CN110232383A - A kind of lesion image recognition methods and lesion image identifying system based on deep learning model - Google Patents
A kind of lesion image recognition methods and lesion image identifying system based on deep learning model Download PDFInfo
- Publication number
- CN110232383A CN110232383A CN201910527716.XA CN201910527716A CN110232383A CN 110232383 A CN110232383 A CN 110232383A CN 201910527716 A CN201910527716 A CN 201910527716A CN 110232383 A CN110232383 A CN 110232383A
- Authority
- CN
- China
- Prior art keywords
- image
- lesion
- deep learning
- learning model
- organ
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to technical field of medical equipment, disclose a kind of lesion image recognition methods and lesion image identifying system based on deep learning model.It creates through the invention, provide new method and new system a kind of based on deep learning model and that the state of an illness in doctor's automatic identification medical image can be substituted, i.e. for the medical image to be detected obtained, by successively carrying out the segmentation of minimum organ-tissue image and the image identification of corresponding organ-tissue, it is marked on deep learning model prediction and prediction result figure, the potential state of an illness in doctor's discovery medical image can be substituted, and the identification label of lesion tissue is automatically carried out on medical image, to remind doctor is further to be diagnosed, the state of an illness is made a definite diagnosis in time, and then it can reduce the working strength of doctor, make a definite diagnosis whether have lesion in time, avoid the therapic opportunity of the delay state of an illness, it is particularly useful to the discovery of early lesion.
Description
Technical field
The invention belongs to technical field of medical equipment, know more particularly to a kind of lesion image based on deep learning model
Other method and lesion image identifying system.
Background technique
The medical photographic equipment on Medical Device Market is all that can see medical imaging, such as vasography figure at present
As (Angiography), angiocardiography image (Cardiac angiography), computed tomography image (CT,
Computerized tomography), mammogram image (Mammography), positron emission tomoscan image (PET,
Positron emission tomography), Magnetic resonance imaging image (NMRI, Nuclear magnetic
Resonance imaging) and medicine ultrasound examination image (Medical ultrasonography) etc..But for disease
The discovery or identification of stove are still that doctor with the naked eye goes to judge whether there is generation lesion problem by professional knowledge, so in lesion
Initial stage, since lesion is often more small, it is easy to it is ignored, while in the observation of the organ-tissues such as bronchus, by
In needing to watch, position is relatively more, and doctor needs each position to examine, and greatly strengthens the working strength of doctor.Cause
How this, mitigate the working strength of doctor and quickly early find the state of an illness, be a urgent problem.
Summary of the invention
In order to solve the problems, such as that existing Medical Devices cannot mitigate the working strength of doctor and the quick early discovery state of an illness,
It is an object of that present invention to provide a kind of lesion image recognition methods and lesion image identifying system based on deep learning model.
The technical scheme adopted by the invention is as follows:
A kind of lesion image recognition methods based on deep learning model, includes the following steps:
S101. medical image to be detected is obtained；
S102. image dividing processing is carried out to the medical image to be detected, it is mutually disjoint to be detected obtains several
Area image；
S103. it is directed to each area to be tested image, image outline feature is first extracted, if then according to the image wheel of extraction
Wide feature identifies corresponding organ-tissue, then successively executes step S301~S302:
S301. deep learning being applicable in and that lesion image recognition training is completed is determined according to organ-tissue recognition result
Then area to be tested image imported into the deep learning model and carries out prediction operation by model, obtain corresponding lesion figure
As identification types and lesion image recognition accuracy, wherein the lesion image identification types include no lesion type and ill
Stove type；
If the lesion image identification types S302. obtained are to have lesion type and lesion image recognition accuracy not less than the
One threshold value, then on the medical image to be detected and lesion image that position mark corresponding with area to be tested image obtains
Identification types and lesion image recognition accuracy；
S104. the medical image of lesion marking is completed in output.
Specifically, the medical image to be detected is vasography image, Angiogram in the step S101
Picture, computed tomography image, mammogram image, positron emission tomoscan image, Magnetic resonance imaging image and medicine
Ultrasound examination image.
Specifically, the mode for carrying out image dividing processing is the image based on threshold value point in the step S102
Cut mode, the image segmentation mode based on region-growing method, the image segmentation mode based on distorted pattern, the image based on graph theory
Partitioning scheme, the image segmentation mode based on cluster or the image segmentation mode based on classification.
Optimization, further include having the following steps before the step S103:
S300. the standard medical image of a variety of different organ and tissues is obtained, then for the standard doctor of every kind of organ-tissue
Image is treated, extracts corresponding standard picture contour feature respectively.
Specifically, extracting the image outline feature in image using the ASM algorithm based on points distribution models.
Specifically, in the step S103, identified pair come the image outline feature according to extraction as follows
The organ-tissue answered: by the image outline feature of area to be tested image seriatim with the standard picture profile of various organ-tissues
Feature carries out similarity calculation, is more than the organ-tissue of second threshold as the correspondence identified using similarity highest and similarity
Organ-tissue.
Optimization, before the step S301, S201~S203 is to suitable for some organ-tissue in accordance with the following steps
Deep learning model carry out lesion image recognition training:
S201. the sample medical image of the organ-tissue and the disease of mark with the binding of each sample medical image are obtained
Stove image recognition type, wherein be directed to various lesion image identification types, the number of corresponding sample medical image is no less than 1000
?；
S202. using each sample medical image and the corresponding lesion image identification types that marked as primary training sample
This, imported into progress lesion image recognition training in deep learning model, wherein inputs number for sample medical image as sample
According to, corresponding with sample medical image will mark lesion image identification types as sample verify data；
S203. during lesion image recognition training, according to the resulting lesion image identification types of training and sample school
The matching result for testing data continues to optimize deep learning model, until completing training or until the resulting lesion image of training
The matching rate of identification types and sample verification data reaches third threshold value.
Optimization, in the step S301 and by area to be tested image imported into determining deep learning model it
Before: processing is zoomed in and out to the area to be tested image, make that treated picture size length or width and sample medical image
Unanimously.
Specifically, the deep learning model uses the convolutional neural networks based on CNN framework in the step S301
Model.
Another technical solution of the present invention are as follows:
A kind of lesion image identifying system based on deep learning model, including image collection module, image segmentation module,
Organ identification module, deep learning module, lesion marking module and image output module；
Described image obtains module, for obtaining medical image to be detected；
Described image divides module, and communication connection described image obtains module, for the medical image to be detected to acquisition
Image dividing processing is carried out, several mutually disjoint area to be tested images are obtained；
The organ identification module, communication connection described image divides module, for extracting the figure of area to be tested image
As contour feature, corresponding organ-tissue is then identified according to the image outline feature of extraction；
The deep learning module communicates to connect the organ identification module, for true according to organ-tissue recognition result
Surely deep learning model applicable and that lesion image recognition training is completed, then imported into the depth for area to be tested image
Prediction operation is carried out in degree learning model, obtains corresponding lesion image identification types and lesion image recognition accuracy, wherein
The lesion image identification types include no lesion type and have lesion type；
The lesion marking module communicates to connect described image respectively and obtains module and the deep learning module, is used for
When the obtained lesion image identification types of discovery are to have lesion type and when lesion image recognition accuracy is greater than first threshold,
On the medical image to be detected and lesion image identification types that position mark corresponding with area to be tested image obtains and
Lesion image recognition accuracy；
Described image output module communicates to connect the lesion marking module, for exporting the medical treatment for completing lesion marking
Image.
The invention has the benefit that
(1) the invention provides a kind of based on deep learning model and can substitute doctor's automatic identification medical treatment figure
The new method and new system of the state of an illness as in, i.e., for the medical image to be detected obtained, by successively carrying out minimum organ-tissue
It is marked on the segmentation of image and the image identification of corresponding organ-tissue, deep learning model prediction and prediction result figure, it can be with
The potential state of an illness in doctor's discovery medical image is substituted, and automatically carries out the identification label of lesion tissue on medical image,
To remind doctor is further to be diagnosed, the state of an illness is made a definite diagnosis in time, and then can reduce the working strength of doctor, making a definite diagnosis in time is
It is no to have lesion, the therapic opportunity of the delay state of an illness is avoided, the discovery of early lesion is particularly useful to.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the flow diagram of lesion image recognition methods provided by the invention.
Fig. 2 is the structural schematic diagram of lesion image identifying system provided by the invention.
Specific embodiment
With reference to the accompanying drawing and specific embodiment come the present invention is further elaborated.It should be noted that for
Although the explanation of these way of example is to be used to help understand the present invention, but and do not constitute a limitation of the invention.It is public herein
The specific structure and function detail opened are only used for description example embodiments of the present invention.However, can with many alternative forms come
The present invention is embodied, and is not construed as limiting the invention in embodiment set forth herein.
It will be appreciated that though various units may be described herein using term first, second etc., but these units
It should not be limited by these terms.These terms are only used to distinguish a unit and another unit.Such as it can be by
Unit one is referred to as second unit, and similarly second unit can be referred to as first unit, shows without departing from of the invention
The range of example embodiment.
It should be appreciated that being only a kind of pass for describing affiliated partner to the term "and/or" being likely to occur in this article
Connection relationship indicates may exist three kinds of relationships, for example, A and/or B, can indicate: individualism A, individualism B are deposited simultaneously
In tri- kinds of situations of A and B；To the term "/and " being likely to occur in this article, it is to describe another affiliated partner relationship, indicates
There may be two kinds of relationships, for example, A/ and B, can indicate: two kinds of situations of individualism A, individualism A and B；In addition, for
The character "/" being likely to occur herein, typicallying represent forward-backward correlation object is a kind of "or" relationship.
If should be appreciated that, it can when unit being referred to as with another unit " connection ", " connected " or " coupling " herein
To be directly connected with another unit or couple or temporary location may exist.Relatively, if herein by unit be referred to as with
When another unit " being connected directly " or " direct-coupling ", indicate that temporary location is not present.Additionally, it should solve in a similar manner
Release for describing the relationship between unit other words (for example, " ... between " to " between directly existing ... ", " adjacent "
To " direct neighbor " etc.).
It should be appreciated that terms used herein are only used for description specific embodiment, it is not intended to limit example of the invention
Embodiment.If used herein, singular "a", "an" and "the" is intended to include plural form, unless context
Contrary is explicitly indicated.If being also understood that, term " includes ", " including ", "comprising" and/or " containing " are herein
When being used, specify stated feature, integer, step, operation, unit and/or component existence, and be not excluded for one
Or other multiple features, quantity, step, operation, unit, component and/or their combination existence or increase.
It should be appreciated that it will be further noted that the function action occurred may go out with attached drawing in some alternative embodiments
Existing sequence is different.Such as related function action is depended on, it can actually substantially be executed concurrently, or sometimes
Two figures continuously shown can be executed in reverse order.
It should be appreciated that providing specific details, in the following description in order to which example embodiment is understood completely.
However those of ordinary skill in the art are it is to be understood that implementation example embodiment without these specific details.
Such as system can be shown in block diagrams, to avoid with unnecessary details come so that example is unclear.In other instances, may be used
Or not show well-known process, structure and technology unnecessary details, to avoid making example embodiment unclear.
Embodiment one
As shown in Figure 1, the lesion image recognition methods based on deep learning model provided in this embodiment, including such as
Lower step S101~S104.
S101. medical image to be detected is obtained.
In the step S101, the medical image to be detected specifically can be from the output interface of existing medical photographic equipment
Export obtains in (such as USB interface or HDMI interface).Specifically, the medical image to be detected can be, but not limited to as blood vessel
Photographs, angiocardiography image, computed tomography image, mammogram image, positron emission tomoscan image, core
Magnetic resonance imaging image and medicine ultrasound examination image etc..
S102. image dividing processing is carried out to the medical image to be detected, it is mutually disjoint to be detected obtains several
Area image.
In the step S102, due to that generally can include the image of multiple organ-tissues in medical image to be detected,
Therefore by image dividing processing, the image of single organ-tissue can be limited to (part in an area to be tested image
In area to be tested image may not yet organ-tissue image), so as to it is subsequent whether be lesion group to single organ-tissue
It knits and is effectively identified.Specifically, the mode for carrying out image dividing processing can be, but not limited to as the image based on threshold value
Partitioning scheme, the image segmentation mode based on region-growing method, the image segmentation mode based on distorted pattern, the figure based on graph theory
Image segmentation mode as partitioning scheme, based on cluster or the image segmentation mode based on classification etc., aforementioned image dividing processing
Mode is the prior art, is not being repeated in this.
S103. it is directed to each area to be tested image, image outline feature is first extracted, if then according to the image wheel of extraction
Wide feature identifies corresponding organ-tissue, then successively executes step S301~S302.
In the step S103, it can be, but not limited to extract area to be detected using the ASM algorithm based on points distribution models
Image outline feature in area image.ASM (the Active Shape Model) algorithm is a kind of existing based on point distribution
The algorithm of model (Point Distribution Model, PDM), in PDM, object similar for shape, such as heart,
The geometry of lung etc. can be sequentially connected in series by the coordinate by several key feature points forms a shape vector to indicate,
So after handling using ASM algorithm the area to be tested image, corresponding image outline feature can be extracted.
Before the step S103, corresponding organ can be identified according to the image outline feature of extraction to realize
The purpose of tissue needs to prepare the standard picture contour feature of a variety of different organ and tissues, to be compared identification, i.e., also wraps
Include and have the following steps: S300. obtains the standard medical image of a variety of different organ and tissues, is then directed to the mark of every kind of organ-tissue
Quasi- medical image extracts corresponding standard picture contour feature respectively.Wherein, it still can be, but not limited to use based on point minute
Image outline feature in the ASM algorithm extraction standard medical image of cloth model.
Further specifically, corresponding organ group can be identified come the image outline feature according to extraction as follows
It knits: the image outline feature of area to be tested image is seriatim subjected to phase with the standard picture contour feature of various organ-tissues
It is calculated like degree, is more than the organ-tissue of second threshold as the correspondence organ-tissue identified using similarity highest and similarity.
Wherein, used similarity calculation algorithm is existing algorithm, and the second threshold both can be preset threshold value, can also
To be default value, such as 50%.Due to there is the case where cannot recognize that corresponding organ-tissue, i.e., all similarities are below institute
State in second threshold or the area to be tested image the just not no image of organ-tissue, so can skip step S301~
S302。
S301. deep learning being applicable in and that lesion image recognition training is completed is determined according to organ-tissue recognition result
Then area to be tested image imported into the deep learning model and carries out prediction operation by model, obtain corresponding lesion figure
As identification types and lesion image recognition accuracy, wherein the lesion image identification types include no lesion type and ill
Stove type.
In the step S301 and before being predicted, need to be selected according to organ-tissue recognition result suitably and
The deep learning model of lesion image recognition training is completed, i.e., for different organ and tissues such as heart and lungs, since lesion is existing
As difference, their image recognition feature is also inevitable different, it is therefore desirable to select different and corresponding lesion image knowledge is completed
Not Xun Lian deep learning model predicted.The deep learning model is that a kind of application joins similar to cerebral nerve cynapse
The structure that connects carries out the mathematics computing model of information process-, by a large amount of node (or neuron) and between be coupled to each other
It constitutes, and a kind of specific output function of each node on behalf, referred to as excitation function (activation function)；Every two
Connection between a node all represents a weighted value for passing through the connection signal, referred to as weight, this is equivalent to artificial mind
Memory through network；The output of network then rely on the connection type of neural network, the difference of weighted value and excitation function and it is different.
Thus the deep learning model can pass through a learning method based on mathematical statistics type (Learning Method)
The features such as being optimised, having the function of the function that self-learning function, connection entropy and high speed find optimization solution and superiority.Specifically
, the deep learning model can be, but not limited to using the convolutional neural networks model based on CNN framework, wherein CNN
(Convolutional Neural Networks, convolutional neural networks) are the general names of a neural network, can be passed through
Caffe (a kind of specific implementation of CNN framework) Lai Shixian, make prediction model have upper quick-moving, speed it is fast, can modularization, opening
Property good and community cultule it is good the features such as.After carrying out prediction operation by deep learning model, meeting exports automatically and area to be tested
The corresponding lesion image identification types of image and lesion image recognition accuracy.In addition, described have lesion type can also basis
State of an illness type and severity are further segmented, such as lung, can also be subdivided into pulmonary tuberculosis early stage, pulmonary tuberculosis
Mid-term, pulmonary tuberculosis advanced stage, lung neoplasm early stage, lung neoplasm mid-term and lung neoplasm advanced stage etc..
Before the step S301, it is necessary to carry out corresponding lesion to the deep learning model of selected organ-tissue in advance
Image recognition training S201~S203 can carry out the deep learning model for being suitable for some organ-tissue in accordance with the following steps
Lesion image recognition training: S201. obtains the sample medical image of the organ-tissue and binds with each sample medical image
The lesion image identification types of mark, wherein be directed to various lesion image identification types, the number of corresponding sample medical image
No less than 1000；S202. using each sample medical image and the corresponding lesion image identification types that marked as primary instruction
Practice sample, imported into progress lesion image recognition training in deep learning model, wherein sample medical image is defeated as sample
Enter data, corresponding with sample medical image will mark lesion image identification types as sample and verify data；S203. in disease
In stove image recognition training process, the matching result of data is verified according to the resulting lesion image identification types of training and sample,
Deep learning model is continued to optimize, until completing training or until the resulting lesion image identification types of training and sample verify
The matching rate of data reaches third threshold value.In the step S203, the third threshold value both can be preset threshold value,
It is also possible to default value, such as 99%.A kind of Caffe framework (i.e. tool of CNN framework is based on when the deep learning model uses
Body way of realization) convolutional neural networks model when, can use by accuracy layers of recognition accuracy got as training
The matching rate of resulting lesion image identification types and sample verification data, recognition accuracy is higher, i.e., matching rate is higher, matching
Property is better.In addition, in order to further enhance forecasting accuracy, optimization, in the step S301 and by area to be tested figure
As before importeding into determining deep learning model: zooming in and out processing to the area to be tested image, the image that makes that treated
Dimensions length or width are consistent with sample medical image.
If the lesion image identification types S302. obtained are to have lesion type and lesion image recognition accuracy not less than the
One threshold value, then on the medical image to be detected and lesion image that position mark corresponding with area to be tested image obtains
Identification types and lesion image recognition accuracy.
In the step S302, the first threshold both can be preset threshold value, be also possible to default value, example
Such as 75%.From there through aforementioned lesion marking method, the medical image for marking a variety of lesion situations can be automatically derived, to mention
The doctor that wakes up further is diagnosed, and makes a definite diagnosis the state of an illness in time.
S104. the medical image of lesion marking is completed in output.
To sum up, using based on the lesion image recognition methods of deep learning model, having as follows provided by the present embodiment
Technical effect:
(1) it present embodiments provides a kind of based on deep learning model and doctor's automatic identification medical image can be substituted
The new method of the middle state of an illness, i.e., for the medical image to be detected obtained, by the segmentation for successively carrying out minimum organ-tissue image
And the image of corresponding organ-tissue is identified, is marked on deep learning model prediction and prediction result figure, can substitute doctor's hair
The potential state of an illness in existing medical image, and the identification label of lesion tissue is automatically carried out on medical image, to remind doctor
Life is further to be diagnosed, and makes a definite diagnosis the state of an illness in time, and then can reduce the working strength of doctor, whether make a definite diagnosis in time has lesion hair
It is raw, the therapic opportunity of the delay state of an illness is avoided, the discovery of early lesion is particularly useful to.
Embodiment two
As shown in Fig. 2, the present embodiment is provided a kind of based on identical inventive concept and is based on deep relative to embodiment one
Spend the lesion image identifying system of learning model, including image collection module, image segmentation module, organ identification module, depth
Study module, lesion marking module and image output module；Described image obtains module, for obtaining medical image to be detected；
Described image divides module, and communication connection described image obtains module, for carrying out image to the medical image to be detected of acquisition
Dividing processing obtains several mutually disjoint area to be tested images；The organ identification module communicates to connect described image
Divide module, for extracting the image outline feature of area to be tested image, is then identified according to the image outline feature of extraction
Corresponding organ-tissue out；The deep learning module communicates to connect the organ identification module, for knowing according to organ-tissue
Other result determines deep learning model being applicable in and that lesion image recognition training is completed, then leads area to be tested image
Enter and carry out prediction operation into the deep learning model, obtains corresponding lesion image identification types and lesion image identification is accurate
Rate, wherein the lesion image identification types include no lesion type and have lesion type；The lesion marking module, respectively
It communicates to connect described image and obtains module and the deep learning module, the lesion image identification types for obtaining when discovery are
When thering is lesion type and lesion image recognition accuracy to be greater than first threshold, on the medical image to be detected and with it is to be detected
The lesion image identification types and lesion image recognition accuracy that the corresponding position mark of area image obtains；Described image output
Module communicates to connect the lesion marking module, for exporting the medical image for completing lesion marking.
The particular technique details and total technical effect of each functional module, can refer to embodiment always in the present embodiment
It connects and is derived by, do not repeated in this.
Multiple embodiments described above are only schematical, if being related to unit as illustrated by the separation member,
It may or may not be physically separated；If being related to component shown as a unit, can be or
It can not be physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to reality
Some or all of the units may be selected to achieve the purpose of the solution of this embodiment for the needs on border.Those of ordinary skill in the art
Without paying creative labor, it can understand and implement.
The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to the foregoing embodiments
Invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each implementation
Technical solution documented by example is modified or equivalent replacement of some of the technical features.And these modification or
Replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.
Finally it should be noted that the present invention is not limited to above-mentioned optional embodiment, anyone is in enlightenment of the invention
Under can all obtain other various forms of products.Above-mentioned specific embodiment should not be understood the limit of pairs of protection scope of the present invention
System, protection scope of the present invention should be subject to be defined in claims, and specification can be used for explaining that right is wanted
Seek book.
Claims (10)
1. a kind of lesion image recognition methods based on deep learning model, which comprises the steps of:
S101. medical image to be detected is obtained；
S102. image dividing processing is carried out to the medical image to be detected, obtains several mutually disjoint area to be tested
Image；
S103. it is directed to each area to be tested image, first extracts image outline feature, if then special according to the image outline of extraction
Sign identifies corresponding organ-tissue, then successively executes step S301~S302:
S301. deep learning mould being applicable in and that lesion image recognition training is completed is determined according to organ-tissue recognition result
Then area to be tested image imported into the deep learning model and carries out prediction operation, obtains corresponding lesion image by type
Identification types and lesion image recognition accuracy, wherein the lesion image identification types include no lesion type and have lesion
Type；
If the lesion image identification types S302. obtained are to have lesion type and lesion image recognition accuracy not less than the first threshold
Value, then on the medical image to be detected and the identification of lesion image that position mark corresponding with area to be tested image obtains
Type and lesion image recognition accuracy；
S104. the medical image of lesion marking is completed in output.
2. a kind of lesion image recognition methods based on deep learning model as described in claim 1, it is characterised in that: in institute
It states in step S101, the medical image to be detected is vasography image, angiocardiography image, CT Scan figure
Picture, mammogram image, positron emission tomoscan image, Magnetic resonance imaging image and medicine ultrasound examination image.
3. a kind of lesion image recognition methods based on deep learning model as described in claim 1, it is characterised in that: in institute
It states in step S102, the mode for carrying out image dividing processing is image segmentation mode based on threshold value, is based on region growing
The image segmentation mode of method, the image segmentation mode based on distorted pattern, the image segmentation mode based on graph theory, based on cluster
Image segmentation mode or image segmentation mode based on classification.
4. a kind of lesion image recognition methods based on deep learning model as described in claim 1, which is characterized in that in institute
Further include having the following steps before stating step S103:
S300. the standard medical image of a variety of different organ and tissues is obtained, the standard medical figure of every kind of organ-tissue is then directed to
Picture extracts corresponding standard picture contour feature respectively.
5. a kind of lesion image recognition methods based on deep learning model as described in claim 1 or 4, it is characterised in that:
Image outline feature in image is extracted using the ASM algorithm based on points distribution models.
6. a kind of lesion image recognition methods based on deep learning model as claimed in claim 4, which is characterized in that in institute
It states in step S103, corresponding organ-tissue is identified according to the image outline feature of extraction as follows: will be to be checked
The image outline feature for surveying area image seriatim carries out similarity calculation with the standard picture contour feature of various organ-tissues,
It is more than the organ-tissue of second threshold as the correspondence organ-tissue identified using similarity highest and similarity.
7. a kind of lesion image recognition methods based on deep learning model as described in claim 1, which is characterized in that in institute
Before stating step S301, S201~S203 carries out disease to the deep learning model for being suitable for some organ-tissue in accordance with the following steps
The training of stove image recognition:
S201. the sample medical image of the organ-tissue and the lesion figure of mark with the binding of each sample medical image are obtained
As identification types, wherein be directed to various lesion image identification types, the number of corresponding sample medical image is no less than 1000；
S202. it using each sample medical image and the corresponding lesion image identification types that marked as a training sample, leads
Enter into deep learning model and carry out lesion image recognition training, wherein, will using sample medical image as sample input data
The lesion image identification types that marked corresponding with sample medical image are as sample verification data；
S203. during lesion image recognition training, according to the resulting lesion image identification types of training and sample check number
According to matching result, deep learning model is continued to optimize, until completing training or until training resulting lesion image identification
The matching rate of type and sample verification data reaches third threshold value.
8. a kind of lesion image recognition methods based on deep learning model as described in claim 1, which is characterized in that in institute
It states in step S301 and area to be tested image is imported into before determining deep learning model: to the area to be tested image
Processing is zoomed in and out, making that treated, picture size length or width is consistent with sample medical image.
9. a kind of lesion image recognition methods based on deep learning model as described in claim 1, it is characterised in that: in institute
It states in step S301, the deep learning model uses the convolutional neural networks model based on CNN framework.
10. a kind of lesion image identifying system based on deep learning model, it is characterised in that: including image collection module, figure
As segmentation module, organ identification module, deep learning module, lesion marking module and image output module；
Described image obtains module, for obtaining medical image to be detected；
Described image divides module, and communication connection described image obtains module, for the medical image to be detected progress to acquisition
Image dividing processing obtains several mutually disjoint area to be tested images；
The organ identification module, communication connection described image divides module, for extracting the image wheel of area to be tested image
Then wide feature identifies corresponding organ-tissue according to the image outline feature of extraction；
The deep learning module communicates to connect the organ identification module, suitable for being determined according to organ-tissue recognition result
And the deep learning model of lesion image recognition training is completed, area to be tested image is then imported into the depth
It practises in model and carries out prediction operation, obtain corresponding lesion image identification types and lesion image recognition accuracy, wherein described
Lesion image identification types include no lesion type and have lesion type；
The lesion marking module communicates to connect described image and obtains module and the deep learning module, respectively for when hair
The lesion image identification types now obtained are when having lesion type and lesion image recognition accuracy to be greater than first threshold, described
On medical image to be detected and lesion image identification types and lesion that position mark corresponding with area to be tested image obtains
Image recognition accuracy rate；
Described image output module communicates to connect the lesion marking module, for exporting the medical image for completing lesion marking.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910527716.XA CN110232383B (en) | 2019-06-18 | 2019-06-18 | Focus image recognition method and focus image recognition system based on deep learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910527716.XA CN110232383B (en) | 2019-06-18 | 2019-06-18 | Focus image recognition method and focus image recognition system based on deep learning model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110232383A true CN110232383A (en) | 2019-09-13 |
| CN110232383B CN110232383B (en) | 2021-07-02 |
Family
ID=67859748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910527716.XA Active CN110232383B (en) | 2019-06-18 | 2019-06-18 | Focus image recognition method and focus image recognition system based on deep learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110232383B (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766701A (en) * | 2019-10-31 | 2020-02-07 | 北京推想科技有限公司 | Network model training method and device, and region division method and device |
| CN110837572A (en) * | 2019-11-15 | 2020-02-25 | 北京推想科技有限公司 | Image retrieval method and device, readable storage medium and electronic equipment |
| CN110867233A (en) * | 2019-11-19 | 2020-03-06 | 西安邮电大学 | System and method for generating electronic laryngoscope medical test reports |
| CN110859642A (en) * | 2019-11-26 | 2020-03-06 | 北京华医共享医疗科技有限公司 | Method, device, equipment and storage medium for realizing medical image auxiliary diagnosis based on AlexNet network model |
| CN110897865A (en) * | 2019-12-25 | 2020-03-24 | 中科彭州智慧产业创新中心有限公司 | Auricular point guiding device and method |
| CN111192249A (en) * | 2019-12-30 | 2020-05-22 | 辽宁师范大学 | Medical image lesion region segmentation method based on energy functional model of machine learning |
| CN111382785A (en) * | 2020-03-04 | 2020-07-07 | 武汉精立电子技术有限公司 | GAN network model and method for realizing automatic cleaning and auxiliary marking of sample |
| CN111383328A (en) * | 2020-02-27 | 2020-07-07 | 西安交通大学 | 3D visualization method and system for breast cancer focus |
| CN111383211A (en) * | 2020-03-04 | 2020-07-07 | 深圳大学 | Bone case identification method, device, server and storage medium |
| CN111753831A (en) * | 2020-06-28 | 2020-10-09 | 上海联影医疗科技有限公司 | Image analysis method and device, image acquisition equipment and storage medium |
| CN111899848A (en) * | 2020-08-05 | 2020-11-06 | 中国联合网络通信集团有限公司 | Image recognition method and device |
| CN112168193A (en) * | 2020-10-14 | 2021-01-05 | 北京赛迈特锐医疗科技有限公司 | System and method for acquiring patella anatomical parameters based on patella axial position X-ray image |
| CN112263269A (en) * | 2020-09-22 | 2021-01-26 | 北京赛迈特锐医疗科技有限公司 | Intelligent detection system and method for urinary system X-ray plain stone |
| CN112614568A (en) * | 2020-12-28 | 2021-04-06 | 东软集团股份有限公司 | Inspection image processing method and device, storage medium and electronic equipment |
| CN112991166A (en) * | 2019-12-16 | 2021-06-18 | 无锡祥生医疗科技股份有限公司 | Intelligent auxiliary guiding method, ultrasonic equipment and storage medium |
| CN113456093A (en) * | 2021-06-09 | 2021-10-01 | 北京东软医疗设备有限公司 | Image processing method, device and image processing system |
| CN114429649A (en) * | 2022-04-07 | 2022-05-03 | 青岛美迪康数字工程有限公司 | Target image identification method and device |
| CN114782397A (en) * | 2022-05-12 | 2022-07-22 | 广东德澳智慧医疗科技有限公司 | Artificial intelligence tumour diagnostic system based on medical image and machine learning |
| TWI792055B (en) * | 2020-09-25 | 2023-02-11 | 國立勤益科技大學 | Establishing method of echocardiography judging model with 3d deep learning, echocardiography judging system with 3d deep learning and method thereof |
| CN115953555A (en) * | 2022-12-29 | 2023-04-11 | 南京鼓楼医院 | Adenomyosis modeling method based on ultrasonic measured value |
| CN116664580A (en) * | 2023-08-02 | 2023-08-29 | 经智信息科技(山东)有限公司 | Multi-image hierarchical joint imaging method and device for CT images |
| CN116681717A (en) * | 2023-08-04 | 2023-09-01 | 经智信息科技(山东)有限公司 | CT image segmentation processing method and device |
| CN117059235A (en) * | 2023-08-17 | 2023-11-14 | 经智信息科技(山东)有限公司 | Automatic rendering method and device for CT image |
| CN117575999A (en) * | 2023-11-01 | 2024-02-20 | 广州盛安医学检验有限公司 | Focus prediction system based on fluorescent marking technology |
| CN112614568B (en) * | 2020-12-28 | 2024-05-28 | 东软集团股份有限公司 | Method and device for processing inspection image, storage medium and electronic equipment |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070081710A1 (en) * | 2005-10-07 | 2007-04-12 | Siemens Corporate Research, Inc. | Systems and Methods For Segmenting Object Of Interest From Medical Image |
| US20130223704A1 (en) * | 2012-02-28 | 2013-08-29 | Siemens Aktiengesellschaft | Method and System for Joint Multi-Organ Segmentation in Medical Image Data Using Local and Global Context |
| CN106204587A (en) * | 2016-05-27 | 2016-12-07 | 孔德兴 | Multiple organ dividing method based on degree of depth convolutional neural networks and region-competitive model |
| WO2017048744A1 (en) * | 2015-09-18 | 2017-03-23 | University Of Pittsburgh-Of The Commonwealth System Of Higher Education | Automated segmentation of organs, such as kidneys, from magnetic resonance images |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN106940816A (en) * | 2017-03-22 | 2017-07-11 | 杭州健培科技有限公司 | Connect the CT image Lung neoplasm detecting systems of convolutional neural networks entirely based on 3D |
| CN107274402A (en) * | 2017-06-27 | 2017-10-20 | 北京深睿博联科技有限责任公司 | A kind of Lung neoplasm automatic testing method and system based on chest CT image |
| CN107563123A (en) * | 2017-09-27 | 2018-01-09 | 百度在线网络技术（北京）有限公司 | Method and apparatus for marking medical image |
| CN108364006A (en) * | 2018-01-17 | 2018-08-03 | 超凡影像科技股份有限公司 | Medical Images Classification device and its construction method based on multi-mode deep learning |
| CN108510489A (en) * | 2018-03-30 | 2018-09-07 | 四川元匠科技有限公司 | A kind of pneumoconiosis detection method and system based on deep learning |
| CN108665456A (en) * | 2018-05-15 | 2018-10-16 | 广州尚医网信息技术有限公司 | The method and system that breast ultrasound focal area based on artificial intelligence marks in real time |
-
2019
- 2019-06-18 CN CN201910527716.XA patent/CN110232383B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070081710A1 (en) * | 2005-10-07 | 2007-04-12 | Siemens Corporate Research, Inc. | Systems and Methods For Segmenting Object Of Interest From Medical Image |
| US20130223704A1 (en) * | 2012-02-28 | 2013-08-29 | Siemens Aktiengesellschaft | Method and System for Joint Multi-Organ Segmentation in Medical Image Data Using Local and Global Context |
| WO2017048744A1 (en) * | 2015-09-18 | 2017-03-23 | University Of Pittsburgh-Of The Commonwealth System Of Higher Education | Automated segmentation of organs, such as kidneys, from magnetic resonance images |
| CN106204587A (en) * | 2016-05-27 | 2016-12-07 | 孔德兴 | Multiple organ dividing method based on degree of depth convolutional neural networks and region-competitive model |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN106940816A (en) * | 2017-03-22 | 2017-07-11 | 杭州健培科技有限公司 | Connect the CT image Lung neoplasm detecting systems of convolutional neural networks entirely based on 3D |
| CN107274402A (en) * | 2017-06-27 | 2017-10-20 | 北京深睿博联科技有限责任公司 | A kind of Lung neoplasm automatic testing method and system based on chest CT image |
| CN107563123A (en) * | 2017-09-27 | 2018-01-09 | 百度在线网络技术（北京）有限公司 | Method and apparatus for marking medical image |
| CN108364006A (en) * | 2018-01-17 | 2018-08-03 | 超凡影像科技股份有限公司 | Medical Images Classification device and its construction method based on multi-mode deep learning |
| CN108510489A (en) * | 2018-03-30 | 2018-09-07 | 四川元匠科技有限公司 | A kind of pneumoconiosis detection method and system based on deep learning |
| CN108665456A (en) * | 2018-05-15 | 2018-10-16 | 广州尚医网信息技术有限公司 | The method and system that breast ultrasound focal area based on artificial intelligence marks in real time |
Cited By (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766701A (en) * | 2019-10-31 | 2020-02-07 | 北京推想科技有限公司 | Network model training method and device, and region division method and device |
| CN110766701B (en) * | 2019-10-31 | 2020-11-06 | 北京推想科技有限公司 | Network model training method and device, and region division method and device |
| CN110837572A (en) * | 2019-11-15 | 2020-02-25 | 北京推想科技有限公司 | Image retrieval method and device, readable storage medium and electronic equipment |
| CN110867233A (en) * | 2019-11-19 | 2020-03-06 | 西安邮电大学 | System and method for generating electronic laryngoscope medical test reports |
| CN110859642A (en) * | 2019-11-26 | 2020-03-06 | 北京华医共享医疗科技有限公司 | Method, device, equipment and storage medium for realizing medical image auxiliary diagnosis based on AlexNet network model |
| CN110859642B (en) * | 2019-11-26 | 2024-01-23 | 北京华医共享医疗科技有限公司 | Method, device, equipment and storage medium for realizing medical image auxiliary diagnosis based on AlexNet network model |
| CN112991166A (en) * | 2019-12-16 | 2021-06-18 | 无锡祥生医疗科技股份有限公司 | Intelligent auxiliary guiding method, ultrasonic equipment and storage medium |
| CN110897865A (en) * | 2019-12-25 | 2020-03-24 | 中科彭州智慧产业创新中心有限公司 | Auricular point guiding device and method |
| CN111192249A (en) * | 2019-12-30 | 2020-05-22 | 辽宁师范大学 | Medical image lesion region segmentation method based on energy functional model of machine learning |
| CN111192249B (en) * | 2019-12-30 | 2023-06-20 | 辽宁师范大学 | Medical image lesion region segmentation method based on machine learning energy functional model |
| CN111383328A (en) * | 2020-02-27 | 2020-07-07 | 西安交通大学 | 3D visualization method and system for breast cancer focus |
| CN111382785A (en) * | 2020-03-04 | 2020-07-07 | 武汉精立电子技术有限公司 | GAN network model and method for realizing automatic cleaning and auxiliary marking of sample |
| CN111383211A (en) * | 2020-03-04 | 2020-07-07 | 深圳大学 | Bone case identification method, device, server and storage medium |
| CN111382785B (en) * | 2020-03-04 | 2023-09-01 | 武汉精立电子技术有限公司 | GAN network model and method for realizing automatic cleaning and auxiliary marking of samples |
| CN111753831A (en) * | 2020-06-28 | 2020-10-09 | 上海联影医疗科技有限公司 | Image analysis method and device, image acquisition equipment and storage medium |
| CN111899848A (en) * | 2020-08-05 | 2020-11-06 | 中国联合网络通信集团有限公司 | Image recognition method and device |
| CN111899848B (en) * | 2020-08-05 | 2023-07-07 | 中国联合网络通信集团有限公司 | Image recognition method and device |
| CN112263269B (en) * | 2020-09-22 | 2024-04-19 | 北京赛迈特锐医疗科技有限公司 | Intelligent detection system and method for urinary X-ray flat-piece calculus |
| CN112263269A (en) * | 2020-09-22 | 2021-01-26 | 北京赛迈特锐医疗科技有限公司 | Intelligent detection system and method for urinary system X-ray plain stone |
| TWI792055B (en) * | 2020-09-25 | 2023-02-11 | 國立勤益科技大學 | Establishing method of echocardiography judging model with 3d deep learning, echocardiography judging system with 3d deep learning and method thereof |
| CN112168193B (en) * | 2020-10-14 | 2024-04-23 | 北京赛迈特锐医疗科技有限公司 | System and method for acquiring patella anatomical parameters based on patella axial X-ray image |
| CN112168193A (en) * | 2020-10-14 | 2021-01-05 | 北京赛迈特锐医疗科技有限公司 | System and method for acquiring patella anatomical parameters based on patella axial position X-ray image |
| CN112614568A (en) * | 2020-12-28 | 2021-04-06 | 东软集团股份有限公司 | Inspection image processing method and device, storage medium and electronic equipment |
| CN112614568B (en) * | 2020-12-28 | 2024-05-28 | 东软集团股份有限公司 | Method and device for processing inspection image, storage medium and electronic equipment |
| CN113456093A (en) * | 2021-06-09 | 2021-10-01 | 北京东软医疗设备有限公司 | Image processing method, device and image processing system |
| CN114429649A (en) * | 2022-04-07 | 2022-05-03 | 青岛美迪康数字工程有限公司 | Target image identification method and device |
| CN114782397A (en) * | 2022-05-12 | 2022-07-22 | 广东德澳智慧医疗科技有限公司 | Artificial intelligence tumour diagnostic system based on medical image and machine learning |
| CN114782397B (en) * | 2022-05-12 | 2022-12-23 | 中晗控股集团有限公司 | Artificial intelligence tumor diagnosis system based on medical image and machine learning |
| CN115953555A (en) * | 2022-12-29 | 2023-04-11 | 南京鼓楼医院 | Adenomyosis modeling method based on ultrasonic measured value |
| CN115953555B (en) * | 2022-12-29 | 2023-08-22 | 南京鼓楼医院 | Uterine adenomyosis modeling method based on ultrasonic measurement value |
| CN116664580A (en) * | 2023-08-02 | 2023-08-29 | 经智信息科技(山东)有限公司 | Multi-image hierarchical joint imaging method and device for CT images |
| CN116664580B (en) * | 2023-08-02 | 2023-11-28 | 经智信息科技(山东)有限公司 | Multi-image hierarchical joint imaging method and device for CT images |
| CN116681717B (en) * | 2023-08-04 | 2023-11-28 | 经智信息科技(山东)有限公司 | CT image segmentation processing method and device |
| CN116681717A (en) * | 2023-08-04 | 2023-09-01 | 经智信息科技(山东)有限公司 | CT image segmentation processing method and device |
| CN117059235A (en) * | 2023-08-17 | 2023-11-14 | 经智信息科技(山东)有限公司 | Automatic rendering method and device for CT image |
| CN117575999A (en) * | 2023-11-01 | 2024-02-20 | 广州盛安医学检验有限公司 | Focus prediction system based on fluorescent marking technology |
| CN117575999B (en) * | 2023-11-01 | 2024-04-16 | 广州盛安医学检验有限公司 | Focus prediction system based on fluorescent marking technology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110232383B (en) | 2021-07-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110232383A (en) | A kind of lesion image recognition methods and lesion image identifying system based on deep learning model | |
| CN109886273B (en) | CMR image segmentation and classification system | |
| CN106056595B (en) | Based on the pernicious assistant diagnosis system of depth convolutional neural networks automatic identification Benign Thyroid Nodules | |
| van Ginneken | Fifty years of computer analysis in chest imaging: rule-based, machine learning, deep learning | |
| EP3770850A1 (en) | Medical image identifying method, model training method, and computer device | |
| CN107644420B (en) | Blood vessel image segmentation method based on centerline extraction and nuclear magnetic resonance imaging system | |
| CN108257135A (en) | The assistant diagnosis system of medical image features is understood based on deep learning method | |
| Kim et al. | Machine-learning-based automatic identification of fetal abdominal circumference from ultrasound images | |
| CN112101451B (en) | Breast cancer tissue pathological type classification method based on generation of antagonism network screening image block | |
| CN110506278A (en) | Target detection in latent space | |
| CN109635846A (en) | A kind of multiclass medical image judgment method and system | |
| CN110555836A (en) | Automatic identification method and system for standard fetal section in ultrasonic image | |
| CN107292884A (en) | The method and device of oedema and hemotoncus in a kind of identification MRI image | |
| Simaiya et al. | MRI brain tumour detection & image segmentation by hybrid hierarchical K-means clustering with FCM based machine learning model | |
| CN108665454A (en) | A kind of endoscopic image intelligent classification and irregular lesion region detection method | |
| CN109902682A (en) | A kind of mammary gland x line image detection method based on residual error convolutional neural networks | |
| CN109460717A (en) | Alimentary canal Laser scanning confocal microscope lesion image-recognizing method and device | |
| Cai et al. | Identifying architectural distortion in mammogram images via a se-densenet model and twice transfer learning | |
| Manikandan et al. | Segmentation and Detection of Pneumothorax using Deep Learning | |
| Nayan et al. | A deep learning approach for brain tumor detection using magnetic resonance imaging | |
| Liu et al. | Automated classification and measurement of fetal ultrasound images with attention feature pyramid network | |
| Cui et al. | DsUnet: a new network structure for detection and segmentation of ultrasound breast lesions | |
| CN111461065A (en) | Tubular structure identification method and device, computer equipment and readable storage medium | |
| CN114360695B (en) | Auxiliary system, medium and equipment for breast ultrasonic scanning and analyzing | |
| CN115908299A (en) | Medical image-based life cycle prediction method, device, equipment and medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CB03 | Change of inventor or designer information |
Inventor after: Do not announce the inventor Inventor before: Do not announce the inventor |
|
| CB03 | Change of inventor or designer information |