[summary of the invention]
Based on this, be necessary to provide a kind of system that represents electronic health record, can carry out the implicit reasoning inquiry to electronic health record, to present its interested electronic health record information to the concrete disease and/or the medical personnel of main suit and different trainings.
A kind of system that represents electronic health record; Said system comprises: the standard management module; The software module that logical relation between electronic health record data standard predetermined data element and the said data element is carried out visual typing, editor and expansion is provided, according to the information generation data element document of typing; Template generator reads said data element document, generates the electronic health record template according to said data element document, and makes up medical knowledge document and search inference rule; Medical knowledge base is stored said medical knowledge document, search inference rule and electronic health record template; The electronic health record generation module; Select and load the electronic health record template according to training classification and/or patient information; Create electronic health record user input interface according to the Data Dynamic in the said electronic health record template; Electronic health record data according to said electronic health record user input interface reception medical personnel typing combine the said electronic health record data data element corresponding with natural language information, generate the electronic health record data element that comprises patient's particular data; Said electronic health record data element is copied to the relevant position of electronic health record, generate electronic health record; The electronic health record storage unit is stored the electronic health record of said generation; The reasoning from logic engine; Obtain electronic health record according to patient information; Disease and/or main suit according to patient call the medical knowledge document from said medical knowledge base; From said medical knowledge base, call search inference rule according to patient's disease and/or main suit and medical personnel's training classification and/or type of service; Utilize search inference rule of being obtained and the logical relation that is included between the data element in the said medical knowledge document that the information in the electronic health record is searched for and the formalization reasoning, obtain electronic health record data subset into training and/or business customizing; Display interface is used to show said electronic health record data subset.
Preferably; Said standard management module is further used for generating the computer representation of the logical relation between electronic health record data standard predetermined data element and the data element; Provide and said data element is carried out additions and deletions change the user interface of looking into; And set up the corresponding relation between natural language and the computing machine semanteme through structurized organizational form; And pass through the structure of data and the logical relation between the said data element of coded representation, and said electronic health record data standard predetermined data element is expanded.
Preferably; Said template generator comprises: electronic health record template generation module; Read data elements document from said standard management module represents the natural language information that comprises in the said data element document through graphical interfaces according to the institutional framework that meets medical logic, and the user is hidden coding and the semantic information of computing machine in the said data element document; On graphical interfaces with the corresponding visual back end of the data element in the said data element document; And receiving the operational order of user to said back end, the data element that said back end is corresponding copies to the correct position in the template file, generates the electronic health record template; The medical knowledge document makes up module; Definition in advance is used to describe medical concept, logical word and the structural description method of logical relation between the said data element; Reorganize according to structural description method pair data element subclass relevant and medical concept subclass through said graphical interfaces with specified disease or main suit; And connect with said medical logic speech, produce the medical knowledge document; Search inference rule is set up module; Definition in advance is used for said medical knowledge base and electronic health record are searched for logical word and operation speech with reasoning; Connect said logical word, operation speech through said graphical interfaces according to said structural description method; And use the medical concept and the data element that define in the medical knowledge document that the parameter in the logical word is carried out instantiation, set up search inference rule to specific training and medical profession.
Further preferably; Said search inference rule is set up the predefined logical word of module and is comprised basic logic speech and combinator speech; Said search inference rule is set up the combinator speech that module is further used for utilizing predefined basic logic speech, operation speech and has defined, makes up said combinator speech through medical concept in said medical knowledge base of structured connection and the electronic health record and medical terminology.
Preferably, it is regular that said display interface also is used for defining in advance the demonstration that is used for according to search and the specific semantic marker of The reasoning results the search The reasoning results being carried out special demonstration.
In addition, also be necessary to provide a kind of method that represents electronic health record, can carry out the implicit reasoning inquiry to electronic health record, to present its interested electronic health record information to the concrete disease and/or the medical personnel of main suit and different trainings.
A kind of method that represents electronic health record; Said method comprising the steps of: through software module between electronic health record data standard predetermined data element and the said data element logical relation carry out visual typing, editor and expansion, generate the data element document; Read said data element document, generate electronic health record template, medical knowledge document and search inference rule, and store said electronic health record template, medical knowledge document and search inference rule through visual graphical interfaces; Create electronic health record user input interface according to the Data Dynamic in the said electronic health record template; Receive the electronic health record data of medical personnel's typing according to said electronic health record user input interface; The said electronic health record data data element corresponding with natural language information combined; Generation comprises the electronic health record data element of patient's particular data, and said electronic health record data element is copied to the relevant position of electronic health record, generates electronic health record and storage; Obtain electronic health record according to patient information; Disease and/or main suit according to patient call the medical knowledge document; Call medical knowledge document and search inference rule according to patient's disease and/or main suit and medical personnel's training classification and/or type of service; The logical relation of utilizing the said search inference rule of obtaining and being included between the data element in the said medical knowledge document is searched for and the formalization reasoning the information in the electronic health record, obtains the electronic health record data subset into training and/or business customizing; Show said electronic health record data subset.
Preferably, the step of said generation data element document comprises: generate the computer representation of the logical relation between electronic health record data standard predetermined data element and the said data element, provide and said data element is carried out additions and deletions change the user interface of looking into; Set up the natural language of electronic health record and the corresponding relation between the semantic description of computing machine through structurized organizational form; Predetermined data element in the electronic health record data standard is expanded; Generate the data element document through said user interface.
Preferably; The step of said generation electronic health record template is specifically: read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document; Receive the operational order of user to said back end, the data element that said back end is corresponding copies to the correct position in the template file, generates the electronic health record template.
Preferably; The step of said structure medical knowledge document is specifically: read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document; Definition in advance is used to describe medical concept, logical word and the structural description method of logical relation between the said data element; Receive user's operational order, reorganize according to structural description method pair data element subclass relevant and medical concept subclass through said graphical interfaces, and connect, produce the medical knowledge document with said medical logic speech with specified disease or main suit.
Preferably; The step of said foundation search inference rule is specifically: read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document; Definition in advance is used for said medical knowledge base and electronic health record are searched for logical word and operation speech with reasoning, and logical word and operation speech also show as visual back end on graphical interfaces; Through said graphical interfaces according to said structural description method with medical concept in the said medical knowledge document and the said logical word of term instantiation in the data element,, set up search inference rule to specific training and medical profession.
Further preferably; Said predefined logical word comprises basic logic speech and combinator speech; The step that makes up the combinator speech is specifically: the combinator speech that utilizes predefined basic logic speech, operation speech and defined; Through medical concept in the said medical knowledge base of structured connection and the medical terminology in the said data element, make up said combinator speech.
Preferably, the step of said foundation search inference rule also comprises: in the search reasoning process, the certain semantic information in the Search Results is carried out mark, and define the demonstration rule of Search Results being carried out special demonstration according to semantic marker in advance.
Preferably, the step of said demonstration electronic health record data subset comprises: when having set up said demonstration rule, said electronic health record data subset is shown according to the demonstration rule in the said search inference rule.
The above-mentioned system and method that represents electronic health record; Through setting up medical knowledge base and search inference rule; Can realize implicit reasoning inquiry, can make full use of the standardized structural information that is included in the electronic health record and carry out complicated reasoning based on semanteme to disease and scene; The part of search reasoning all is an electronic health record itself according to the data source of (like disease, state, time, process etc.) and search reasoning, and the logical source of search reasoning is in structurized medical knowledge document and search inference rule.Wherein, The content reaction of medical knowledge document notion and the rules with disease association such as physiology, pathology, pharmacology; It is relevant with specific disorders and/or main suit; And irrelevant with medical personnel's training and concrete medical profession, thereby can to set up be the medical knowledge document that is used for all doctors and healthcare structure.Search inference rule is used for describing to specific disorders, main suit, and specific specialist is interested electronic health record content in specific medical scene.The doctor interested electronic health record content often can have the combinator construed of parameter with minority, each such combinator speech has corresponding natural language expressing form.For example sick drug effect and the spinoff of XX, the complication time of occurrence that XX is sick etc.Wherein XX is a parameter.These combinator speech can be connected with corresponding data element or medical concept through some basic logic speech, operation speech and the small grain size combinator speech that has defined and define, and make it to have the search and the reasoning semanteme that can be computer understanding.Can be independent of concrete specialist, scene, disease or main suit to the definition procedure of these minority combinator speech and carry out, thereby can accomplish by the developer.Can be reduced to the process that this minority combinator speech is reconfigured and parameter carried out instantiation to the process of setting up of specific training user and scene and specified disease and/or main suit's search inference rule.Since each doctor in daily diagnosis and treatment process institute often the quantity such as disease, main suit, scene of processing are often less and relatively stable; Thereby can be when system deployment; Help to use visual natural language interface to accomplish the reorganization and the instantiation of above-mentioned combinator speech down the engineering staff by the doctor, generate search and inference rule that particular doctor or section office use in day to day operation.Because these search inference rules comprise being used for of can be that computing machine understands the electronic health record of ad hoc structure and computing machine semanteme and medical knowledge document are searched for the instruction with reasoning; Computing machine can use these rules to carry out the formalization reasoning; Implicitly excavate out the medical logic between the electronic health record record traditional Chinese medicine fact, this process does not require that medical personnel have deep understanding to the storage organization of data.In addition; The medical knowledge document is organized according to specific disorders and main suit; Search inference rule then customizes according to specific disorders and main suit to specific specialist and scene; These organizational forms all meet the custom of doctor's clinical thinking and the exchanges data characteristics in the clinical business, and information such as relevant disease, main suit, doctor's training, scene can clearly obtain in medical procedures, thereby this method has good support to clinical business.Because system can show particular doctor interested electronic health record content under special scenes according to organizational form that meets clinical thinking and the demonstration rule with logical expression ability; Medical personnel can obtain patient's various diagnosis and treatment relevant informations fast; Improved accurate rate of diagnosis, reduced mistaken diagnosis and failed to pinpoint a disease in diagnosis mistake and control the incidence of Lou controlling.
[embodiment]
Fig. 1 shows the system that represents electronic health record among the embodiment, and this system comprises standard management module 10, template generator 20, medical knowledge base 30, electronic health record generation module 40, electronic health record storage unit 50, reasoning from logic engine 60, patient management module 70, user management module 80 and display module 90.Wherein:
(1) standard management module 10 is used to provide the software module that the logical relation between electronic health record data standard predetermined data element and the said data element is carried out visual typing, editor and expansion, according to the information generation data element document of typing.The electronic health record data standard can be the various electronic health record standards of having stipulated; For example health ministry has been issued " electronic health record basic framework and data standard (exposure draft) " August in this year, and the term that wherein should follow electronic health record, coding and storage organization have been made regulation.In one embodiment; Standard management module 10 performed functions specifically comprise: (1) generates the computer representation of the logical relation between electronic health record data standard predetermined data element and the data element, provides said data element is carried out additions and deletions to change the user interface of looking into; (2) set up the corresponding relation between natural language and the computing machine semanteme through structurized organizational form, and through the structure of data and the logical relation between the said data element of coded representation; (3) electronic health record data standard predetermined data element is expanded, comprised self-defining metadata and the semantic information of computing machine of increasing.
Fig. 2 A shows the user interface the user clicks the interpolation top mode in the standard management device after, dynamically generates the data element that corresponding data element input interface is used for the standard code of typing electronic health record in right panel.Click ACK button behind user's typing corresponding information, will produce interface shown in Fig. 2 B.Visible from Fig. 2 B, the data element information of typing among Fig. 2 A has produced a new node (being designated as red node) at Left-Hand Panel.Fig. 2 C shows the user adds the subdata node element in the standard management device interface.The user selects a data node (the current node of choosing of red node representative) at the template manager Left-Hand Panel, clicks then and adds the child node button, will dynamically produce interface shown in Fig. 2 C right panel.When the user selects " floating number " option in the corresponding combo box of Fig. 2 C right panel " value type " label, will dynamically produce interface shown in Fig. 2 D right panel.Shown in the mark among Fig. 2 D, the typing item that dynamically produces had both comprised the electronic health record standard-required " floating number " data element of type requires the information (like unit) of typing, also comprises the information of system extension, like ULN, normal lower limit etc.When user's typing corresponding information and click confirming button in the interface of Fig. 2 D; Template manager is with interface shown in the displayed map 2E; Can find out that from Fig. 2 E the data element of typing has produced a new node (being designated as red node) at the Left-Hand Panel of Fig. 2 E among Fig. 2 C and Fig. 2 D.
(2) template generator 20 is used for the read data elements document, generates the electronic health record template according to the data element document, and makes up medical knowledge document and search inference rule.As shown in Figure 3, in one embodiment, template generator 20 comprises that electronic health record template generation module 210, medical knowledge document make up module 220 and search inference rule is set up module 230.Wherein:
Electronic health record generation module 210 is used for from standard management module 10 read data elements documents; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document; On graphical interfaces with the corresponding visual back end of the data element in the said data element document; And receiving the operational order of user to said back end, the data element that said back end is corresponding copies to the correct position in the template file, generates the electronic health record template to various disease and training;
Fig. 4 shows among the embodiment interface that generates the electronic health record template, and wherein, the form through tree-shaped or tabulation represents the natural language information of the data element of electronic health record standard code.In the panel in Fig. 4 left side; The corresponding data elements (comprising corresponding medical terminology, coding, metadata etc.) of the back end of each natural language, the institutional framework of the node tree of being formed have been reacted the logical relation between the different pieces of information element of electronic health record prescribed by standard; In the panel on right side; The visual structure that has represented the electronic health record template of this disease of the infection of the upper respiratory tract; Wherein each visual back end is still corresponding to a data elements, but be included in this template data element subclass and institutional framework all with on breathe that to infect this specified disease relevant.
Shown in Fig. 4 Left-Hand Panel; Electronic health record module generation module 210 has represented the natural language information that comprises in the data element through this graphical interfaces according to the institutional framework that meets medical logic, and the user has been hidden coding and the semantic information of computing machine that comprises in the data element.The user pulls and/or duplicates stickup through the data node is carried out mouse; Visual this back end is inserted into relevant position in the right panel; Simultaneously, the data element that this visual back end is corresponding copies to position suitable in the electronic health record template document.The user can generate the electronic health record template to various disease and training through the position of the selection of controlling visual node, the order that pulls and placement.In one embodiment; The user carries out mouse to the back end that is directed against specified disease or clinical business in Fig. 4 Left-Hand Panel and pulls or duplicate stickup; This process will cause that the corresponding data element of visual node is dissolved in the corresponding electronic health record template document of right panel; Thereby can create electronic health record template to specified disease or clinical business; The electronic health record template that is generated has kept the full detail of selected data element subclass, and has increased and specified disease and/or the relevant logical relation of clinical business, is structurized medical science document.
The medical knowledge document makes up module 220 and is used to make up the medical knowledge document to specified disease or main suit; It defines medical concept, logical word and the structural description method that is used to describe logical relation between the said data element in advance; Reorganize according to the subclass of the structural description method pair data element relevant and the subclass of medical concept through said graphical interfaces with specified disease or main suit; And connect with said medical logic speech, produce the medical knowledge document.The medical knowledge document is (for example to be used to be described under the context of specified disease, main suit, physiological system the logical relation between the related data elements; Under the context of flu; Symptoms such as generating heat, sneeze often appears; Cough, expectoration etc. appear in certain hour afterwards, thereby change has taken place the logical communication link between these symptoms under the prerequisite of flu).
In one embodiment; Make up the medical knowledge document and adopt similar graphical interfaces as shown in Figure 5; In the electronic health record standard the predetermined data element, the medical knowledge document make up module 220 defined in advance the medical concept (for example " complication ", " indication ", " spinoff " etc.) that is used to describe logical relation between the said data element, logical word (for example " and have ... performance ", " ... before the .. ... the time " etc.) and structural description method.These medical concept and logical word be also corresponding back end on this graphical interfaces; Carry out mouse through the user and pull or duplicate stickup; According to divisions such as specific disorders classification, main suit, physiological systems relevant data element subclass is reorganized according to predefined structural description method with relevant medical concept subclass; And utilize predefined logical word to carry out suitable connection, thereby can produce structurized medical knowledge document.
Search inference rule set up module 230 define in advance be used for to medical knowledge base 30 and electronic health record search for the logical word of reasoning (as " with ", " or ", " greater than " etc.) and operate speech (as " search ", " except " etc.); It adopts similar graphical interfaces shown in Figure 6; On graphical interfaces, search inference rule is set up 230 predefined logical words of module and the also corresponding different back end of operation speech.Search inference rule is set up module 230 and is connected predefined logical word, operation speech through this graphical interfaces according to the structural description method; And use the medical concept and the data element that define in the medical knowledge document that the parameter in the logical word is carried out instantiation, set up search inference rule (as shown in Figure 7) to specific training and medical profession.
In one embodiment; Search inference rule is set up module 230 predefined logical words and is comprised basic logic speech and combinator speech, wherein the basic logic speech be for example " with ", " or ", " greater than ", " less than " etc. play the word that is connected effect basically.Search plain inference rule and set up the combinator speech that module 230 is further used for utilizing predefined basic logic speech, operation speech and has defined; Through the medical concept of definition in the structurized connection medical knowledge base 30 and the term in the related data elements; Produce the definition of combinator speech; Parameter can be with in defined combinator speech, for example " searches the XX side effects of pharmaceutical drugs ", " search XX sick medicine " etc.The combinator speech of being set up is also corresponding different back end on graphical interfaces, and can be used for defining the more combinator speech of coarsegrain.Search inference rule is set up 230 of modules and through graphical interfaces different logical speech and operation time is made up and parameter instantiation, sets up the search inference rule to different trainings and medical profession.In another embodiment, search inference rule is set up module 230 and is also defined in advance the certain semantic information of search in the The reasoning results is marked, and is defined on the display interface demonstration rule of the semantic information of mark being carried out special demonstration.
(3) medical knowledge base 30 is used to store medical knowledge document, search inference rule and the electronic health record template of generation.In one embodiment; Medical knowledge base 30 comprises medical knowledge document storing unit 310, search inference rule storage unit 320 and electronic health record template storage unit 330, and wherein: medical knowledge document storing unit 310 storage medical knowledge documents make up the constructed structurized medical knowledge document of module 220; Search inference rule storage unit 320 memory search inference rules are set up the search inference rule that module 230 is set up; The electronic health record template that electronic health record template storage unit 330 store electrons case history template generation modules 210 are generated.
(4) electronic health record generation module 40 is used for selecting and load the electronic health record template according to training classification and/or patient information; Create electronic health record user input interface according to the Data Dynamic in the said electronic health record template; Receive the electronic health record data of medical personnel's typing according to said electronic health record user input interface; The said electronic health record data data element corresponding with natural language information combined; Generation comprises the electronic health record data element of patient's particular data, and said electronic health record data element is copied to the relevant position of electronic health record, generates electronic health record.
In a by way of example; Electronic health record generation module 40 generates engine through input interface the electronic health record template of selecting is resolved; According to natural language information dynamic creation electronic health record user input interface wherein, receive the electronic health record data of medical personnel's typing through this user's input interface.Supply the form of typing to comprise quantitative typing (like the numerical value check result), sxemiquantitative typing (like feminine gender, the positive etc.), template or selection typing (as from limited option, selecting symptom, sign, medication, drug effect, complication, spinoff etc.) and free typing (by the own fill message of medical personnel etc.).The corresponding data element of natural language information combines on the electronic health record data of typing and the interface, produces the electronic health record data element and also copies in the electronic health record of generation.Medical personnel also can mark reinforcement (for example explain that some symptoms be difficult to explain, significant etc.) to certain some medical record information.
(5) electronic health record storage unit 50 is used to store the electronic health record of said generation.
(6) reasoning from logic engine 60 is used for obtaining electronic health record according to patient information; Disease and/or main suit according to patient call the medical knowledge document from medical knowledge base 30; From medical knowledge base 30, call search inference rule according to patient's disease and/or main suit and medical personnel's training classification and/or type of service; Search inference rule that utilization is obtained and the logical relation that is included between the data element in the medical knowledge document are searched for and the formalization reasoning the information in the electronic health record, obtain the electronic health record data subset to specified disease and/or main suit into training and/or business customizing.In one embodiment; Reasoning from logic engine 60 recognizes that the doctor is interested in patient's diabetes; And recognize in the medical knowledge document that from medical knowledge base 30, calls that the common medicine of treatment of diabetes has 300 kinds; In patient's electronic health record, find it and used 5 kinds of medicines wherein; Then variations such as the medical science fact retrieval symptom that electronic health record write down before and after these 5 kinds of medicines of 60 pairs of uses of reasoning from logic engine, sign, auxiliary examination; And search have or not with the medical knowledge document in the drug effect, the content that spinoff is complementary that write down, thereby according to effective, invalid, spinoff arranged, have no side effect, dosage reason of changes etc., represent relevant electronic health record data according to certain visual cues and logical organization.
(7) patient management module 70 is used to manage patient information.Comprise patient's essential information (like name, age etc.), at the information of registering of hospital (like training classification that patient hung, kinds of Diseases etc.).
(8) user management module 80 is used for managing user information.Comprise that medical personnel belong to section office, affiliated training classification etc.
(9) 90 of display interfaces are used for the electronic health record data subset that 60 search of display logic inference engine and reasoning obtain.In one embodiment, display interface 90 has defined in advance and has been used for carrying out the demonstration rule of special demonstration according to search and the specific semantic marker of The reasoning results to searching for The reasoning results.In search and reasoning process, search inference engine 60 adds predefined semantic marker according to the semanteme of search result content in Search Results.Display interface 60 shows the electronic health record data subset that obtains according to special semantic marker in the search inference rule and predefined demonstration rule.
Fig. 8 shows the method flow that represents electronic health record among the embodiment, and detailed process is following:
In step S10, through software module between electronic health record data standard predetermined data element and the said data element logical relation carry out visual typing, editor and expansion, generate the data element document.Fig. 9 shows the method flow that generates the data element document among the embodiment, and this method flow is specially:
In step S101, generate electronic health record data standard predetermined data element, and safeguard the logical relation between the said data element through the institutional framework and the coding of data, provide and said data element is carried out additions and deletions change the user interface of looking into.Fig. 2 A shows the user interface the user clicks the interpolation top mode in the standard management device after, dynamically generates the data element that corresponding data element input interface is used for the standard code of typing electronic health record in right panel.Click ACK button behind user's typing corresponding information, will produce interface shown in Fig. 2 B.Visible from Fig. 2 B, the data element information of typing among Fig. 2 A has produced a new node (being designated as red node) at Left-Hand Panel.Fig. 2 C shows the user adds the subdata node element in the standard management device interface.The user selects a data node (the current node of choosing of red node representative) at the template manager Left-Hand Panel, clicks then and adds the child node button, will dynamically produce interface shown in Fig. 2 C right panel.When the user selects " floating number " option in the corresponding combo box of Fig. 2 C right panel " value type " label, will dynamically produce interface shown in Fig. 2 D right panel.Shown in the mark among Fig. 2 D, the typing item that dynamically produces had both comprised the electronic health record standard-required " floating number " data element of type requires the information (like unit) of typing, also comprises the information of system extension, like ULN, normal lower limit etc.When user's typing corresponding information and click confirming button in the interface of Fig. 2 D; Template manager is with interface shown in the displayed map 2E; Can find out that from Fig. 2 E the data element of typing has produced a new node (being designated as red node) at the Left-Hand Panel of Fig. 2 E among Fig. 2 C and Fig. 2 D.
In step S102, set up the natural language of electronic health record and the corresponding relation between the semantic description of computing machine through structurized organizational form.Owing to stipulated constraint and scope, the correlation properties of input data in the corresponding medical terminology of basic terms collection (data element set) that electronic health record should use in writing, position (reaction between the different pieces of information unit) that different pieces of information unit should occur and each data element, coding, the case history etc. in the electronic health record standard in the electronic health record that generates in the classificatory logical relation of medical terminology.A part is the natural language information (like medical terminology) that supplies doctor's read-write in these information; Another partial information (like coding) is used to help computing machine that retrieval analyzed in the semanteme of electronic health record; Corresponding relation between this two has been done clear and definite regulation in the electronic health record standard, thereby can safeguard this corresponding relation through structurized computer representation.
In step S103, predetermined data element in the electronic health record data standard is expanded.
In step S104, generate the data element document through said user interface.
In step S20, read said data element document, generate the electronic health record template through visual graphical interfaces, and make up medical knowledge document and search inference rule, and store said electronic health record template, medical knowledge document and search inference rule.
Figure 10 shows the method flow that generates the electronic health record template among the embodiment, and this method flow detailed process is following:
In step S201; Read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document.In one embodiment, the interface that makes up the electronic health record template is as shown in Figure 4, and the form through tree-shaped or tabulation is with the logical relation between natural language demonstrating data element and the element.
In step S202, receive the operational order of user to said back end, the data element that said back end is corresponding copies to the correct position in the template file, generates the electronic health record template.As noted earlier; In one embodiment; Can pull and/or duplicate stickup through the data node is carried out mouse, visually corresponding data element copied to position suitable in the template document, the user can be directed against specified disease or clinical service selection back end; Be inserted into the relevant position of electronic health record template through the mode that pulls and/or the duplicate data element that back end is corresponding, and on user interface, produce the visualization display of the electronic health record template of upgrading.The electronic health record template that is generated has kept the full detail of selected data element subclass, and has increased and specified disease and/or the relevant logical relation of clinical business, is structurized medical science document.
Figure 11 shows the method flow that makes up the medical knowledge document among the embodiment, and detailed process is following:
In step S211; Read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document.In one embodiment, make up the medical knowledge document and can adopt graphical interfaces as shown in Figure 5.
In step S212, definition in advance is used to describe medical concept, logical word and the structural description method of logical relation between the said data element.Medical concept, logical word be also corresponding visual back end (as shown in Figure 6) on graphical interfaces.
In step S213; Receive user's operational order; Use said graphical interfaces to pull or duplicate the method pair data element subclass relevant such as stickup through mouse and the medical concept subclass reorganizes according to the structural description method with specified disease and/or main suit; Utilize said medical logic speech to connect, produce the medical knowledge document.
Figure 12 shows the method flow of setting up search inference rule among the embodiment, and detailed process is following:
In step S231; Read said data element document; Represent the natural language information that comprises in the said data element document according to the institutional framework that meets medical logic through graphical interfaces; The user is hidden coding and the semantic information of computing machine in the said data element document, on graphical interfaces with the visual back end of the correspondence of the data element in the said data element document.
In step S232, definition in advance is used for said medical knowledge base and electronic health record are searched for logical word and operation speech with reasoning.Predefined logical word comprises basic logic speech and combinator speech, wherein the basic logic speech be for example " with ", " or ", " greater than ", " less than " etc. play the word that is connected effect basically, also corresponding different back end on graphical interfaces.The combinator speech is to define through medical concept, the medical terminology in the related data elements and the combinator speech that has defined that connects definition in basic logic speech, the medical knowledge base 30, and it can be accomplished through graphical interfaces visually.The combinator speech that has defined also on graphical interfaces corresponding to corresponding data interface, be used to define new combinator speech or make up search inference rule.Parameter can be with in the combinator speech, and parameter is medical concept or term or their combination.
In step S233, synthetic logical word is made up to specific training, medical profession, specified disease or main suit through said graphical interfaces, and parameter instantiation is set up search inference rule.In one embodiment, when setting up search inference rule, also define the demonstration rule of the electronic health record data subset being carried out special demonstration according to the semantic marker in the search The reasoning results in advance.
The electronic health record template that generates, medical knowledge document and search inference rule are stored in the medical knowledge base 30.
In step S30; Create electronic health record user input interface according to the Data Dynamic in the said electronic health record template; Electronic health record data according to said electronic health record user input interface reception medical personnel typing combine the said electronic health record data data element corresponding with natural language information, generate the electronic health record data element that comprises patient's particular data; Said electronic health record data element is copied to relevant position, generation electronic health record and the storage of electronic health record.
In a by way of example; Electronic health record generation module 40 generates engine through input interface the electronic health record template of selecting is resolved; According to natural language information dynamic creation electronic health record user input interface wherein, receive the electronic health record data of medical personnel's typing through this user's input interface.Supply the form of typing to comprise quantitative typing (like the numerical value check result), sxemiquantitative typing (like feminine gender, the positive etc.), template or selection typing (as from limited option, selecting symptom, sign, medication, drug effect, complication, spinoff etc.) and free typing (by the own fill message of medical personnel etc.).The corresponding data element of natural language information combines on the electronic health record data of typing and the interface, produces the electronic health record data element and also copies in the electronic health record of generation.Medical personnel also can mark reinforcement (for example explain that some symptoms be difficult to explain, significant etc.) to certain some medical record information.The electronic health record that is generated can be stored in the electronic health record storage unit 50.
In step S40; Obtain electronic health record according to patient information; Training classification and/or type of service according to patient's disease and/or main suit, medical personnel are called medical knowledge document and search inference rule; The logical relation of utilizing the said search inference rule of obtaining and being included between the data element in the said medical knowledge document is searched for and the formalization reasoning the information in the electronic health record; Obtain electronic health record data subset, show said electronic health record data subset for training and/or business customizing.In one embodiment, defined the demonstration rule that is used to show the electronic health record data subset in advance, then shown the electronic health record data subset that obtains according to this predefined rule that shows.
Because electronic health record has write down patient's health and fitness information, is the data of patient's personalization; The medical knowledge document is independent of concrete patient, user and medical institutions, describes by the logical relation between the medical science fact of objective laws such as the physiology of disease, pathology, pharmacology decision, and be the Knowledge Base that the clinician carries out clinical thinking; Search inference rule then depends on the business characteristic of doctor when handling different diagnosis and treatment tasks of different trainings; Therefore searching for inference rule has determined which certain applications of medical knowledge in the search and reasoning of electronic health record; And it has stipulated how to use (i.e. the path of search and reasoning) to obtain the electronic health record data subset; Search inference rule depends on factors such as user's training classification and medical institutions, needs to carry out the personalization configuration according to user's training classification or medical institutions.Because medical profession has a lot of general character, thereby the medical profession routine can define on than coarsegrain.For example search " diabetes medicament is to the drug effect and the spinoff of given patient " and search " hypertension drug is to the drug effect and the spinoff of given patient ", its implementation procedure has very big similarity.Use the combinator speech of " the drug withdrawal reason that XX is sick " this band parameter, can set up the semantic search inference rule of corresponding calculated and be used to handle above-mentioned two processes.
Because electronic health record, medical knowledge document and search inference rule have all adopted the identical terminology that meets the electronic health record standard code, coding and structure, thereby aspect semantic, communicate.Electronic health record standard and medical knowledge document have comprised medical science true conceptual abstraction and relationship description through structuring, and it is that the reasoning from logic rule description method of setting up coarsegrain lays the first stone that these notions tie up to the semantic meaning representation method with the pass.The combinator speech of search in the inference rule specified and how to be used the subclass of specific notion and relation that the electronic health record data are searched for and reasoning.For example search " drug side-effect that XX is sick " this combinator speech; Comprise and search corresponding medical knowledge document to confirm all medicines and each side effects of pharmaceutical drugs of specified disease; The medicine that from patient's electronic health record record, finds this patient to use then, and search the symptom that has or not appearance and used drug side-effect to be complementary in the symptom description that makes medicament future trouble person and describe.In said process; Specified disease is included in the medical knowledge document to the constraint of hunting zone and has reacted doctor's thinking search procedure; What Search Results was presented to the user is not only a medical science fact, but diagnosis and treatment process, specialist's thinking logic and the case history physiology logic of disease previously.
In one embodiment; Use (like insurance company, be in hospital etc.) for some non-outpatient services, the medical knowledge document (like the medical knowledge document according to physiological system or organ-tissue) by other clue tissues also can be provided and search for inference rule accordingly to satisfy the corresponding business demand.Because these document form are relatively independent of application scenarios, can use these documents electronic health record is carried out periodic analysis and arrangement, and analysis result is saved in the electronic health record.When the user need understand patient's health status according to the organizational form of physiological system when writing system review or definite insurance strategy of case history in hospital (as), then can use this knowledge document and corresponding search inference rule that patient's electronic health record is analyzed.
The above embodiment has only expressed several kinds of embodiments of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to claim of the present invention.Should be pointed out that for the person of ordinary skill of the art under the prerequisite that does not break away from the present invention's design, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with accompanying claims.