CN108170677A - A kind of medical terms abstracting method and device - Google Patents
A kind of medical terms abstracting method and device Download PDFInfo
- Publication number
- CN108170677A CN108170677A CN201711448103.4A CN201711448103A CN108170677A CN 108170677 A CN108170677 A CN 108170677A CN 201711448103 A CN201711448103 A CN 201711448103A CN 108170677 A CN108170677 A CN 108170677A
- Authority
- CN
- China
- Prior art keywords
- model
- chapters
- sections
- document
- medical terms
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Abstract
This application discloses a kind of medical terms abstracting method and devices, after the free text in obtaining target electronic case history, information extraction can be carried out to the free text using the Information Extraction Model pre-established, obtain the medical terms set that described information extraction model is drawn into；Then, due to having pre-defined continuous relationship between certain medical terms, in the medical terms set, corresponding continuous relationship is established for pre-defined every group of medical terms for having continuous relationship.In this way, can emerge from the continuous relationship between the medical terms extracted, so as to accurately be sorted out to the medical terms extracted according to pre defined attribute, avoiding extraction result, there are deviations, the accuracy of result is extracted, and then can meet the needs of doctor is to free text-processing result so as to improve medical terms.
Description
Technical field
This application involves field of computer technology more particularly to a kind of medical terms abstracting methods and device.
Background technology
Case history is diagnosed a disease including door (urgency) to be gone through and inpatient cases, case history be medical worker to the generation of patient disease, development, turn
Return, the record of curative activities process such as checked, diagnosed, treated and collected data is concluded, arranges, is comprehensive
Analysis is closed, the patient medical health account write by defined form and requirement.Case history be both clinical practice work summary and
Be explore disease rule and handle medical tangle legal basis, case history to medical treatment, prevention, teaching, scientific research, hospital management etc. all
There is important role.
Electronic health record is also the medical record system of computerization or computer based patient record.It is to use electronic equipment
The medical records of digitized patient that (computer, health card etc.) preserves, manages, transmits and reappear replaces hand-written paper disease
It goes through, its content includes all information of paper case history.US National Institute for Medical Research is defined as：Electronic health record is to be based on
The electronic patient record of one particular system, the system provide data, warning, prompting and the clinic that user accesses complete and accurate
The ability of DSS.
The information included in electronic health record text is carried out structuring processing by information extraction technique, it is the same to become table
Organizational form.What is inputted in information extraction system is original electron case history text, output be set form information point.Letter
Breath point is extracted from various texts, is then integrated in the form of unified, here it is information extractions
Main task.The benefit that information is integrated in the form of unified is to facilitate inspection and compare, and passes through the information point after extraction
Big data analysis and machine learning are carried out, meets medical worker in research work to the needs of certain indexs, for example, needing
2015 to 2017 are counted using morbidity of the pectoralgia as the patient with angina pectoris that cardinal symptom is admitted to hospital between all patient with angina pectoris
Probability.
With the raising of medical worker scientific research level, the requirement to information point is also increasingly thinner, multi-level into practising medicine
The extraction for treating term increasingly becomes necessity of demand.At present, the information extraction of medical text, mostly using dictionary pattern matching and canonical
Expression formula extracts.
Dictionary pattern matching is the dictionary for establishing different attribute, and word in text is determined by the specific object of word in dictionary
Type, such as " influenza " is a disease word, and by the matching of disease dictionary, computer is known that " popular
Flu " is the word of disease category, and then can correctly segment.
Regular expression is by establishing different classes, and by analysis and the relevant byte of class, it is this to make corresponding byte
The attribute of a class, such as " lung can and moist rales " in medicine text, expression be by doctor's auscultation, can be in the lung of patient
Hear " moist rales " that the position attribute of this class of the output result and " moist rales " of regular expression is in portion：Lung, flag bit
Attribute for "Yes" (can and the meaning, a kind of in different contexts, the structure of expression output equivalent in meaning for standard).
Free text can export simple logical relation by dictionary pattern matching or regular expression, but expression medically is patrolled
Comparatively increasingly complex, such as to cough with expectoration relationship is collected, is matched according to medical dictionary, both of which is symptom, but is coughed
For the premise of expectoration, there is expectoration to have cough certainly, not vice versa.It can not be solved by the matching of dictionary and regular expression
The medically subsumption problem of some Words ' Attributes, such as will appear nausea,vomiting,diarrhea symptom after Chemotherapy of Tumor Patients, if
It is simple by dictionary pattern matching, then extract result and have deviation, and can't resolve demand of the doctor to free text-processing result.
Invention content
In view of this, the main purpose of the embodiment of the present application is to provide a kind of medical terms abstracting method and device, energy
Enough improve the accuracy that medical terms extract result.
In a first aspect, this application provides a kind of medical terms abstracting method, including：
Obtain the free text in target electronic case history；
Information extraction is carried out to the free text using the Information Extraction Model pre-established, wherein, described information is taken out
Modulus type includes at least two attribute models, and the attribute model is for extraction and the relevant at least one information of pre defined attribute
The corresponding medical terms of point；
Obtain the medical terms set that described information extraction model is drawn into；
If there is each medical art at least one set of medical terms and every group of medical terms in the medical terms set
Continuous relationship is pre-defined between language, then the undertaking established between each medical terms in every group of medical terms is closed
System, wherein, every group of medical terms include at least two medical terms.
Optionally, when the target electronic case history includes at least one document document and the document document is included at least
During one document section, described information extraction model further includes the chapters and sections mould for carrying out information extraction to the document section
Type；
Then, it is described that information extraction is carried out to the free text using the Information Extraction Model pre-established, including：
Using each document section in the target electronic case history as target chapters and sections；
The corresponding chapters and sections model of the target chapters and sections is called, called chapters and sections model is made to call at least two attributes mould
Attribute model in type carries out information extraction to the text message for belonging to the target chapters and sections in the free text.
Optionally, the text message to belonging to the target chapters and sections in the free text carries out information extraction, packet
It includes：
For belonging to the text message of the target chapters and sections in the free text, taken out according to writing main line into row information
It takes.
Optionally, the method further includes：
According to the electronic health record of different mode to the division result of document section, establish the electronic health record of each pattern with it is right
Answer the first correspondence between the chapters and sections model of document section；
Then, it is described to call the corresponding chapters and sections model of the target chapters and sections, including：
According to first correspondence, the corresponding chapters and sections model of the target chapters and sections is called.
Optionally, the method further includes：
At least one document section that the pre defined attribute occurred is counted in advance；
Establish the corresponding attribute model of pre defined attribute chapters and sections model corresponding at least one document section
Between the second correspondence；
Then, the attribute model that called chapters and sections model is made to call at least two attribute model, including：
According to second correspondence, called chapters and sections model is made to call the attribute at least two attribute model
Model.
Optionally, the chapters and sections model corresponds at least one document document；The method further includes：
At least one document document that the document section occurred is counted in advance；
Establish that third between the corresponding chapters and sections model of the document section and at least one document document is corresponding to close
System；
Then, it is described to call the corresponding chapters and sections model of the target chapters and sections, including：
According to the third correspondence, the corresponding chapters and sections model of the target chapters and sections is called.
Second aspect, this application provides a kind of medical terms draw-out device, including：
Text acquiring unit, for obtaining the free text in target electronic case history；
Information extracting unit, for being taken out using the Information Extraction Model pre-established to the free text into row information
It takes, wherein, described information extraction model includes at least two attribute models, and the attribute model is for extraction and pre defined attribute
The corresponding medical terms of relevant at least one information point；
Gather acquiring unit, for obtaining the medical terms set that described information extraction model is drawn into；
First establishing unit, if for there is at least one set of medical terms and every group of medical treatment art in the medical terms set
Continuous relationship has been pre-defined between each medical terms in language, then has established each medical art in every group of medical terms
Continuous relationship between language, wherein, every group of medical terms include at least two medical terms.
Optionally, when the target electronic case history includes at least one document document and the document document is included at least
During one document section, described information extraction model further includes the chapters and sections mould for carrying out information extraction to the document section
Type；
Then, described information extracting unit, including：
Target chapters and sections determination subelement, for using each document section in the target electronic case history as target chapter
Section；
Chapters and sections model calls subelement, for calling the corresponding chapters and sections model of the target chapters and sections；
Text Information Extraction subelement, for called chapters and sections model to be made to call the category at least two attribute model
Property model, to belong in the free text target chapters and sections text message carry out information extraction.
Optionally, the Text Information Extraction subelement, specifically for for belonging to the target in the free text
The text message of chapters and sections carries out information extraction according to writing main line.
Optionally, described device further includes：
Second establishes unit, for the electronic health record according to different mode to the division result of document section, establishes each
The first correspondence between the electronic health record of pattern and the chapters and sections model of corresponding document section；
Then, the chapters and sections model calls subelement, specifically for according to first correspondence, calling the target chapter
Save corresponding chapters and sections model.
Optionally, described device further includes：
First statistic unit, for counting at least one document section that the pre defined attribute occurred in advance；
Third establishes unit, for establishing the corresponding attribute model of the pre defined attribute and at least one document chapter
Save the second correspondence between corresponding chapters and sections model；
Then, the Text Information Extraction subelement, specifically for according to second correspondence, making called chapters and sections mould
Type calls the attribute model at least two attribute model.
Optionally, the chapters and sections model corresponds at least one document document；Described device further includes：
Second statistic unit, for counting at least one document document that the document section occurred in advance；
4th establishes unit, for establishing the corresponding chapters and sections model of the document section and at least one document document
Between third correspondence；
Then, the chapters and sections model calls subelement, specifically for according to the third correspondence, calling the target chapter
Save corresponding chapters and sections model.
A kind of medical terms abstracting method and device provided by the embodiments of the present application, in target electronic case history is obtained from
After text, information extraction can be carried out to the free text using the Information Extraction Model pre-established, obtain the letter
The medical terms set that breath extraction model is drawn into；Then, due to having pre-defined continuous relationship between certain medical terms,
Therefore, it in the medical terms set, establishes corresponding accept for pre-defined every group of medical terms for having continuous relationship and closes
System.In this way, it can emerge from the continuous relationship between the medical terms extracted, so as to according to pre defined attribute pair
The medical terms extracted are accurately sorted out, and avoid and extract result there are deviation, so as to improve medical terms extraction
As a result accuracy, and then can meet the needs of doctor is to free text-processing result.
Description of the drawings
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or it will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is the application
Some embodiments, for those of ordinary skill in the art, without creative efforts, can also basis
These attached drawings obtain other attached drawings.
Fig. 1 is a kind of one of method flow diagram of medical terms abstracting method that the embodiment of the present application one provides；
Fig. 2 is the drug model schematic that the embodiment of the present application one provides；
Fig. 3 is the symptom model schematic that the embodiment of the present application one provides；
Fig. 4 is the symptom model and effluent model schematic that the embodiment of the present application one provides；
Fig. 5 is the two of a kind of method flow diagram of medical terms abstracting method that the embodiment of the present application two provides；
Fig. 6 is one of present illness history model schematic that the embodiment of the present application two provides；
Fig. 7 is the two of present illness history model schematic that the embodiment of the present application two provides；
Fig. 8 is the chapters and sections model schematic of Different hospital different department that the embodiment of the present application three provides；
Fig. 9 is the correspondence schematic diagram between the chapters and sections model and attribute model that the embodiment of the present application four provides；
Figure 10 is the correspondence schematic diagram between the chapters and sections model that the embodiment of the present application four provides and document document；
Figure 11 is the apparatus structure block diagram of a kind of medical terms draw-out device that the embodiment of the present application five provides.
Specific embodiment
Purpose, technical scheme and advantage to make the embodiment of the present application are clearer, below in conjunction with the embodiment of the present application
In attached drawing, the technical solution in the embodiment of the present application is clearly and completely described, it is clear that described embodiment is
Some embodiments of the present application, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art
All other embodiments obtained without making creative work shall fall in the protection scope of this application.
Through applicants have found that, it is existing in free text abstracting method, mainly passing through dictionary pattern matching or canonical table
Simple logical relation is exported up to formula.But expression logic medically is comparatively increasingly complex, passes through dictionary and regular expressions
The matching of formula can not solve the subsumption problem of medically some Words ' Attributes, if simple by dictionary pattern matching, can cause
Result is extracted there are deviation, and also can't resolve demand of the doctor to free text-processing result.
To solve the above problems, the embodiment of the present application provides a kind of medical terms abstracting method and device, extracting
After medical terms in free text, pre-set continuous relationship between medical terms can be utilized, to the medical terms
It is extracted, avoiding extraction result, there are deviations, and the accuracy of result is extracted, and then can expire so as to improve medical terms
Podiatrist gives birth to the demand to free text-processing result.
The side that a kind of medical terms shown in the application exemplary embodiment are extracted below in conjunction with attached drawing 1 to attached drawing 10
Method is introduced.
Embodiment one
Referring to Fig. 1, for the method flow diagram of a kind of medical terms abstracting method that the embodiment of the present application one provides, this method
It may comprise steps of：
S101：Obtain the free text in target electronic case history.
In the present embodiment, it would be desirable to the electronic health record for carrying out medical terms extraction is known as target electronic case history, wherein, it is described
Electronic health record can be the medical records that patient is preserved, manages, transmitting, reappeared with electronic equipment, it can be understood as traditionally on paper
The digitized version of case history.
Free text in electronic health record can be the unstructured data in electronic health record, for example, free text can be with
Including master control, present illness history, past medical history, menstrual history, marriage childbearing history in admission records etc., can also include in audit report
It checks data and checks conclusion (such as word description part in electrocardiogram report).
S102：Information extraction is carried out to the free text using the Information Extraction Model pre-established, wherein, the letter
It ceases extraction model and includes at least two attribute models, the attribute model is relevant at least one with pre defined attribute for extracting
The corresponding medical terms of information point.
In the present embodiment, the pre defined attribute can be the attribute classified in advance based on medical knowledge, such as in advance
Defined attribute can be symptom, sign, inspection, inspection, drug, disease etc..It wherein, can with the relevant information point of pre defined attribute
To be the relevant information of the pre defined attribute, if for example, pre defined attribute be symptom, can be with the relevant information point of symptom
Symptom title, symptom inducement, symptom property etc..
The attribute model can be based on constructed by pre defined attribute and relative information point, it is and each
A attribute model corresponds to a pre defined attribute, in this way, attribute model can be utilized, free text is extracted, is obtained
To medical terms corresponding with the information point of pre defined attribute.For example, as shown in Fig. 2, when pre defined attribute is drug, it can be with
Based on drug and with the relevant information point of drug (such as the nomenclature of drug of the drug, dosage, the medication frequency, administration way
Diameter, medication duration etc.), drug model (i.e. attribute model) corresponding with drug is built, it is right so as to utilize the drug model
Free text is extracted, and obtains the relevant information for a certain drug.Certainly, drug model can be used as an attribute mould
Type, symptom model, sign model, testing model, inspection model, disease model etc. can also be respectively as a kind of attribute model.
After the completion of each attribute model structure, at least two attribute models can be built an Information Extraction Model.
Wherein, at least two attribute model can include the corresponding attribute model of different pre defined attributes, such as including above-mentioned medicine
At least two attribute models in product model, symptom model, sign model, testing model, inspection model, disease model etc..
After the free text in target electronic case history is got by step S101, it can be directed in Information Extraction Model
The corresponding pre defined attribute of each attribute model, using the attribute model in the free text text message carry out
It extracts, obtains medical terms corresponding with the relevant at least one information point of the pre defined attribute.
Next, the symptom model with reference to included by Information Extraction Model shown in Fig. 3, illustrates how to utilize information
Extraction model carries out information extraction to free text.Assuming that a free text message in the target electronic case history got is
" breathing problem leads to throat dry cough, and cough continues one week, and cough continues two days with expectoration, yellow sputum, expectoration ".Then
The relevant information that symptom is cough can be extracted from the free text first with the symptom model in the Information Extraction Model,
The symptom being drawn into is that the relevant information of cough is following five medical terms：The entitled cough of symptom, symptom inducement is exhales
Tract disease is inhaled, symptom position is throat, and symptom duration is one week, and simultaneous phenomenon is expectoration.Then, the information extraction is utilized
Symptom model in model, then the relevant information that symptom is expectoration is extracted from the free text, the symptom being drawn into is cough
The relevant information of phlegm is following three medical terms：The entitled expectoration of symptom, symptom property be yellow sputum, symptom duration
It is two days.
S103：Obtain the medical terms set that described information extraction model is drawn into.
Continue by taking the example of above-mentioned S102 as an example, it, can be with after each medical terms are drawn into using Information Extraction Model
Using all medical terms being drawn into as medical terms set.
S104：If there is each doctor at least one set of medical terms and every group of medical terms in the medical terms set
Continuous relationship has been pre-defined between treatment term, then has established the undertaking between each medical terms in every group of medical terms
Relationship, wherein, every group of medical terms include at least two medical terms.
In the present embodiment, before medical terms set is got, if existing in the medical terms set respectively for more
Multiple medical terms of a pre defined attribute, and this is directed between multiple medical terms of multiple pre defined attributes and determines in advance respectively
Justice continuous relationship can be this respectively for the multiple of multiple pre defined attributes then after medical terms set is got
Medical terms set continuous relationship.
If for example, as shown in figure 3, go out the doctor that symptom (i.e. pre defined attribute) is " cough " using symptom model extraction
Term is treated, and goes out the medical terms that symptom (i.e. pre defined attribute) is " expectoration " using symptom model extraction；Due to " cough
Cough " continuous relationship is pre-set between " expectoration ", therefore, continuous relationship is established for " cough " and " expectoration ", represents " cough
Phlegm " is the symptomatic consequence of " cough ".
In another example symptom model and effluent model included by Information Extraction Model as shown in Figure 4, if utilizing symptom
Model extraction goes out the medical terms that symptom (i.e. pre defined attribute) is " vomiting ", and goes out discharge using effluent model extraction
Object (i.e. pre defined attribute) is a medical terms of " gastric juice "；Due to a medical terms of the symptom for " vomiting " and the row
Go out object and pre-set continuous relationship between a medical terms of " gastric juice ", therefore, held for " vomiting " and " gastric juice " foundation
Relationship is connect, it is the symptomatic consequence of " vomiting " to represent " gastric juice ".
As it can be seen that the embodiment of the present application, after the free text in obtaining target electronic case history, can utilize what is pre-established
Information Extraction Model carries out information extraction to the free text, obtains the medical terms that described information extraction model is drawn into
Set；Then, due to having pre-defined continuous relationship between certain medical terms, in the medical terms set, be
Pre-defined every group of medical terms for having continuous relationship establish corresponding continuous relationship.In this way, the medical art extracted can be made
Continuous relationship between language emerges from, so as to accurately be returned to the medical terms extracted according to pre defined attribute
Class, avoiding extraction result, there are deviations, and the accuracy of result is extracted, and then can meet doctor so as to improve medical terms
Demand to free text-processing result.
Embodiment two
It should be noted that compared with embodiment one, the present embodiment crunode introduces the S102 in embodiment one.
Referring to Fig. 5, for the method flow diagram of a kind of medical terms abstracting method that the embodiment of the present application two provides, this method
It may comprise steps of：
S501：Obtain the free text in target electronic case history.
In the present embodiment, target electronic case history can include at least one document document, for example the paperwork document can be with
For admission records, for the first time progress note, daily progress note, operation record, discharge record etc..Also, the paperwork document can be with
Including at least one document section, the document chapters and sections can be the text structure that different content is stated in document document, for example, working as
When document document is admission records, document section which is included can be chief complaint, present illness history, past medical history, family
History, individual are, menstrual history, marriage childbearing history etc..
It should be noted that S501 is similar to the S101 that above-described embodiment is a kind of, the phase in above-described embodiment one is referred to
It closes and introduces.
S502：Using each document section in the target electronic case history as target chapters and sections.
In the present embodiment, for ease of description, the document section for needing to carry out information extraction can be known as target chapters and sections.
S503：The corresponding chapters and sections model of the target chapters and sections is called, called chapters and sections model is made to call described at least two
Attribute model in attribute model carries out information extraction to the text message for belonging to the target chapters and sections in the free text.
The chapters and sections model can be based on constructed by a document section in target electronic case history, and each
Chapters and sections model corresponds to a document section.For example, when document document is admission records, it can be according to institute in the admission records
Comprising " present illness history " document section build corresponding present illness history model (i.e. chapters and sections model).It should be noted that chapters and sections
Model can also include main suit's model, past medical history model, marriage childbearing history model etc. other than present illness history model.
After determining target chapters and sections, chapters and sections model corresponding with the target chapters and sections can be first called, and utilize the chapters and sections model,
For the attribute model in the free text message recalls information extraction model in the target chapters and sections.Then, the attribute mould is utilized
Type carries out information extraction to the text message in the target chapters and sections, obtains medical art corresponding with the information point of pre defined attribute
Language.
Next, the present illness history model with reference to included by Information Extraction Model shown in fig. 6, illustrates how to utilize letter
It ceases extraction model and information extraction is carried out to free text.As shown in fig. 6, the Information Extraction Model includes a present illness history model
Eight attribute models such as (i.e. chapters and sections model) and mechanism model, symptom model, sign model.Assuming that target chapters and sections are target
" present illness history " in electronic health record, and the text message of target chapters and sections " present illness history " got for " skin surface neoplasm,
First People's Hospital carries out test alive and carries out laser therapy ".It can be first according to target chapters and sections " present illness history ", in the letter
It ceases in extraction model, calls the corresponding present illness history model of target chapters and sections " present illness history ", and call and be somebody's turn to do using the present illness history model
Eight attribute models such as mechanism model, symptom model, sign model in Information Extraction Model.Then, this eight attributes are utilized
Model carries out information extraction to the text message of the target chapters and sections " present illness history ", obtains following medical terms：Organization names are
One the People's Hospital, symptom property are skin surface neoplasm, examine entitled test alive, treat entitled laser therapy.
S504：Obtain the medical terms set that described information extraction model is drawn into.
S505：If there is each doctor at least one set of medical terms and every group of medical terms in the medical terms set
Continuous relationship is pre-set between treatment term, then establishes the undertaking between each medical terms in every group of medical terms
Relationship, wherein, every group of medical terms include at least two medical terms.
It should be noted that S504, S505 are identical with S103, S104 in above-described embodiment one, above-mentioned implementation is referred to
Related introduction in example one.
Since the text message of each document section in document document is write according to certain style of writing pattern
's.For example, the text message of the present illness history (i.e. document section) in admission records (i.e. document document) can be based on the time
Line is write, i.e., is to follow time order and function sequence for the relevant information of present illness history in the text message of present illness history
To be write.
Therefore, in a kind of realization method of the present embodiment, in S503 " to belonging to the target in the free text
The step of text message progress information extraction of chapters and sections ", can include：For belonging to the target chapters and sections in the free text
Text message, according to writing main line carry out information extraction.
Wherein, the writing main line can be the style of writing pattern of the text message of target chapters and sections.If for example, target chapters and sections
Text message follow time order and function sequence write, then the writing main line can be served as theme with the time.In another example
If the text message of the target chapters and sections is to follow the sequence of disease to be write, which can be based on disease
Line.
For example, the present illness history model included by Information Extraction Model as shown in Figure 7, the Information Extraction Model include
One present illness history model (i.e. chapters and sections model), a upper timing node correspond to eight, mechanism model, symptom model, sign model etc.
Attribute model, future time node also correspond to eight attribute models such as mechanism model, symptom model, sign model.Assuming that mesh
It is " present illness history " to mark chapters and sections, and writing main line was served as theme with the time, then when to the text messages of target chapters and sections " present illness history " into
When row information extracts, in the Information Extraction Model, the target chapters and sections can be called " existing first according to target chapters and sections " present illness history "
The corresponding present illness history model of medical history ".Then, the corresponding eight attribute moulds of a upper timing node are called using the present illness history model
Type, and information extraction is carried out to the text message of the target chapters and sections " present illness history " using this eight attribute models.Then, it recycles
The present illness history model calls corresponding eight attribute models of future time node, and using this eight attribute models to the target chapter
The text message for saving " present illness history " carries out information extraction.
As it can be seen that in the embodiment of the present application, after the medical terms in extracting each target chapters and sections, can will pre-define
Continuous relationship is established between the multiple medical terms for having continuous relationship so that the continuous relationship between the medical terms extracted obtains
To embody, so as to accurately be sorted out to the medical terms extracted according to pre defined attribute, extraction result is avoided
There are deviations, and the accuracy of result is extracted, and then can meet doctor to free text-processing knot so as to improve medical terms
The demand of fruit.
Embodiment three
Under normal conditions, electronic health record has certain pattern, and the electronic health record of different mode can lead to electronic health record
Also can be different to the division of document section, i.e., for the electronic health record of different mode, the structure of document section also can be different.Example
Such as, electronic health record used in different hospital or section office may be different, specifically, the note of being admitted to hospital of the Occupational Medicine of certain hospital
It can include " occupational disease history " this document section in record, and not to occupation in the admission records of other clinical departments of the hospital
The description of medical history；And in the admission records of another hospital, then the description of occupational disease history is put into the " personal of the admission records
In this document section of history ".
It therefore, can be according to the electronic health record for stroke of document section to realize the electronic health record for different mode
Point, corresponding model is neatly called, in a kind of realization method of the embodiment of the present application, can also be included the following steps：According to
The electronic health record of different mode establishes the electronic health record of each pattern and corresponding document section to the division result of document section
The first correspondence between chapters and sections model；Then, " the corresponding chapters and sections mould of the target chapters and sections is called in the S503 of embodiment two
The step of type ", can include：According to first correspondence, the corresponding chapters and sections model of the target chapters and sections is called.
For example, as shown in Figure 8, it is assumed that the admission records of B section office of A hospitals and D section office of C hospitals are (i.e. in electronic health record
Document document) be to the division result of document section：" present illness history ", " past medical history ", " progress note ", " diagnosis and treatment process ",
" physical examination " and " auxiliary examination ".Come with the admission records of B section office of A hospitals for example, it is possible, firstly, to for the note of being admitted to hospital
Each document section in record, determine chapters and sections model corresponding with the document chapters and sections, so as to obtain in the admission records
The corresponding present illness history model of document section, past medical history model, progress note model, diagnosis and treatment by model, physical examination model
With six chapters and sections models such as auxiliary examination model.Then, the first couple between the admission records and this six chapters and sections models is established
It should be related to.When need in the admission records text message carry out information extraction when, can according to the admission records and this six
The first correspondence between a chapters and sections model calls this six chapters and sections models, and goes out this using this six chapters and sections model extractions
Medical terms in admission records, so as to obtain the admission records structural data of B section office of A hospitals.Similarly, C hospitals D section office
The first correspondence between admission records and corresponding chapters and sections model, can equally establish according to above-mentioned steps.
As it can be seen that in the embodiment of the present application, can be built according to the electronic health record of different mode to the division result of document section
Vertical the first correspondence between the electronic health record of each pattern and the chapters and sections model of corresponding document section, so that can basis
First correspondence calls the corresponding chapters and sections model of the target chapters and sections, so as to obtain meeting demand, accurately
Extract result.
Example IV
Under normal conditions, the medical terms of same pre defined attribute are belonged in electronic health record can repeat repeatedly, for example,
The description of symptom can be appeared in simultaneously in the document sections such as " main suit ", " present illness history ", also, the description of symptom is corresponding pre-
Defined attribute is relatively-stationary.
Therefore, it to avoid the redundancy of Information Extraction Model, in a kind of realization method of the embodiment of the present application, can also wrap
Include following steps：At least one document section that the pre defined attribute occurred is counted in advance, establishes the predefined category
The second correspondence between the corresponding attribute model of property chapters and sections model corresponding at least one document section；Then, it is real
The step of applying " called chapters and sections model is made to call the attribute model at least two attribute model " in the S503 of example two
It can include：According to second correspondence, called chapters and sections model is made to call the category at least two attribute model
Property model.
For example, as shown in Figure 9, it is assumed that " symptom ", " sign ", " inspection ", " inspection ", " drug " and " disease " this six
A pre defined attribute occurred in " present illness history " and " past medical history " the two document sections.Come with " present illness history " document section
For example, first, then each pre defined attribute that can be directed in this six pre defined attributes determines and the predefined category
The corresponding attribute model of property, so as to obtain respectively with this corresponding symptom model of six pre defined attributes, sign model,
Six attribute models such as testing model, inspection model, drug model and disease model.Then, " present illness history " document section is established
The second correspondence between corresponding present illness history model (i.e. chapters and sections model) and this six attribute models.When needs are to " existing disease
It, can be according to corresponding " the present illness history mould of " present illness history " document section when text message in history " document section carries out information extraction
The second correspondence between type " and this six attribute models makes present illness history model call this six attribute models, and utilizes this
Six attribute models extract the medical terms in " present illness history " document section, so as to obtain present illness history structural data.Similarly,
The second correspondence between the corresponding past medical history model of " past medical history " document section and this six attribute models, equally can root
It is established according to above-mentioned steps.
As it can be seen that in the embodiment of the present application, at least one document that the pre defined attribute occurred can be counted in advance
Chapters and sections are established between the corresponding attribute model of pre defined attribute chapters and sections model corresponding at least one document section
The second correspondence so that the corresponding attribute mould of the pre defined attribute can be called according to second correspondence
Type so that attribute model can repeat to be called by chapters and sections model, avoids the redundancy of Information Extraction Model, and advantageous
In big data analysis.
Under normal conditions, same document section can be come across simultaneously in multiple document documents, such as " present illness history " document
In the documents documents such as chapters and sections can come across admission records simultaneously, twenty four hours enters discharge record, discharge record.Therefore, to keep away
Exempt from the redundancy of Information Extraction Model, in a kind of realization method of the embodiment of the present application, if chapters and sections model is corresponding at least one
Document document can also then include the following steps：At least one document document that the document section occurred is counted in advance,
Establish the third correspondence between the corresponding chapters and sections model of the document section and at least one document document；Then, it is real
The step of applying " calling the corresponding chapters and sections model of the target chapters and sections " in the S503 of example two can include：According to the third pair
It should be related to, call the corresponding chapters and sections model of the target chapters and sections.
For example, as shown in Figure 10, it is assumed that " present illness history ", " past medical history ", " progress note ", " diagnosis and therapy recording ", " physique
Inspection " and " auxiliary examination " this six document sections occurred in the documents document such as admission records and other medicine literature.
Come with admission records for example, it is possible, firstly, to for each document section in this six document sections, determined and this article
The corresponding chapters and sections model of shelves chapters and sections, so as to obtain respectively with this corresponding present illness history model of six document sections, the past
Six chapters and sections models such as history model, progress note model, diagnosis and therapy recording model, physical examination model and auxiliary examination model.It connects
It, establishes the third correspondence between the admission records and this six chapters and sections models.When needs are to the text in the admission records
When this information carries out information extraction, it can be adjusted according to the third correspondence between the admission records and this six chapters and sections models
With this six chapters and sections models, and go out using this six chapters and sections model extractions the medical terms in the admission records, so as to be somebody's turn to do
Admission records structural data.Similarly, the third correspondence between other medicine documents and this six chapters and sections models, equally may be used
To be established according to above-mentioned steps.
As it can be seen that in the embodiment of the present application, at least one document text that the document section occurred can be counted in advance
Shelves, establish the third correspondence between the corresponding chapters and sections model of the document section and at least one document document, with
Allow to according to the third correspondence, the corresponding chapters and sections model of the paperwork document is called, so that chapters and sections mould
Type can repeat to be called by the paperwork document, avoid the redundancy of Information Extraction Model, and be conducive to big data analysis.
Embodiment five
Based on a kind of medical terms abstracting method that above example provides, the embodiment of the present application additionally provides a kind of medical treatment
Its operation principle is described in detail below in conjunction with the accompanying drawings in terminology extraction device.
Referring to Figure 11, which is a kind of apparatus structure block diagram of medical terms draw-out device provided by the embodiments of the present application.
A kind of medical terms draw-out device provided in this embodiment, including：
Text acquiring unit 1101, for obtaining the free text in target electronic case history；
Information extracting unit 1102, for using the Information Extraction Model that pre-establishes to the free text into row information
It extracts, wherein, described information extraction model includes at least two attribute models, and the attribute model belongs to for extracting with predefined
The corresponding medical terms of the relevant at least one information point of property；
Gather acquiring unit 1103, for obtaining the medical terms set that described information extraction model is drawn into；
First establishing unit 1104, if for there is at least one set of medical terms and every group of doctor in the medical terms set
Continuous relationship has been pre-defined between each medical terms in treatment term, then has established each doctor in every group of medical terms
The continuous relationship between term is treated, wherein, every group of medical terms include at least two medical terms.
Optionally, when the target electronic case history includes at least one document document and the document document is included at least
During one document section, described information extraction model further includes the chapters and sections mould for carrying out information extraction to the document section
Type；
Then, described information extracting unit 1102, including：
Target chapters and sections determination subelement, for using each document section in the target electronic case history as target chapter
Section；
Chapters and sections model calls subelement, for calling the corresponding chapters and sections model of the target chapters and sections；
Text Information Extraction subelement, for called chapters and sections model to be made to call the category at least two attribute model
Property model, to belong in the free text target chapters and sections text message carry out information extraction.
Optionally, the Text Information Extraction subelement, specifically for for belonging to the target in the free text
The text message of chapters and sections carries out information extraction according to writing main line.
Optionally, described device further includes：
Second establishes unit, for the electronic health record according to different mode to the division result of document section, establishes each
The first correspondence between the electronic health record of pattern and the chapters and sections model of corresponding document section；
Then, the chapters and sections model calls subelement, specifically for according to first correspondence, calling the target chapter
Save corresponding chapters and sections model.
Optionally, described device further includes：
First statistic unit, for counting at least one document section that the pre defined attribute occurred in advance；
Third establishes unit, for establishing the corresponding attribute model of the pre defined attribute and at least one document chapter
Save the second correspondence between corresponding chapters and sections model；
Then, the Text Information Extraction subelement, specifically for according to second correspondence, making called chapters and sections mould
Type calls the attribute model at least two attribute model.
Optionally, the chapters and sections model corresponds at least one document document；Described device further includes：
Second statistic unit, for counting at least one document document that the document section occurred in advance；
4th establishes unit, for establishing the corresponding chapters and sections model of the document section and at least one document document
Between third correspondence；
Then, the chapters and sections model calls subelement, specifically for according to the third correspondence, calling the target chapter
Save corresponding chapters and sections model.
As seen through the above description of the embodiments, those skilled in the art can be understood that above-mentioned implementation
All or part of step in example method can add the mode of required general hardware platform to realize by software.Based on such
Understand, the part that the technical solution of the application substantially in other words contributes to the prior art can be in the form of software product
It embodies, which can be stored in storage medium, such as ROM/RAM, magnetic disc, CD, including several
Instruction is used so that computer equipment (can be personal computer, the network communications such as server or Media Gateway
Equipment, etc.) perform method described in certain parts of each embodiment of the application or embodiment.
It should be noted that each embodiment is described by the way of progressive in this specification, each embodiment emphasis is said
Bright is all difference from other examples, and just to refer each other for identical similar portion between each embodiment.For reality
For applying device disclosed in example, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related part
Referring to method part illustration.
It should also be noted that, herein, relational terms such as first and second and the like are used merely to one
Entity or operation are distinguished with another entity or operation, without necessarily requiring or implying between these entities or operation
There are any actual relationship or orders.Moreover, term " comprising ", "comprising" or its any other variant are intended to contain
Lid non-exclusive inclusion, so that process, method, article or equipment including a series of elements not only will including those
Element, but also including other elements that are not explicitly listed or further include as this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
Also there are other identical elements in process, method, article or equipment including the element.
The foregoing description of the disclosed embodiments enables professional and technical personnel in the field to realize or using the application.
A variety of modifications of these embodiments will be apparent for those skilled in the art, it is as defined herein
General Principle can in other embodiments be realized in the case where not departing from spirit herein or range.Therefore, the application
The embodiments shown herein is not intended to be limited to, and is to fit to and the principles and novel features disclosed herein phase one
The most wide range caused.
Claims (12)
1. a kind of medical terms abstracting method, which is characterized in that including：
Obtain the free text in target electronic case history；
Information extraction is carried out to the free text using the Information Extraction Model pre-established, wherein, described information extracts mould
Type includes at least two attribute models, and the attribute model is for extraction and the relevant at least one information point pair of pre defined attribute
The medical terms answered；
Obtain the medical terms set that described information extraction model is drawn into；
If exist in the medical terms set each medical terms at least one set of medical terms and every group of medical terms it
Between pre-defined continuous relationship, then establish the continuous relationship between each medical terms in every group of medical terms,
In, every group of medical terms include at least two medical terms.
2. according to the method described in claim 1, it is characterized in that, when the target electronic case history includes at least one document text
Shelves and the document document include at least one document section when, described information extraction model is further included for the document
Chapters and sections carry out the chapters and sections model of information extraction；
Then, it is described that information extraction is carried out to the free text using the Information Extraction Model pre-established, including：
Using each document section in the target electronic case history as target chapters and sections；
The corresponding chapters and sections model of the target chapters and sections is called, called chapters and sections model is made to call at least two attribute model
Attribute model, to belong in the free text target chapters and sections text message carry out information extraction.
It is 3. according to the method described in claim 2, it is characterized in that, described to belonging to the target chapters and sections in the free text
Text message carry out information extraction, including：
For belonging to the text message of the target chapters and sections in the free text, information extraction is carried out according to writing main line.
4. according to the method described in claim 2, it is characterized in that, the method further includes：
According to the electronic health record of different mode to the division result of document section, the electronic health record of each pattern and corresponding text are established
The first correspondence between the chapters and sections model of shelves chapters and sections；
Then, it is described to call the corresponding chapters and sections model of the target chapters and sections, including：
According to first correspondence, the corresponding chapters and sections model of the target chapters and sections is called.
5. according to claim 2 to 4 any one of them method, which is characterized in that the method further includes：
At least one document section that the pre defined attribute occurred is counted in advance；
It establishes between the corresponding attribute model of pre defined attribute chapters and sections model corresponding at least one document section
The second correspondence；
Then, the attribute model that called chapters and sections model is made to call at least two attribute model, including：
According to second correspondence, called chapters and sections model is made to call the attribute mould at least two attribute model
Type.
6. according to claim 2 to 4 any one of them method, which is characterized in that the chapters and sections model corresponds at least one
Document document；The method further includes：
At least one document document that the document section occurred is counted in advance；
Establish the third correspondence between the corresponding chapters and sections model of the document section and at least one document document；
Then, it is described to call the corresponding chapters and sections model of the target chapters and sections, including：
According to the third correspondence, the corresponding chapters and sections model of the target chapters and sections is called.
7. a kind of medical terms draw-out device, which is characterized in that including：
Text acquiring unit, for obtaining the free text in target electronic case history；
Information extracting unit, for carrying out information extraction to the free text using the Information Extraction Model pre-established,
In, described information extraction model includes at least two attribute models, and the attribute model is related to pre defined attribute for extracting
The corresponding medical terms of at least one information point；
Gather acquiring unit, for obtaining the medical terms set that described information extraction model is drawn into；
First establishing unit, if for existing at least one set of medical terms and every group of medical terms in the medical terms set
Each medical terms between pre-defined continuous relationship, then establish each medical terms in every group of medical terms it
Between continuous relationship, wherein, every group of medical terms include at least two medical terms.
8. device according to claim 7, which is characterized in that when the target electronic case history includes at least one document text
Shelves and the document document include at least one document section when, described information extraction model is further included for the document
Chapters and sections carry out the chapters and sections model of information extraction；
Then, described information extracting unit, including：
Target chapters and sections determination subelement, for using each document section in the target electronic case history as target chapters and sections；
Chapters and sections model calls subelement, for calling the corresponding chapters and sections model of the target chapters and sections；
Text Information Extraction subelement, for called chapters and sections model to be made to call the attribute mould at least two attribute model
Type carries out information extraction to the text message for belonging to the target chapters and sections in the free text.
9. device according to claim 8, which is characterized in that the Text Information Extraction subelement, specifically for for
Belong to the text message of the target chapters and sections in the free text, information extraction is carried out according to writing main line.
10. device according to claim 8, which is characterized in that described device further includes：
Second establishes unit, for the electronic health record according to different mode to the division result of document section, establishes each pattern
Electronic health record and the chapters and sections model of corresponding document section between the first correspondence；
Then, the chapters and sections model calls subelement, specifically for according to first correspondence, calling the target chapters and sections pair
The chapters and sections model answered.
11. according to claim 8 to 10 any one of them device, which is characterized in that described device further includes：
First statistic unit, for counting at least one document section that the pre defined attribute occurred in advance；
Third establishes unit, for establishing the corresponding attribute model of the pre defined attribute and at least one document section pair
The second correspondence between the chapters and sections model answered；
Then, the Text Information Extraction subelement, specifically for according to second correspondence, making called chapters and sections model tune
With the attribute model at least two attribute model.
12. according to claim 8 to 10 any one of them device, which is characterized in that the chapters and sections model corresponds at least one
A document document；Described device further includes：
Second statistic unit, for counting at least one document document that the document section occurred in advance；
4th establishes unit, for establishing between the corresponding chapters and sections model of the document section and at least one document document
Third correspondence；
Then, the chapters and sections model calls subelement, specifically for according to the third correspondence, calling the target chapters and sections pair
The chapters and sections model answered.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711448103.4A CN108170677B (en) | 2017-12-27 | 2017-12-27 | Medical term extraction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711448103.4A CN108170677B (en) | 2017-12-27 | 2017-12-27 | Medical term extraction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108170677A true CN108170677A (en) | 2018-06-15 |
| CN108170677B CN108170677B (en) | 2022-01-04 |
Family
ID=62518255
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711448103.4A Active CN108170677B (en) | 2017-12-27 | 2017-12-27 | Medical term extraction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108170677B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110674244A (en) * | 2019-08-20 | 2020-01-10 | 南京医渡云医学技术有限公司 | Structured processing method and device for medical text |
| CN111415751A (en) * | 2020-03-19 | 2020-07-14 | 北京嘉和海森健康科技有限公司 | Theme segmentation method, device and system for electronic medical record data |
| CN111429985A (en) * | 2020-03-02 | 2020-07-17 | 北京嘉和海森健康科技有限公司 | Electronic medical record data processing method and system |
| CN111681728A (en) * | 2020-06-09 | 2020-09-18 | 北京嘉和海森健康科技有限公司 | Content quality control method and device for electronic medical record |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080228040A1 (en) * | 2007-03-16 | 2008-09-18 | Arthur Solomon Thompson | International medical expert diagnosis |
| CN103530513A (en) * | 2013-10-10 | 2014-01-22 | 中国中医科学院 | Input system for achieving fast input of electronic medical record |
| CN105894088A (en) * | 2016-03-25 | 2016-08-24 | 苏州赫博特医疗信息科技有限公司 | Medical information extraction system and method based on depth learning and distributed semantic features |
| CN106897568A (en) * | 2017-02-28 | 2017-06-27 | 北京大数医达科技有限公司 | The treating method and apparatus of case history structuring |
| CN106897559A (en) * | 2017-02-24 | 2017-06-27 | 黑龙江特士信息技术有限公司 | A kind of symptom and sign class entity recognition method and device towards multi-data source |
| CN106933806A (en) * | 2017-03-15 | 2017-07-07 | 北京大数医达科技有限公司 | The determination method and apparatus of medical synonym |
-
2017
- 2017-12-27 CN CN201711448103.4A patent/CN108170677B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080228040A1 (en) * | 2007-03-16 | 2008-09-18 | Arthur Solomon Thompson | International medical expert diagnosis |
| CN103530513A (en) * | 2013-10-10 | 2014-01-22 | 中国中医科学院 | Input system for achieving fast input of electronic medical record |
| CN105894088A (en) * | 2016-03-25 | 2016-08-24 | 苏州赫博特医疗信息科技有限公司 | Medical information extraction system and method based on depth learning and distributed semantic features |
| CN106897559A (en) * | 2017-02-24 | 2017-06-27 | 黑龙江特士信息技术有限公司 | A kind of symptom and sign class entity recognition method and device towards multi-data source |
| CN106897568A (en) * | 2017-02-28 | 2017-06-27 | 北京大数医达科技有限公司 | The treating method and apparatus of case history structuring |
| CN106933806A (en) * | 2017-03-15 | 2017-07-07 | 北京大数医达科技有限公司 | The determination method and apparatus of medical synonym |
Non-Patent Citations (1)
| Title |
|---|
| 吴东: "基于潜在语义相关算法的电子病历检索的研究与应用", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110674244A (en) * | 2019-08-20 | 2020-01-10 | 南京医渡云医学技术有限公司 | Structured processing method and device for medical text |
| CN111429985A (en) * | 2020-03-02 | 2020-07-17 | 北京嘉和海森健康科技有限公司 | Electronic medical record data processing method and system |
| CN111429985B (en) * | 2020-03-02 | 2023-10-27 | 北京嘉和海森健康科技有限公司 | Electronic medical record data processing method and system |
| CN111415751A (en) * | 2020-03-19 | 2020-07-14 | 北京嘉和海森健康科技有限公司 | Theme segmentation method, device and system for electronic medical record data |
| CN111415751B (en) * | 2020-03-19 | 2023-08-08 | 北京嘉和海森健康科技有限公司 | Topic segmentation method, device and system for electronic medical record data |
| CN111681728A (en) * | 2020-06-09 | 2020-09-18 | 北京嘉和海森健康科技有限公司 | Content quality control method and device for electronic medical record |
| CN111681728B (en) * | 2020-06-09 | 2023-10-24 | 北京嘉和海森健康科技有限公司 | Content quality control method and device for electronic medical records |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108170677B (en) | 2022-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lindvall et al. | Natural language processing to assess end-of-life quality indicators in cancer patients receiving palliative surgery | |
| CN108170677A (en) | A kind of medical terms abstracting method and device | |
| Stearns et al. | SNOMED clinical terms: overview of the development process and project status. | |
| Aramaki et al. | Extraction of adverse drug effects from clinical records | |
| Coiera | Recent advances: medical informatics | |
| CN107845422A (en) | A kind of remote medical consultation with specialists session understanding and method of abstracting based on the fusion of multi-modal clue | |
| Roberts | Language, structure, and reuse in the electronic health record | |
| CN109313648A (en) | It is the system and method for hierarchy chart shape data structure by Document Modeling based on the semantic relation between the clinical concept being present in free text clinical document | |
| CN108320808A (en) | Analysis of medical record method and apparatus, equipment, computer readable storage medium | |
| Uhm et al. | The impact of a standardized inter‐department handover on nurses’ perceptions and performance in Republic of Korea | |
| Kim | Application of machine and deep learning algorithms in intelligent clinical decision support systems in healthcare | |
| Selvaraj et al. | Medication regimen extraction from medical conversations | |
| CN109448811A (en) | Checking prescription improved method, device, electronic equipment and storage medium | |
| US20220189486A1 (en) | Method of labeling and automating information associations for clinical applications | |
| Kim et al. | Information Extraction from Patient Care Reports for Intelligent Emergency Medical Services | |
| Cutler | New rules for the pandemic era | |
| Patel et al. | Automatic identification of self-reported COVID-19 vaccine information from vaccine adverse events reporting system | |
| Rink et al. | A supervised framework for resolving coreference in clinical records | |
| CN105205334A (en) | Doctor-patient data processing method | |
| Wang et al. | Constructing a Chinese medical conversation corpus annotated with conversational structures and actions | |
| CN108831560A (en) | A kind of method and apparatus of determining medical data attribute data | |
| Fairchild et al. | Eliciting disease data from Wikipedia articles | |
| CA3163953A1 (en) | Privacy preservation in a queryable database built from unstructured texts | |
| Ghadage et al. | Voice-Based Prescription Generation using Artificial Intelligence | |
| CN110070927A (en) | A kind of management system and application method of electronic health record |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190729 Address after: Room 2307, 3 storeys, No. 7 Pioneer Road, Shangdi Information Industry Base, Haidian District, Beijing 100085 Applicant after: Beijing Jiahesen Health Technology Co., Ltd. Address before: 100085 Haidian District city on the base of the information industry base, Pioneer Road, building No. 7, section I, layer three, layer Applicant before: Goodwill Information Technology Co., Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: 100085 room 2208, 2nd floor, building 1, No.7 Kaifa Road, Shangdi Information Industry base, Haidian District, Beijing Patentee after: Beijing Jiahesen Health Technology Co.,Ltd. Address before: 100085 room 2307, third floor, building 1, No.7 Kaifa Road, Shangdi Information Industry base, Haidian District, Beijing Patentee before: Beijing Jiahesen Health Technology Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |