the preset threshold; When the type of the rule conclusion is the rule conclusion of symptoms and treatment methods, add the current matching rule as new evidence to the existing evidence set; When the type of rule conclusion is a rule conclusion of disease and syndrome, a syndrome hypothesis set is constructed according to the necessary symptom comparison table of the six meridian diseases and syndromes in the Treatise on Febrile Diseases, and reverse reasoning is performed based on each syndrome hypothesis until the reasoning result is obtained or the reasoning end condition is met; the necessary symptom comparison table of the six meridian diseases and syndromes in the Treatise on Febrile Diseases is constructed according to the characteristics of the six meridian syndrome differentiation in the Treatise on Febrile Diseases under the guidance of classical prescription experts.

4. The method according to claim 3, wherein When the rule conclusion type is a disease and syndrome rule conclusion, a syndrome hypothesis set is constructed based on the necessary symptom comparison table of the six meridian diseases and syndromes in the Treatise on Febrile Diseases. Reverse reasoning is performed based on each syndrome hypothesis until the reasoning result is obtained or the reasoning end condition is met, including: When the rule conclusion type is the rule conclusion of disease and syndrome, the syndrome hypothesis set is constructed according to the necessary symptom comparison table of six meridian diseases and syndromes in Treatise on Febrile Diseases; Processing each syndrome hypothesis in the syndrome hypothesis set: when the syndrome hypothesis is established, the syndrome hypothesis and its prescription are added to the inference result set; when the syndrome hypothesis is not established, the conclusion and the current matching rule are deleted from the agenda queue; After all syndrome hypotheses are processed, if the obtained inference result set is not empty, the result set is returned and the inference ends; otherwise, the existing evidence set is used to continue the next round of generative mixed reasoning until the inference end condition is met.

5. The method according to claim 3, wherein When the conclusion set included in the current matching rule includes clinical syndromes and prescriptions, a threshold value may be preset based on the credibility of the clinical syndromes and prescriptions to determine the processing method of the current matching rule, including: When the conclusion set included in the current matching rule contains clinical syndromes and prescriptions, and when the credibility of the clinical syndromes and prescriptions is not less than the preset threshold, the syndromes and prescriptions are returned as reasoning results, and the reasoning process ends; When the credibility of a clinical syndrome and prescription is less than a preset threshold, the syndrome and prescription are taken as an inference result and added to the inference result set, and the current matching rule is removed from the agenda queue.

6. A digital syndrome differentiation device for Treatise on Cold Damage Disorders, characterized in that, The device comprises: The module for determining the conceptual hierarchical structure of Six-Jing Diseases is used to analyze and mine the original text of Treatise on Febrile Diseases to obtain the conceptual hierarchical structure of diseases, syndromes, symptoms, methods, and prescriptions for Six-Jing Diseases and their variant syndromes; A module for determining a digital syndrome differentiation rule base for the Treatise on Febrile Diseases is used to extract knowledge and construct generative reasoning rules based on the provisions on the outline syndromes or disease and syndrome definitions, prescription treatments, prohibited and misused treatment methods discussed in the original text of the Treatise on Febrile Diseases, as well as the concept hierarchy structure of diseases, syndromes, symptoms, treatments, and prescriptions for the six meridian diseases and their variants, thereby obtaining a digital syndrome differentiation rule base for the Treatise on Febrile Diseases; the rule base includes reasoning rules for the outline syndromes of the six meridian diseases, reasoning rules for prescription and syndrome, reasoning rules for prohibited and misused treatment methods, and reasoning rules for common knowledge; The Digital Differentiation and Reasoning Module of Treatise on Cold Damage Disorders is used to take the symptom set and its credibility set input by the user as the evidence set; adopt the Rete algorithm to perform pattern matching between the evidence set and the rule base to obtain a set of matching rules; calculate the conclusion credibility of each rule in the set of matching rules, and sort the matching rules in the order of decreasing conclusion credibility, and select the rules with credibility values greater than a predetermined threshold to form an agenda queue; perform production hybrid reasoning on each matching rule in the agenda queue according to the type of rule conclusion until an inference result is obtained or the inference end condition is met. Among them, the Digital Differentiation and Reasoning Module of Treatise on Cold Damage Disorders is also used to set the credibility of rules according to preset principles, and set the credibility of evidence according to the type of evidence; evidence is divided into initial evidence and intermediate evidence; calculate the conclusion credibility according to the credibility of evidence and the credibility of rules; sort the matching rules in the order of decreasing conclusion credibility, and select the rules with credibility values greater than a predetermined threshold to form an agenda queue. The Digital Differentiation and Reasoning Module of Treatise on Cold Damage Disorders is also used to set the credibility of the Six Meridians Disease Outline Syndrome Reasoning Rules and the Common Knowledge Reasoning Rules to 1; set the credibility of the Treatment Prohibition and Misuse Reasoning Rules to -1; divide the credibility of the Formula-Syndrome Correspondence Reasoning Rules into four levels, including recommended, suitable, permitted, and prohibited, and set the credibility to 1, 0.8, 0.6, and -1 respectively; when the evidence is initial evidence, set the credibility of the evidence to the value given by the user providing the evidence; when the evidence is intermediate evidence, set the credibility of the evidence to the credibility of the inference conclusion obtained in the previous stage; set the credibility of the negative form of the evidence to the opposite of the credibility of the evidence; when the evidence E is combined evidence, calculate the credibility of the conjunctive evidence set and the credibility of the disjunctive evidence respectively; the calculation formula for the credibility of the conjunctive evidence is: CF(E 和 ) = min{CF(E1), CF(E2), …, CF(E n )} Among them, CF(E 和 ) is the credibility of the conjunctive evidence, E is the conjunctive evidence, E = E1 AND E2 AND…AND E n , E1, E2, ……, E n are the evidence 1, evidence 2, ……, evidence n respectively, n is the number of evidence, and is an integer greater than 1; are CF(E1), CF(E2), ……, CF(E n ) are the credibility of evidence 1, evidence 2, ……, evidence n respectively; The calculation formula for the credibility of the disjunctive evidence is: CF(E 析 ) = max{CF(E1), CF(E2), …, CF(E n )} Among them, CF(E 析 ) is the credibility of the disjunctive evidence, and E 析 is the disjunctive evidence, where E 析 = E1 OR E2 OR…OR E n ; The calculation of the conclusion credibility in the Digital Differentiation and Reasoning Module of Treatise on Cold Damage Disorders includes two cases: when the conclusion is the conclusion deduced by one rule, calculate the conclusion credibility of the rule according to the credibility of the evidence and the rule; the calculation method formula for the conclusion credibility is: CF(H) = CF(H, E) * max{0, CF(E)} Among them, CF(E) is the credibility of the evidence E, and calculate the conclusion credibility of each rule according to the evidence and the credibility of each rule. When the same conclusion is deduced by multiple rules, calculate the conclusion credibility of each rule according to the evidence and the credibility of each rule; and calculate the comprehensive credibility according to the conclusion credibility of multiple rules to obtain the conclusion credibility; the specific steps include: When the same conclusion is deduced by k rules, calculate the conclusion credibility of each rule according to the evidence and the credibility of each rule. According to the conclusion credibility of the first rule and the conclusion credibility of the second rule, calculate the comprehensive credibility of the conclusion credibility of the first rule and the conclusion credibility of the second rule; the calculation formula for the comprehensive credibility is: Among them, CF1(H) is the confidence of the conclusion of the first rule, and CF2(H) is the confidence of the conclusion of the second rule; CF 1-2 (H) is the combined confidence of the conclusion of the first rule and the conclusion of the second rule; P(H) is the prior probability of the conclusion H; Calculate the combined confidence of the conclusions of the first 3 rules based on the combined confidence of the conclusion confidence of the first rule, the conclusion confidence of the second rule, and the conclusion confidence of the third rule; continue in this way until the aforementioned k rules are all synthesized to obtain the combined confidence of the k rules, and use the combined confidence of the k rules as the confidence of the conclusion.

Images