An AI-based clinical intelligent question-answering system

Abstract

The present invention relates to the field of artificial intelligence technology, and in particular to an AI-based clinical intelligent question-answering system, comprising a question recognition module, which is used to recognize user questions and determine user needs; a semantic analysis module, which is used to determine the clarity of the user's question; a keyword extraction module, which is used to determine the extraction of keywords of the sentences in the user's question according to the clarity; a result determination module, which is used to search for relevant information in a database according to the keywords extracted by the keyword extraction module to determine preliminary results; a user feedback module, which is used to determine the accuracy of the preliminary results according to user feedback; and an adjustment module, which is used to determine the adjustment of the extraction process of the keyword extraction module or the process of determining the preliminary results of the result determination module according to the accuracy. The present invention solves the problem of low accuracy of the results of the intelligent question-answering system in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AI-based clinical intelligent question-answering system. Background Art

[0002] Clinical intelligent question-answering systems are increasingly widely used in the modern medical field. Based on big data analysis and artificial intelligence technology, they provide great convenience for medical professionals and patients. By analyzing and processing massive amounts of medical information, clinical intelligent question-answering systems can quickly and accurately provide users with symptom self-diagnosis, disease prediction, and intelligent doctor assistance services. Clinical intelligent question-answering systems play an important role in improving the efficiency of medical services, reducing the workload of medical staff, and providing personalized health advice. With the continuous development of artificial intelligence technology, its application areas will be further expanded, injecting new vitality into the development of the medical industry.

[0003] Chinese Patent Publication No.: CN113590783B discloses a TCM health care intelligent question-answering system based on NLP natural language processing, including: a user terminal and a management server, wherein the management server includes a TCM health care text segmentation module, a part-of-speech tagging module, a named entity recognition and classification module, a dependency analysis module, a semantic similarity calculation module and an answer retrieval module. The TCM health care text segmentation module performs segmentation processing on a clinical manifestation text according to a predetermined segmentation vocabulary, the named entity recognition and classification module performs named entity recognition and classification on the segmented words, the dependency analysis module analyzes the relationship between the various components of the clinical manifestation text to obtain the semantics of the clinical manifestation text, the semantic similarity calculation module calculates the semantic similarity value between the user question and the stored question based on the semantics of the clinical manifestation text, and the answer retrieval module uses a predetermined retrieval mechanism to retrieve the answer corresponding to the user question from the question-answering database according to the semantic similarity value as a health care suggestion.

[0004] It can be seen that there is a problem in the prior art that the results of the intelligent question-answering system are not accurate enough. Summary of the invention

[0005] To this end, the present invention provides an AI-based clinical intelligent question and answer system to overcome the problem of low accuracy of the results of the intelligent question and answer system in the prior art.

[0006] To achieve the above objectives, the present invention provides an AI-based clinical intelligent question-answering system, comprising:

[0007] A problem identification module is used to identify the user's problem and determine the user's needs;

[0008] A semantic analysis module, connected to the question identification module, for determining the clarity of the user's question;

[0009] A keyword extraction module, connected to the semantic analysis module, for determining the extraction of keywords from the sentence in the user's question according to the clarity;

[0010] A result determination module, which is connected to the keyword extraction module and is used to search for relevant information in a database according to the keywords extracted by the keyword extraction module to determine a preliminary result;

[0011] A user feedback module, which is connected to the result determination module and is used to determine the accuracy of the preliminary result according to user feedback;

[0012] An adjustment module is connected to the user feedback module and is used to adjust the extraction process of the keyword extraction module or the process of determining the preliminary result of the result determination module according to the accuracy.

[0013] Furthermore, the keyword extraction module determines the extraction of keywords from the sentence in the user's question according to the clarity, including determining a clarity evaluation value, so as to determine the extraction of keywords from the sentence in the user's question according to a comparison result between the clarity evaluation value and a preset clarity evaluation value.

[0014] Furthermore, the step of extracting keywords from the sentence in the user's question according to the comparison result between the clarity evaluation value and the preset clarity evaluation value includes:

[0015] Under the condition that the clarity evaluation value is greater than a preset clarity evaluation value, extracting keywords from the sentence in the user's question;

[0016] Under the condition that the clarity evaluation value is less than or equal to a preset clarity evaluation value, keywords and words and sentences associated with the keywords in the sentence in the user's question are extracted.

[0017] Furthermore, the result determination module searches for relevant information in a database according to the keywords extracted by the keyword extraction module to determine a preliminary result, including processing the relevant information to determine the preliminary result, and determining the eligibility of the preliminary result.

[0018] Furthermore, the processing of the relevant information includes determining features in the relevant information to determine the screening of the relevant information according to the number of the features.

[0019] Further, the screening of the relevant information according to the number of features includes:

[0020] When the number of features is less than or equal to the preset number of features, identical information in the related information is filtered and deduplication is performed;

[0021] When the number of features is greater than a preset number of features, all features are associated with symptoms, and the number of associated symptoms is determined;

[0022] The relevant information after screening is determined to be the preliminary results.

[0023] Furthermore, the result determination module determines that the preliminary result is unqualified under the condition that the number of the associated symptoms is greater than a preset number.

[0024] Furthermore, the result determination module extracts similar symptoms in the associated disease as the preliminary result under the condition that the preliminary result is unqualified.

[0025] Furthermore, the adjustment module determines to adjust the extraction process of the keyword extraction module or the process of determining the preliminary result of the result determination module, including determining the adjustment process according to the difference between the accuracy evaluation value and the preset accuracy evaluation value and the ratio of the preset accuracy evaluation value difference.

[0026] Further, the determining of the adjustment process according to the comparison result of the difference between the accuracy evaluation value and the preset accuracy evaluation value and the difference between the preset accuracy evaluation values ​​comprises:

[0027] Under the condition that the difference between the accuracy evaluation value and the preset accuracy evaluation value is less than or equal to the preset accuracy evaluation value difference, adjusting the preset number of features according to the adjustment coefficient;

[0028] Under the condition that the difference between the accuracy evaluation value and the preset accuracy evaluation value is greater than the preset accuracy evaluation value difference, the preset clarity evaluation value is adjusted according to the adjustment coefficient.

[0029] Compared with the prior art, the beneficial effect of the present invention lies in that, by setting up a semantic analysis module, the clarity evaluation value of the user's question is determined, and the extraction method of the keyword extraction module is determined according to the comparison result of the clarity evaluation value and the preset clarity evaluation value, which can improve the accuracy of keyword extraction, optimize the utilization of system resources, avoid unnecessary waste, and at the same time lay a good foundation for improving the accuracy of the results.

[0030] Furthermore, the present invention sets a result determination module to search for relevant information in the data according to the extracted keywords, and adopts different screening methods for the relevant information according to the different comparison results between the number of features and the preset number of features M, thereby improving the accuracy of the results.

[0031] Furthermore, the present invention sets a user feedback module to determine the accuracy evaluation value of the preliminary results according to the user's feedback, and adopts different adjustment methods according to the difference between the accuracy evaluation value and the preset accuracy evaluation value and the comparison result of the difference between the preset accuracy evaluation values, which can improve the accuracy of the adjustment and further improve the accuracy of the results.

[0032] Furthermore, the present invention can improve the accuracy of adjustment and avoid the occurrence of over-adjustment by setting the adjustment module and setting specific adjustment coefficients under different adjustment modes. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 Schematic diagram of the structure of the AI-based clinical intelligent question-answering system described in an embodiment of the present invention;

[0034] Figure 2 This is a workflow diagram of a keyword extraction module of an AI-based clinical intelligent question-answering system described in an embodiment of the present invention;

[0035] Figure 3 This is a workflow diagram of a result determination module of an AI-based clinical intelligent question-answering system according to an embodiment of the present invention;

[0036] Figure 4 This is a workflow diagram of the adjustment module of the AI-based clinical intelligent question-answering system described in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0038] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0039] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0040] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0041] Example

[0042] See also Figure 1-Figure 4 As shown, the AI-based clinical intelligent question-answering system of the embodiment of the present invention includes:

[0043] A problem identification module is used to identify the user's problem and determine the user's needs;

[0044] A semantic analysis module, connected to the question identification module, for determining the clarity of the user's question;

[0045] A keyword extraction module, connected to the semantic analysis module, for determining the extraction of keywords from the sentence in the user's question according to the clarity;

[0046] A result determination module, which is connected to the keyword extraction module and is used to search for relevant information in a database according to the keywords extracted by the keyword extraction module to determine a preliminary result;

[0047] A user feedback module, which is connected to the result determination module and is used to determine the accuracy of the preliminary result according to user feedback;

[0048] An adjustment module is connected to the user feedback module and is used to adjust the extraction process of the keyword extraction module or the process of determining the preliminary result of the result determination module according to the accuracy.

[0049] Specifically, the keyword extraction module determines the extraction of keywords from the sentence in the user's question according to the clarity, including determining a clarity evaluation value P, so as to determine the extraction of keywords from the sentence in the user's question according to the comparison result between the clarity evaluation value P and a preset clarity evaluation value P0.

[0050] In the embodiment of the present invention, the clarity evaluation value P is determined according to the following formula:

[0051]

[0052] Among them, P represents the clarity evaluation value, N represents the number of keywords, K represents the correlation between keywords, and L represents the length of the question.

[0053] In an embodiment of the present invention, the keywords include disease names, body parts, and symptoms; the correlation between the keywords can be determined by but not limited to using a text similarity algorithm in natural language processing, and its value range is 0-1; the length of the question is determined by counting the number of characters in the user's question.

[0054] Specifically, the step of extracting keywords from the sentence in the user's question according to the comparison result between the clarity evaluation value P and the preset clarity evaluation value P0 includes:

[0055] If P>P0, the keyword extraction module extracts keywords from the user's question;

[0056] If P≤P0, the keyword extraction module extracts keywords and words and sentences associated with the keywords in the sentence in the user's question.

[0057] In the embodiment of the present invention, the preset clarity evaluation value is 0.8.

[0058] In an embodiment of the present invention, the keywords in the user's question include disease names, body parts, and symptoms; the words and phrases associated with the keywords include symptoms similar to the disease names. For example, if the user's question mentions "cold", then the words and phrases associated with "cold" including "fever", "flu", and "tonsillitis" should also be used as keywords; for example, if the user's question only mentions "stomach ache", then the disease that causes "stomach ache" should be used as a keyword.

[0059] In the embodiment of the present invention, based on the different comparison results between the clarity evaluation value of the user's question and the preset clarity evaluation value, different extraction methods are used for keywords, which can improve the accuracy of keyword extraction, optimize the utilization of system resources, avoid unnecessary waste, and lay a good foundation for improving the accuracy of the results.

[0060] Specifically, the result determination module searches for relevant information in a database according to the keywords extracted by the keyword extraction module to determine the preliminary result, including processing the relevant information to determine the preliminary result, and determining the eligibility of the preliminary result.

[0061] In an embodiment of the present invention, the database includes disease information, drug information, treatment information, a medical knowledge base, and historical question and answer information.

[0062] Specifically, the processing of the relevant information includes determining features in the relevant information to determine the screening of the relevant information according to a comparison result between the feature quantity M and a preset feature quantity M0.

[0063] In the embodiment of the present invention, the features include the name of the disease and the symptoms corresponding to the disease.

[0064] Specifically, the determining of screening of the relevant information according to the number of features M includes:

[0065] If M≤M0, the same information in the related information is filtered and de-duplicated.

[0066] If M>MO, then all features are associated with symptoms and the number of associated symptoms B is determined;

[0067] The relevant information after screening is determined to be the preliminary results.

[0068] In the embodiment of the present invention, the value of the preset feature number M0 is 20.

[0069] In the embodiment of the present invention, under the condition that the number B of the associated symptoms is less than or equal to the preset number B0, the preliminary result is directly output.

[0070] Specifically, under the condition that the number B of the associated symptoms is greater than a preset number B0, the preliminary result is determined to be unqualified.

[0071] Specifically, under the condition that the preliminary result is unqualified, similar symptoms in the associated disease are extracted as the preliminary result.

[0072] In the embodiment of the present invention, when the preliminary result is unqualified, the symptoms of the disease associated with each feature are extracted, and similar parts of these symptoms are determined as preliminary results.

[0073] Specifically, the user feedback module determines the accuracy evaluation value of the preliminary result according to the user feedback. In the embodiment of the present invention, the accuracy evaluation value Q is calculated according to the following formula:

[0074] Q=α×F+β×T

[0075] Among them, Q represents the accuracy evaluation value, F represents the user's satisfaction with the answer given by the system, T represents the time it takes for the system to give a reply, and α and β represent weight coefficients.

[0076] In the embodiment of the present invention, the user's satisfaction with the answer given by the system ranges from 0 to 100, and the weight coefficients α and β are 0.5 and 0.5 respectively.

[0077] Specifically, the adjustment module determines to adjust the extraction process of the keyword extraction module or the process of determining the preliminary result of the result determination module, including determining the adjustment process based on the difference ΔQ between the accuracy evaluation value Q and the preset accuracy evaluation value Q0 and the ratio of the preset accuracy evaluation value difference ΔQ0.

[0078] In the embodiment of the present invention, the value of the preset accuracy evaluation value Q0 is 80, and the value of the preset accuracy evaluation value difference ΔQ0 is 10.

[0079] Specifically, the adjustment process is determined according to the comparison result of the difference ΔQ between the accuracy evaluation value Q and the preset accuracy evaluation value Q0 and the preset accuracy evaluation value difference ΔQ0, including:

[0080] If ΔQ≤ΔQ0, the adjustment module adjusts the preset feature quantity M0 according to the adjustment coefficient X1.

[0081] If ΔQ>ΔQ0, the adjustment module adjusts the preset clarity evaluation value P0 according to the adjustment coefficient X2.

[0082] In the embodiment of the present invention, the adjustment coefficient X1 is The adjustment coefficient X2 is ΔQ represents the difference between the accuracy evaluation value Q and the preset accuracy evaluation value Q0. The adjusted preset feature quantity Mm=M0×X1, and the adjusted preset clarity evaluation value Pp=P0×X2.

[0083] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may be modified and varied in various ways.

[0085] manual

[0086] Any modifications, equivalent substitutions, improvements, etc. made should be included in the protection scope of the present invention.

Claims

1. A clinical intelligent question-answering system based on AI, characterized in that: include: A problem identification module is used to identify the user's problem and determine the user's needs; A semantic analysis module, connected to the question identification module, for determining the clarity of the user's question; A keyword extraction module, connected to the semantic analysis module, for determining the extraction of keywords from the sentence in the user's question according to the clarity; A result determination module, which is connected to the keyword extraction module and is used to search for relevant information in a database according to the keywords extracted by the keyword extraction module to determine a preliminary result; A user feedback module, which is connected to the result determination module and is used to determine the accuracy of the preliminary result according to user feedback; an adjustment module connected to the user feedback module, for adjusting the extraction process of the keyword extraction module or the process of determining the preliminary result of the result determination module according to the accuracy, and determining the adjustment process according to a comparison result of a difference between the accuracy evaluation value and a preset accuracy evaluation value and a difference between the preset accuracy evaluation values; The keyword extraction module determines the extraction of keywords from the sentence in the user's question according to the clarity, including determining a clarity evaluation value, so as to determine the extraction of keywords from the sentence in the user's question according to a comparison result between the clarity evaluation value and a preset clarity evaluation value; Wherein, under the condition that the clarity evaluation value is greater than a preset clarity evaluation value, extracting keywords from the sentence in the user's question; Under the condition that the clarity evaluation value is less than or equal to a preset clarity evaluation value, extracting keywords and words and sentences associated with the keywords in the sentence in the user's question; The clarity evaluation value P is determined according to the following formula: Among them, P represents the clarity evaluation value, N represents the number of keywords, K represents the relevance between keywords, and L represents the length of the question; The accuracy evaluation value Q is calculated according to the following formula: Among them, Q represents the accuracy evaluation value, F represents the user's satisfaction with the answer given by the system, T represents the time it takes for the system to give a reply, and α and β represent weight coefficients.

2. The AI-based clinical intelligent question-answering system according to claim 1, characterized in that: The result determination module searches for relevant information in a database according to the keywords extracted by the keyword extraction module to determine a preliminary result, including processing the relevant information to determine the preliminary result, and determining the eligibility of the preliminary result.

3. The AI-based clinical intelligent question-answering system according to claim 2, characterized in that: The processing of the relevant information includes determining features in the relevant information to determine the screening of the relevant information according to the number of the features.

4. The AI-based clinical intelligent question-answering system according to claim 3, characterized in that: The screening of the relevant information according to the number of features includes: When the number of features is less than or equal to the preset number of features, identical information in the related information is filtered and deduplication is performed; When the number of features is greater than a preset number of features, all features are associated with symptoms, and the number of associated symptoms is determined; The relevant information after screening is determined to be the preliminary results.

5. The AI-based clinical intelligent question-answering system according to claim 4, characterized in that: The result determination module determines that the preliminary result is unqualified under the condition that the number of the associated symptoms is greater than a preset number.

6. The AI-based clinical intelligent question-answering system according to claim 5, characterized in that: The result determination module extracts similar symptoms in the associated disease as the preliminary result under the condition that the preliminary result is unqualified.

7. The AI-based clinical intelligent question-answering system according to claim 4, characterized in that: The step of determining the adjustment process according to the comparison result of the difference between the accuracy evaluation value and the preset accuracy evaluation value and the difference between the preset accuracy evaluation values ​​comprises: Under the condition that the difference between the accuracy evaluation value and the preset accuracy evaluation value is less than or equal to the preset accuracy evaluation value difference, adjusting the preset number of features according to the adjustment coefficient; Under the condition that the difference between the accuracy evaluation value and the preset accuracy evaluation value is greater than the preset accuracy evaluation value difference, the preset clarity evaluation value is adjusted according to the adjustment coefficient.

Images