CN107463786A - Medical image Knowledge Base based on structured report template - Google Patents
Medical image Knowledge Base based on structured report template Download PDFInfo
- Publication number
- CN107463786A CN107463786A CN201710707594.3A CN201710707594A CN107463786A CN 107463786 A CN107463786 A CN 107463786A CN 201710707594 A CN201710707594 A CN 201710707594A CN 107463786 A CN107463786 A CN 107463786A
- Authority
- CN
- China
- Prior art keywords
- report template
- medical image
- disease
- feature
- knowledge base
- 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.)
- Pending
Links
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The present invention relates to a kind of medical image Knowledge Base based on structured report template, belong to medical information processing and application field.The present invention is converted into the medical image of multi-modal storage the big data of structuring for the purpose of improving availability of data, realizes data innovation and increment, and providing precision data for medical research supports.Step of the present invention includes:The foundation of medical image feature noumenon, the standard for defining disease, exchanged as the design of the report template of element, the formation of knowledge element with information using structural data and learning rule foundation.The present invention builds the relation of diagnosis and Findings, recall precision is high, accuracy greatly enhances by the structure of ontology library, the definition of disease, text report template, picture and text report template and CDA documents;By the foundation of learning rule, constantly improve knowledge element is not only helpful to medical research, and is favorably improved the level of the medical image of the whole industry, has extremely extensive applying value.
Description
Technical field
The invention belongs to medical image information processing and application field, is related to a kind of medical image data and collects acquisition method,
Changed more particularly to a kind of based on the structuring that bulk form is carried out to medical image feature, the method for establishing knowledge base.
Background technology
As medical field enters " big data epoch ", incident is exactly following two problems.One problem, such as
What utilizes data Innovation Exploring medical science, how in huge data resource quick obtaining information is to lift medical level,
It is the realistic problem urgently inquired into.Another problem, with the implementation of " specification-precisely medical treatment " theory, personalized treatment
Specification and comprehensive assessment are carried out it is required that showing disease.
It is well known that maximum information source of the medical image as signs of disease, accounts for whole clinical medicine information data amounts
More than 80%, be disease maximum information source.As medical treatment enters big data epoch, the unstructured properties of medical image
It can not meet that big data analyzes requirement to the quality of data, and the knowledge acquisition method of traditional " stochastic analysis " or " sample investigation "
Total data resource can not effectively be utilized.At present, still need and want one kind to believe by standardizing flow and instrument medical image
Breath carries out structuring conversion, the method for establishing image evidence-based knowledge base, provides medical research accurately image data and supports.
The content of the invention
In view of the foregoing defects the prior art has, for the purpose of improving availability of data, the definition of body, disease are passed through
Structuring, crucial image extraction expression, the structure of medical ontology knowledge base, the medical image of multi-modal storage is converted into
" big data " of structuring, data innovation and increment are realized, providing precision data for medical research supports.
Medical image Knowledge Base of the present invention based on structured report template, comprises the following steps:
Step 1:The foundation of medical image feature noumenon:The classification of medical image feature noumenon includes anatomical structure, image
Feature, disease performance and diagnosis;
Step 2:Define the standard of disease:Using Consensus of experts, diagnosis guide as foundation, to the portion involved by report template
Position pre-defines disease that may be present and diagnosis, and disease performance and diagnosis are retrieved by code;
Step 3:Design using structural data as the report template of element:Report template include text report template and
Picture and text report template;
Step 4:The formation of knowledge element exchanges with information:Report template is stored according to CDA, by manually solving
Read, convert images into computer-readable feature;
Step 5:Learning rule is established:Input value using features described above as deep learning, is looked into using deep learning method
Look for imaging manifestations features and the relation of diagnostic result.
Preferably, in the step 1, the foundation of medical image feature noumenon includes following small step:
S1:List the entry involved by every one kind;
S2:Concluded and changed according to the build-in attribute of entry and exclusive feature, classification and level are established to entry
The disaggregated model of change;
S3:Unified coding is carried out to above-mentioned entry, adds related information;
S4:As needed, add example as concept tool as.
Preferably, the classification of the medical image feature noumenon, i.e., with coding mark, hierarchical relationship, subordinate relation, search
The standardization Chinese terminology in rope path, accurately to be retrieved using computer, wherein, dissection knot is defined according to analyzing system
The inheritance of structure;Interpretive classification is defined according to image feature;The disease table according to corresponding to disease performance classification by histoorgan
Now it is defined;ACR code databases and ICD-10 code databases are established to follow-up case in database according to diagnostic imaging conclusion respectively
Diagnosis encoded.
Preferably, in the step 2, the standard of disease, including following small step are defined:
S5:In each region of anatomy, detailed description of the disease foundation including Normal appearances to predefined, description
In the image feature that is related to use image feature term in medical image feature noumenon to be marked;
S6:When image is understood and writes report template, after doctor selects nomenclature of diseases corresponding to the position image, it is corresponding
Detailed description be automatically filled in;
S7:Doctor is appropriately modified to description content on this basis;
S8:What report template outward appearance showed is the word description of image feature, and implicit content includes the generation of disease
Position, nomenclature of diseases and coding, turn into searchable knowledge content;
S9:Above-mentioned report template content is cited in other document templates, reaches using limited report template to determine
All documents of justice.
Preferably, in the step 3, text report template includes scanning position, mode, the title of disease and code, sweeps
That looks into position is categorized as head, neck, chest, abdomen and four limbs, and mode is divided into CT, MRI and x-ray.
Preferably, in the step 3, picture and text report template includes quantizating index, crucial image and annotation of images;Quantify
Index includes the region of anatomy, measured value and number of quantitative analysis, and picture material is become to analyze the normal data of retrieval.
Preferably, in the step 4, knowledge element is built using CDA standards;During the foundation of knowledge element, root
Constructed according to CDA and XML standards, be a CDA documents per a iconography report template.
Preferably, in the step 5, deep learning uses enforcement mechanisms learning method, establishes multilayer neural network mould
Type, multilayer neural network model include input layer, hidden layer and output layer;
Input layer is:A=f (wTx+b)
Hidden layer is:Qj=f (∑ wij×xi-qj)
Output model is:Yk=f (Tjk×Oj-qk)
Wherein, f is non-linear action function, and q is the threshold value of neural unit, wijFor input node xiTo the shadow of hidden node
Ring weight;
Error calculating is the function for reflecting neutral net desired output and calculating error size between output, and it is calculated
It is as follows:
Wherein TpExported for network objectives;
The learning process of multilayer neural network model, that is, connect the weight matrix w between lower level node and upper layer nodeij's
Setting and error correction process.
The beneficial effects of the invention are as follows:Using the medical image knowledge base of the present invention based on structured report template
Method for building up, structuring conversion is carried out to medical image information by standardizing flow and instrument, establishes image evidence-based knowledge base,
Accurately image data is provided medical research to support.The present invention passes through text report template, picture and text report template and CDA texts
Shelves, the relation of diagnosis and Findings is built, recall precision is high, accuracy greatly enhances;By the foundation of learning rule,
Constantly improve knowledge element, it is not only helpful to medical research, and the level of the medical image of the whole industry is favorably improved,
With extremely extensive applying value.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of the present invention.
Fig. 2 is the inheritance figure of the anatomical structure of the present invention.
The image feature that Fig. 3 is the present invention defines interpretive classification figure.
Fig. 4 is the disease performance classification chart of the present invention.
Fig. 5 is the text report Prototype drawing of the present invention.
Fig. 6 is the picture and text report Prototype drawing of the present invention.
Fig. 7 is the CDA document architecture figures of the present invention.
Fig. 8 is the learning rule foundation figure of the present invention.
Embodiment
In order that the object of the invention, technical scheme are more clearly understood, with reference to embodiment, the present invention is made further
Describe in detail.
The present invention follows Medical Imaging feature, Chinese version body lexicon of the structure with inner link, according to international public
The Consensus of experts recognized, the specification that the common disease of different parts is defined using the body lexicon built describe;Using disease storehouse as
Basis, establish the handling process of specification and meet the structured report template of CDA standards, for being used to describe disease in template
Keyword and node noun be marked using above-mentioned unified term.In the writing process of image report template, pass through doctor
Raw deciphering, structured report template video conversion is appreciated that into computer, retrieved, set up KBS.Utilize depth
Spend learning method and data analysis is carried out to existing knowledge, conclude.
As shown in figure 1, the present invention is realized by following step:
Step 1:The foundation of medical image feature noumenon.Medical image feature noumenon (ontology) is to concept system
Clearly, specification formalize, sharable.Text envelope corresponding to image is determined in the presence of medical image feature noumenon
Breath, then foundation index is labeled to image from the processing method defined, the confirmation of knowledge is carried out by professional.This
Sample, the retrieval to medical image are just changed into the retrieval to concept.According to medical image feature, medical image feature noumenon includes
4 classes:Anatomical structure, image feature, disease performance and diagnosis.The construction step of medical image feature noumenon is:S1:Row
Go out the entry involved by every one kind;S2:Concluded and changed according to the build-in attribute of entry and exclusive feature, entry is built
Vertical class and hierarchical disaggregated model;S3:Unified coding is carried out to above-mentioned entry and adds related information;S4:As needed, add
Add example as concept tool as.
1.1 are divided into according to analyzing system:Nerve/neck, chest, heart and blood vessel, digestion, urogenital, skeletal muscle,
Mammary gland etc..By taking chest as an example, the inheritance (Fig. 2) of anatomical structure is defined.
1.2 define interpretive classification (Fig. 3) according to image feature.
1.3 show classification according to disease:Disease performance corresponding to histoorgan is defined.Such as in chest system, its
The nomenclature of diseases (Fig. 4) of middle lung and air flue.
1.4 diagnostic imaging conclusions:Carried out using the region of anatomy+pathology mode.ACR code databases are established respectively and ICD-10 is compiled
Diagnosis of the code storehouse to follow-up case in database encodes, and realizes the standardization of diagnosis.ACR code databases include dissection table and disease
Table is managed, each region of anatomy and pathological diagnosis are encoded using numeral in table, wherein dissection coding up to 4, pathology
Up to 6;ICD-10 code databases include classification chart and diagnosis coding table, classify in each table or diagnose with letter+numeral
Encoded.The template type selected according to diagnostician, it is preferential to import the list of diseases to match with template type, avoid people
Work selects the tedious steps of coding-belt.
Step 2:Using Consensus of experts, diagnosis guide as foundation, the standard scale for defining disease reaches:S5:In each region of anatomy
In, detailed description of the disease foundation including Normal appearances to predefined, the image feature being related in description is using doctor
Image feature term in image feature body is learned to be marked;S6:When being understood specific to image and writing report template, Yi Shengxuan
After selecting nomenclature of diseases corresponding to the position image, detailed description corresponding to it is automatically filled in;S7:Doctor is on this basis to description
Content is appropriately modified;S8:Now, what report template outward appearance showed is the word description of image feature, is wrapped in implicit content
The information such as position, nomenclature of diseases and the coding of the generation of disease are included, turn into searchable knowledge content;S9:Above-mentioned report template
Content can also be cited in other document templates, reach using limited report template to define all documents, as follows
Shown in table 1.
The disease list of table 1 is illustrated
Step 3:Report template using structural data as element designs:Image is established using XML syntax rules and standard
Report template.XML expresses information in a manner of a kind of structuring is based on text formatting, that is, meets the standard of medical document,
There is good autgmentability, interactivity again.XML document is structuring, and its each beginning label has corresponding end mark
Note, and these marks are nested in an orderly way.Because its is structural, XML document can be considered tree construction, and its node is by marking
And formed between mark and corresponding to the information of mark.Image report template is divided into two types:(1) text template report
Slide former, write for Routine report template, report template is established according to different images mode (CT, MR, x-ray and ultrasound);(2)
Picture and text template report template, use for accurate image quided, set according to the Consensus of experts of diagnostic imaging and report template purposes.
The foundation of 3.1 text report templates
It is that foundation establishes template according to scanning position and mode.Mode is divided into CT, MRI and x-ray.Scanning position according to head,
Neck, chest, abdomen, four limbs etc. are classified.In the Findings at each position, acquiescence inserts normal performance.By taking chest CT scans as an example,
Formwork style, as shown in Figure 5.
3.2 picture and text assessment report templates are established
To adapt to personalized diagnostic requirements, the knowledge form in storehouse of enriching one's knowledge, on the basis of foregoing picture and text report template,
By image measurement, the crucial image of reflection disease, annotation of images integrates, and forms a " accurate assessment report mould
Plate ".
Template is established and establishes that content is identical with text report template, and term is using the standard language in ontology knowledge base.Increase
Add image typing region.The content in typing region is limited in advance.As " cardiac shape " region limits user's typing reflection
The image of cardiac shape, as shown in Figure 6.To the imagery exploitation human interpretation of typing, retouched using standardization form technics
State, solve the problems, such as that computer can not understand picture material, in favor of the extraction and data preparation of knowledge.
Step 4:The formation of knowledge element exchanges with information:A iconography signed and issued according to above-mentioned report template is reported
Template represents a knowledge element.Knowledge Element is built using CDA (Clinical Document Architecture) standard
Element.
CDA is a part for HL7 third editions standard (HL7V3), the standardization of specific provision clinical document content.Due to
CDA complies fully with XML standards, therefore image knowledge can be packaged, and can effectively realize the scalability of knowledge base
And platform-neutral.As it was previously stated, XML document is structuring, each beginning label has corresponding end mark, and this
A little marks are nested in an orderly way.Because its is structural, XML document can be considered tree construction, its node by marking and
Formed between mark and corresponding to the information of mark.During the foundation of template, structure is carried out according to CDA and XML standards
Make, be a CDA documents per a iconography report template.
Fig. 7 is the basic building block of CDA frameworks.All components are in fact all RIM models (the one of HL7V3 standards
Part).The second layer (L2) CDA documents include one or more chapters and sections, and they use the structure of composite mode.One chapters and sections can be with
Comprising sub- chapters and sections, sub- chapters and sections include sub- chapters and sections again.The content of each chapters and sections is the descriptive expression encoded by XML, including
Anatomical structure, disease performance, measured value, chart etc., all key contents carry out assignment using ontology knowledge base coding.Calculate
Machine is encoded by following the trail of, it is possible to understand that the content topic described in chapters and sections, and traced.
Below, illustrate a knowledge element part CDA documents it is as follows:
Step 5:Learning rule is established:By above-mentioned form, the artificial solution that the image information of complexity can be passed through specification
Read, obtain the disease performance of standard, form the data that computer is appreciated that study., can when data accumulation is to certain amount
To apply deep learning method, disease and the relation of diagnosis are obtained.
Using enforcement mechanisms learning method, establish a kind of multilayer neural network model and learnt, it is anti-that its structure belongs to band
The feedforward network structure of feedback, model topology structure include input layer, hidden layer and output layer three-decker.Including input and output mould
Type, action function model, error calculating and self learning model.
Wherein, input layer is:
A=f (wTx+b)
Hidden layer is:
Qj=f (∑ wij×xi-qj)
Output layer is:
Yk=f (Tjk×Oj-qk)
Wherein, f is non-linear action function, and q is the threshold value of neural unit, wijFor input node xiTo the shadow of hidden node
Ring weight.
Error calculating is the function for reflecting neutral net desired output and calculating error size between output, and it is calculated
It is as follows:
Wherein TpExported for network objectives.
In the learning process of neutral net, that is, connect the weight matrix w between lower level node and upper layer nodeijSetting and
Error correction process.Decline (stochastic gradient decent) method using gradient to be approached.
By taking prostate cancer magnetic resonance case as an example, prostate T2WI, DWI, DCE of the existing proved by pathology of some examples are obtained
Mp-MRI, wave spectrum image measured value, and the age of patient and PSI desired values, the input as training sample.The phase of setting
Prestige value is diagnosis classification, is scored using PI-RADSv2:(prostate for having clinical meaning is not present in PIRADS1-extremely low
Cancer), PIRADS2-low (prostate cancer that there's almost no clinical meaning), PIRADS3-medium (with there is clinical meaning
Corresponding relation is failed to understand between prostate cancer), PIRADS4-height is (corresponding with there may be between the prostate cancer for having clinical meaning to close
System), PIRADS5-high (prostate cancer for having clinical meaning).
With the conclusion that deep learning obtains and internationally recognized prostate image report template and data system (prostate
Imaging reporting and data system, PI-RADS) compareed, to inspection requirements, assess criteria for classification, skill
Art specification and sweep parameter are verified and Knowledge Discovery, further obtains clinical decision axiom (Fig. 8).
Concrete example illustrates the practice process of the present invention below.
Step 1:Establish iconography knowledge base, establish the medical image feature noumenon of image technics, including anatomy portion
Position, image feature, nomenclature of diseases, classification of diseases, determines the subordinate relation between term;
Step 2:The specific performance of nomenclature of diseases definition being likely to occur to each region of anatomy;
Step 3:Establish report template:Including text report template and picture and text report template;Wherein, text report template
And lesion examination, preoperative evaluation, the accurate assessment report template such as follow-up assessment;Picture and text report template, such as lymph node subregion, disease
Stove measured zone, 3D display region, supply artery of the tumor show area.
Step 4:Case report template is collected, the knowledge element of formation is registered with CDA file modes, forms shadow
As knowledge element.
Step 5:The foundation of learning rule:Neural network model is established, a kind of proved by pathology is selected as needed or has
The disease of typical performance is learnt, and establishes the relation of diagnosis and Findings.
It the foregoing is only presently preferred embodiments of the present invention and oneself, be not limitation with the present invention, all essences in the present invention
Impartial modifications, equivalent substitutions and improvements made within refreshing and principle etc., it should be included in the patent covering scope of the present invention.
Claims (8)
1. a kind of medical image Knowledge Base based on structured report template, it is characterised in that comprise the following steps:
Step 1:The foundation of medical image feature noumenon:It is special that the classification of medical image feature noumenon includes anatomical structure, image
Sign, disease performance and diagnosis;
Step 2:Define the standard of disease:It is pre- to the position involved by report template using Consensus of experts, diagnosis guide as foundation
Disease that may be present and diagnosis are first defined, disease performance and diagnosis are retrieved by code;
Step 3:Design using structural data as the report template of element:Report template includes text report template and picture and text
Report template;
Step 4:The formation of knowledge element exchanges with information:Report template is stored according to CDA, will by human interpretation
Image is converted into computer-readable feature;
Step 5:Learning rule is established:Input value using features described above as deep learning, shadow is searched using deep learning method
As the relation of performance characteristic and diagnostic result.
2. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 1, the foundation of medical image feature noumenon includes following small step:
S1:List the entry involved by every one kind;
S2:Concluded and changed according to the build-in attribute of entry and exclusive feature, classification and hierarchical is established to entry
Disaggregated model;
S3:Unified coding is carried out to above-mentioned entry, adds related information;
S4:As needed, add example as concept tool as.
3. the medical image Knowledge Base according to claim 1 or 2 based on structured report template, its feature
It is, the classification of the medical image feature noumenon, i.e., with coding mark, hierarchical relationship, subordinate relation, the rule of searching route
Generalized Chinese terminology, accurately to be retrieved using computer, wherein, the succession that anatomical structure is defined according to analyzing system is closed
System;Interpretive classification is defined according to image feature;The disease performance according to corresponding to disease performance classification by histoorgan is defined;
The diagnosis of ACR code databases and ICD-10 code databases to follow-up case in database is established respectively according to diagnostic imaging conclusion to compile
Code.
4. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 2, defining the standard of disease, including following small step:
S5:In each region of anatomy, detailed description of the disease foundation including Normal appearances to predefined, related in description
And image feature use medical image feature noumenon in image feature term be marked;
S6:When image is understood and writes report template, after doctor selects nomenclature of diseases corresponding to the position image, it is corresponding in detail
Thin description is automatically filled in;
S7:Doctor is appropriately modified to description content on this basis;
S8:What report template outward appearance showed is the word description of image feature, and implicit content includes the happening part of disease, disease
Title and coding are demonstrate,proved, turns into searchable knowledge content;
S9:Above-mentioned report template content is cited in other document templates, reaches using limited report template to define
Some documents.
5. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 3, text report template includes scanning position, mode, the title of disease and code, the classification at scanning position
For neck, chest, abdomen and four limbs, mode is divided into CT, MRI and x-ray.
6. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 3, picture and text report template includes quantizating index, crucial image and annotation of images;Quantizating index includes dissection
Position, measured value and number of quantitative analysis, picture material is become to analyze the normal data of retrieval.
7. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 4, knowledge element is built using CDA standards;During the foundation of knowledge element, marked according to CDA and XML
Standard is constructed, and is a CDA documents per a iconography report template.
8. the medical image Knowledge Base according to claim 1 based on structured report template, its feature exist
In in the step 5, deep learning uses enforcement mechanisms learning method, establishes multilayer neural network model, multilayer nerve net
Network model includes input layer, hidden layer and output layer;
Input layer is:A=f (wTx+b)
Hidden layer is:Qj=f (∑ wij×xi-qj)
Output model is:Yk=f (Tjk×Oj-qk)
Wherein, f is non-linear action function, and q is the threshold value of neural unit, wijFor input node xiInfluence power to hidden node
Weight;
Error calculating is the function for reflecting neutral net desired output and calculating error size between output, and it is calculated such as
Under:
<mrow>
<mi>E</mi>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>2</mn>
<mi>N</mi>
</mrow>
</mfrac>
<mo>&times;</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>p</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>T</mi>
<mi>p</mi>
</msub>
<mo>-</mo>
<msub>
<mi>Q</mi>
<mi>p</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
Wherein TpExported for network objectives;
The learning process of multilayer neural network model, that is, connect the weight matrix w between lower level node and upper layer nodeijSetting
And error correction process.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710707594.3A CN107463786A (en) | 2017-08-17 | 2017-08-17 | Medical image Knowledge Base based on structured report template |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710707594.3A CN107463786A (en) | 2017-08-17 | 2017-08-17 | Medical image Knowledge Base based on structured report template |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107463786A true CN107463786A (en) | 2017-12-12 |
Family
ID=60549186
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710707594.3A Pending CN107463786A (en) | 2017-08-17 | 2017-08-17 | Medical image Knowledge Base based on structured report template |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107463786A (en) |
Cited By (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108305671A (en) * | 2018-01-23 | 2018-07-20 | 深圳科亚医疗科技有限公司 | By computer implemented medical image dispatching method, scheduling system and storage medium |
CN108376558A (en) * | 2018-01-24 | 2018-08-07 | 复旦大学 | A kind of multi-modal nuclear magnetic resonance image Case report no automatic generation method |
CN108389614A (en) * | 2018-03-02 | 2018-08-10 | 西安交通大学 | The method for building medical image collection of illustrative plates based on image segmentation and convolutional neural networks |
CN108717567A (en) * | 2018-05-03 | 2018-10-30 | 合肥工业大学 | Multi-modal affection data storage method and device |
CN108877369A (en) * | 2018-06-27 | 2018-11-23 | 上海交通大学医学院 | A kind of read tablet skills training is practiced system |
CN109189833A (en) * | 2018-08-28 | 2019-01-11 | 中国银联股份有限公司 | A kind of method for digging and device of knowledge base |
CN109300534A (en) * | 2018-11-30 | 2019-02-01 | 上海藤核智能科技有限公司 | A kind of construction method of medical image knowledge base and application |
CN109448849A (en) * | 2018-09-27 | 2019-03-08 | 广东省科技基础条件平台中心 | A kind of physical examination project indicator normal anomaly determination method based on frame representation |
CN109961834A (en) * | 2019-03-22 | 2019-07-02 | 上海联影医疗科技有限公司 | The generation method and equipment of diagnostic imaging report |
CN110069639A (en) * | 2019-04-01 | 2019-07-30 | 东华大学 | A method of building Thyroid ultrasound domain body |
CN110085299A (en) * | 2019-04-19 | 2019-08-02 | 合肥中科离子医学技术装备有限公司 | A kind of image recognition goes drying method, system and image library |
CN110415791A (en) * | 2019-01-29 | 2019-11-05 | 四川大学华西医院 | System and method is established in a kind of disease library |
CN110827988A (en) * | 2018-08-14 | 2020-02-21 | 上海明品医学数据科技有限公司 | Control method for medical data research based on mobile terminal |
CN110931096A (en) * | 2018-09-19 | 2020-03-27 | 北京赛迈特锐医疗科技有限公司 | System and method for automatically generating image structured report template |
CN111142758A (en) * | 2019-12-26 | 2020-05-12 | 嘉兴太美医疗科技有限公司 | Medical image reading interaction method, system and computer readable medium |
CN111324861A (en) * | 2020-02-28 | 2020-06-23 | 厦门大学 | Deep learning magnetic resonance spectrum reconstruction method based on matrix decomposition |
CN111341437A (en) * | 2020-02-21 | 2020-06-26 | 山东大学齐鲁医院 | Digestive tract disease judgment auxiliary system based on tongue image |
CN111584025A (en) * | 2020-05-06 | 2020-08-25 | 武汉联影医疗科技有限公司 | Image report generation method, device and storage medium |
CN111814478A (en) * | 2020-07-08 | 2020-10-23 | 泰康保险集团股份有限公司 | Imaging examination report structuring method and device |
CN111984805A (en) * | 2020-07-06 | 2020-11-24 | 北京欧应信息技术有限公司 | Method for establishing orthopedics cloud image database based on orthopedics knowledge graph |
CN112350999A (en) * | 2019-12-27 | 2021-02-09 | 北京理工大学 | Packaging transmission method for medical image analysis data set |
CN112614563A (en) * | 2020-12-28 | 2021-04-06 | 北京赛迈特锐医疗科技有限公司 | System and method for dynamically generating structured report in image diagnosis service |
CN112735569A (en) * | 2020-12-31 | 2021-04-30 | 四川大学华西医院 | System and method for outputting glioma operation area result before multi-modal MRI of brain tumor |
CN112863648A (en) * | 2020-12-31 | 2021-05-28 | 四川大学华西医院 | Brain tumor postoperative MRI (magnetic resonance imaging) multi-mode output system and method |
CN113053487A (en) * | 2021-04-22 | 2021-06-29 | 薛蕴菁 | Method and device for automatically giving diagnosis suggestions based on structured report historical data |
CN113077887A (en) * | 2021-03-24 | 2021-07-06 | 四川大学华西医院 | Automatic quantitative analysis system and interpretation method for white matter lesions of brain |
CN113096796A (en) * | 2021-04-01 | 2021-07-09 | 四川大学华西医院 | Intelligent prediction system and method for cerebral hemorrhage hematoma expansion risk |
CN114255835A (en) * | 2021-11-15 | 2022-03-29 | 四川大学华西医院 | Case big data management method and system based on artificial intelligence |
CN114298005A (en) * | 2021-12-29 | 2022-04-08 | 深圳市铱硙医疗科技有限公司 | Brain image auxiliary interpretation report generation system and method |
CN114446467A (en) * | 2021-12-31 | 2022-05-06 | 华中科技大学同济医学院附属协和医院 | Lung function quantitative analysis method based on structured report template |
CN116168796A (en) * | 2023-03-03 | 2023-05-26 | 杭州电子科技大学 | Medical image report structured generation method based on visual question and answer |
US11883687B2 (en) | 2020-09-08 | 2024-01-30 | Shanghai United Imaging Healthcare Co., Ltd. | X-ray imaging system for radiation therapy |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101315652A (en) * | 2008-07-17 | 2008-12-03 | 张小粤 | Composition and information query method of clinical medicine information system in hospital |
US8019621B2 (en) * | 2006-04-07 | 2011-09-13 | Siemens Medical Solutions Usa, Inc. | Medical image report data processing system |
CN102365641A (en) * | 2009-03-26 | 2012-02-29 | 皇家飞利浦电子股份有限公司 | A system that automatically retrieves report templates based on diagnostic information |
CN102651055A (en) * | 2012-04-11 | 2012-08-29 | 华中科技大学 | Method and system for generating file based on medical image |
CN103116643A (en) * | 2013-02-25 | 2013-05-22 | 江苏物联网研究发展中心 | Hadoop-based intelligent medical data management method |
CN104463754A (en) * | 2014-12-30 | 2015-03-25 | 天津迈沃医药技术有限公司 | Method for building medical ontology database based on disease characteristics |
-
2017
- 2017-08-17 CN CN201710707594.3A patent/CN107463786A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8019621B2 (en) * | 2006-04-07 | 2011-09-13 | Siemens Medical Solutions Usa, Inc. | Medical image report data processing system |
CN101315652A (en) * | 2008-07-17 | 2008-12-03 | 张小粤 | Composition and information query method of clinical medicine information system in hospital |
CN102365641A (en) * | 2009-03-26 | 2012-02-29 | 皇家飞利浦电子股份有限公司 | A system that automatically retrieves report templates based on diagnostic information |
CN102651055A (en) * | 2012-04-11 | 2012-08-29 | 华中科技大学 | Method and system for generating file based on medical image |
CN103116643A (en) * | 2013-02-25 | 2013-05-22 | 江苏物联网研究发展中心 | Hadoop-based intelligent medical data management method |
CN104463754A (en) * | 2014-12-30 | 2015-03-25 | 天津迈沃医药技术有限公司 | Method for building medical ontology database based on disease characteristics |
Non-Patent Citations (3)
Title |
---|
张娜等: "基于本体的医学影像信息整合", 《计算机系统应用》 * |
赵斌: "应用结构化报告构建医学影像学大数据", 《医学影像学杂志》 * |
陈晓娟: "《高校本科专业设置预测模型构建》", 30 April 2015 * |
Cited By (48)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108305671B (en) * | 2018-01-23 | 2021-01-01 | 深圳科亚医疗科技有限公司 | Computer-implemented medical image scheduling method, scheduling system, and storage medium |
CN108305671A (en) * | 2018-01-23 | 2018-07-20 | 深圳科亚医疗科技有限公司 | By computer implemented medical image dispatching method, scheduling system and storage medium |
CN108376558A (en) * | 2018-01-24 | 2018-08-07 | 复旦大学 | A kind of multi-modal nuclear magnetic resonance image Case report no automatic generation method |
CN108376558B (en) * | 2018-01-24 | 2021-08-20 | 复旦大学 | Automatic generation method for multi-modal nuclear magnetic resonance image medical record report |
CN108389614A (en) * | 2018-03-02 | 2018-08-10 | 西安交通大学 | The method for building medical image collection of illustrative plates based on image segmentation and convolutional neural networks |
CN108389614B (en) * | 2018-03-02 | 2021-01-19 | 西安交通大学 | Method for constructing medical image map based on image segmentation and convolutional neural network |
CN108717567A (en) * | 2018-05-03 | 2018-10-30 | 合肥工业大学 | Multi-modal affection data storage method and device |
CN108877369A (en) * | 2018-06-27 | 2018-11-23 | 上海交通大学医学院 | A kind of read tablet skills training is practiced system |
CN110827988A (en) * | 2018-08-14 | 2020-02-21 | 上海明品医学数据科技有限公司 | Control method for medical data research based on mobile terminal |
CN109189833A (en) * | 2018-08-28 | 2019-01-11 | 中国银联股份有限公司 | A kind of method for digging and device of knowledge base |
CN109189833B (en) * | 2018-08-28 | 2021-10-01 | 中国银联股份有限公司 | Knowledge base mining method and device |
CN110931096A (en) * | 2018-09-19 | 2020-03-27 | 北京赛迈特锐医疗科技有限公司 | System and method for automatically generating image structured report template |
CN109448849A (en) * | 2018-09-27 | 2019-03-08 | 广东省科技基础条件平台中心 | A kind of physical examination project indicator normal anomaly determination method based on frame representation |
CN109300534A (en) * | 2018-11-30 | 2019-02-01 | 上海藤核智能科技有限公司 | A kind of construction method of medical image knowledge base and application |
CN110415791A (en) * | 2019-01-29 | 2019-11-05 | 四川大学华西医院 | System and method is established in a kind of disease library |
US11574716B2 (en) * | 2019-03-22 | 2023-02-07 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for generating imaging report |
US20200303049A1 (en) * | 2019-03-22 | 2020-09-24 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for generating imaging report |
CN109961834A (en) * | 2019-03-22 | 2019-07-02 | 上海联影医疗科技有限公司 | The generation method and equipment of diagnostic imaging report |
CN110069639B (en) * | 2019-04-01 | 2023-07-07 | 东华大学 | Method for constructing thyroid ultrasound field ontology |
CN110069639A (en) * | 2019-04-01 | 2019-07-30 | 东华大学 | A method of building Thyroid ultrasound domain body |
CN110085299A (en) * | 2019-04-19 | 2019-08-02 | 合肥中科离子医学技术装备有限公司 | A kind of image recognition goes drying method, system and image library |
CN110085299B (en) * | 2019-04-19 | 2020-12-08 | 合肥中科离子医学技术装备有限公司 | Image identification dryness removal method and system and image library |
CN111142758A (en) * | 2019-12-26 | 2020-05-12 | 嘉兴太美医疗科技有限公司 | Medical image reading interaction method, system and computer readable medium |
CN111142758B (en) * | 2019-12-26 | 2021-03-16 | 上海亿锎智能科技有限公司 | Medical image reading interaction method, system and computer readable medium |
CN112350999A (en) * | 2019-12-27 | 2021-02-09 | 北京理工大学 | Packaging transmission method for medical image analysis data set |
CN112350999B (en) * | 2019-12-27 | 2021-08-24 | 北京理工大学 | Packaging transmission method for medical image analysis data set |
CN111341437A (en) * | 2020-02-21 | 2020-06-26 | 山东大学齐鲁医院 | Digestive tract disease judgment auxiliary system based on tongue image |
CN111341437B (en) * | 2020-02-21 | 2022-02-11 | 山东大学齐鲁医院 | Digestive tract disease judgment auxiliary system based on tongue image |
CN111324861A (en) * | 2020-02-28 | 2020-06-23 | 厦门大学 | Deep learning magnetic resonance spectrum reconstruction method based on matrix decomposition |
CN111584025A (en) * | 2020-05-06 | 2020-08-25 | 武汉联影医疗科技有限公司 | Image report generation method, device and storage medium |
CN111984805A (en) * | 2020-07-06 | 2020-11-24 | 北京欧应信息技术有限公司 | Method for establishing orthopedics cloud image database based on orthopedics knowledge graph |
CN111814478A (en) * | 2020-07-08 | 2020-10-23 | 泰康保险集团股份有限公司 | Imaging examination report structuring method and device |
CN111814478B (en) * | 2020-07-08 | 2023-11-24 | 泰康保险集团股份有限公司 | Imaging inspection report structuring method and device |
US11883687B2 (en) | 2020-09-08 | 2024-01-30 | Shanghai United Imaging Healthcare Co., Ltd. | X-ray imaging system for radiation therapy |
CN112614563A (en) * | 2020-12-28 | 2021-04-06 | 北京赛迈特锐医疗科技有限公司 | System and method for dynamically generating structured report in image diagnosis service |
CN112735569A (en) * | 2020-12-31 | 2021-04-30 | 四川大学华西医院 | System and method for outputting glioma operation area result before multi-modal MRI of brain tumor |
CN112863648A (en) * | 2020-12-31 | 2021-05-28 | 四川大学华西医院 | Brain tumor postoperative MRI (magnetic resonance imaging) multi-mode output system and method |
CN112863648B (en) * | 2020-12-31 | 2022-08-26 | 四川大学华西医院 | Brain tumor postoperative MRI (magnetic resonance imaging) multi-mode output system and method |
CN113077887A (en) * | 2021-03-24 | 2021-07-06 | 四川大学华西医院 | Automatic quantitative analysis system and interpretation method for white matter lesions of brain |
CN113096796A (en) * | 2021-04-01 | 2021-07-09 | 四川大学华西医院 | Intelligent prediction system and method for cerebral hemorrhage hematoma expansion risk |
CN113053487A (en) * | 2021-04-22 | 2021-06-29 | 薛蕴菁 | Method and device for automatically giving diagnosis suggestions based on structured report historical data |
CN114255835B (en) * | 2021-11-15 | 2023-07-28 | 四川大学华西医院 | Case big data management method and system based on artificial intelligence |
CN114255835A (en) * | 2021-11-15 | 2022-03-29 | 四川大学华西医院 | Case big data management method and system based on artificial intelligence |
CN114298005A (en) * | 2021-12-29 | 2022-04-08 | 深圳市铱硙医疗科技有限公司 | Brain image auxiliary interpretation report generation system and method |
CN114446467B (en) * | 2021-12-31 | 2023-09-12 | 华中科技大学同济医学院附属协和医院 | Pulmonary function quantitative analysis method based on structured report template |
CN114446467A (en) * | 2021-12-31 | 2022-05-06 | 华中科技大学同济医学院附属协和医院 | Lung function quantitative analysis method based on structured report template |
CN116168796A (en) * | 2023-03-03 | 2023-05-26 | 杭州电子科技大学 | Medical image report structured generation method based on visual question and answer |
CN116168796B (en) * | 2023-03-03 | 2023-11-10 | 杭州电子科技大学 | Medical image report structured generation method based on visual question and answer |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107463786A (en) | Medical image Knowledge Base based on structured report template | |
CN106202955B (en) | Diagnosis associated packets method and system based on intellectual coded adaptation | |
US10878962B2 (en) | System and method for extracting oncological information of prognostic significance from natural language | |
Liu et al. | Medical-vlbert: Medical visual language bert for covid-19 ct report generation with alternate learning | |
WO2016192612A1 (en) | Method for analysing medical treatment data based on deep learning, and intelligent analyser thereof | |
CN107563383A (en) | A kind of medical image auxiliary diagnosis and semi-supervised sample generation system | |
CN108628824A (en) | A kind of entity recognition method based on Chinese electronic health record | |
CN106776711A (en) | A kind of Chinese medical knowledge mapping construction method based on deep learning | |
Özkaraca et al. | Multiple brain tumor classification with dense CNN architecture using brain MRI images | |
CN109657062A (en) | A kind of electronic health record text resolution closed-loop policy based on big data technology | |
CN116860987A (en) | Domain knowledge graph construction method and system based on generation type large language model | |
RU2720363C2 (en) | Method for generating mathematical models of a patient using artificial intelligence techniques | |
CN115293128A (en) | Model training method and system based on multi-modal contrast learning radiology report generation | |
CN117271804B (en) | Method, device, equipment and medium for generating common disease feature knowledge base | |
Khan et al. | IoMT-enabled computer-aided diagnosis of pulmonary embolism from computed tomography scans using deep learning | |
CN109300534A (en) | A kind of construction method of medical image knowledge base and application | |
Seth et al. | Use of artificial intelligence in the advancement of breast surgery and implications for breast reconstruction: a narrative review | |
Xiao et al. | Collaborations of industry, academia, research and application improve the healthy development of medical imaging artificial intelligence industry in China | |
JPWO2009025022A1 (en) | Information processing system and program | |
CN110060749A (en) | Electronic health record intelligent diagnosing method based on SEV-SDG-CNN | |
Xu et al. | Gene mutation classification using CNN and BiGRU network | |
Deng et al. | Automatic cardiopulmonary endurance assessment: a machine learning approach based on GA-XGBOOST | |
JP5682657B2 (en) | Database system | |
CN113887204A (en) | Coding method for clinical examination medical text | |
Erattakulangara et al. | Automatic Multiple Articulator Segmentation in Dynamic Speech MRI Using a Protocol Adaptive Stacked Transfer Learning U-NET Model |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171212 |
|
RJ01 | Rejection of invention patent application after publication |