CN107274386A - A kind of cervical cell liquid-based smear artificial intelligence aids in diagosis system - Google Patents
A kind of cervical cell liquid-based smear artificial intelligence aids in diagosis system Download PDFInfo
- Publication number
- CN107274386A CN107274386A CN201710351064.XA CN201710351064A CN107274386A CN 107274386 A CN107274386 A CN 107274386A CN 201710351064 A CN201710351064 A CN 201710351064A CN 107274386 A CN107274386 A CN 107274386A
- Authority
- CN
- China
- Prior art keywords
- cell
- image
- cervical
- segmentation
- convolutional neural
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Diagosis system is aided in the invention discloses a kind of cervical cell liquid-based smear artificial intelligence, the system includes：Cell image acquisition module, cell image pretreatment module, cell image detection segmentation module carries out automatic detection, while being split automatically to nucleus, cytoplasm, background in same cell component for the different cell components to image cell；Segmentation result is finely adjusted and optimized；Cell rapid classification identification module, for the image after segmentation to be identified, divides into individual cells or cell cluster；Hierarchical identification is carried out to individual cells using the double-current convolutional neural networks in additional knowledge field and the cell knowledge mapping of structure respectively, two kinds of classification results are respectively obtained；Using the identification of the inseparable cell cluster of double-current convolutional neural networks model realization of cell cluster；With interpretation and post-processing module, joint interpretation is carried out for two kinds of classification results to individual cells, clash handle is carried out, obtains the classification results of individual cells.
Description
Technical field
The present invention relates to medical cell image processing field, and in particular to a kind of cervical cell liquid-based smear artificial intelligence is auxiliary
Help diagosis system.
Background technology
Cervical carcinoma is the malignant tumour for betiding uterine cervix, and it has close relationship with high-risk HPV infection.Palace
Early detection, diagnosis and the treatment of neck precancerous lesion are one of reduction cervical cancer pathogenesis rate, the strategy of the death rate.Cervical carcinoma is sieved
Checking method is easy to be economical, and many countries reduce its Disease Spectrum using examination means control cervical cancer pathogenesis rate, the death rate.
2008 to 2016, Beijing completed the free people of examination more than 200 ten thousand of cervical carcinoma, detection cervical carcinoma and precancerosis altogether
More than 4000 examples of change, cervical cytology positive rate 2.3%, according to the data of foreign data combination China some areas,
The positive rate of cervical cytology is substantially between 5-7% in experts guesstimate China normal population, and Pekinese's detection at present
Rate is with this standard deviation away from larger.
Cervical cell liquid-based smear artificial intelligence auxiliary diagosis is in itself for mitigating the labor intensity of diagosis worker, being lifted
Diagosis accuracy rate and operating efficiency, generally investigated applied to large batch of cervical cancer cell in terms of will produce positive effect, its
The rule of development and needs of medical market are followed, meets the research layout of the medical big data of national health, image information and shadow
As intelligent diagnostics have critical role in medical big data industry.
Some famous research institutions of US and European have begun to carry out cervical cell quantitative analysis and auxiliary diagnosis
The slide scan-image analysis system of correlative study, such as U.S. Hao Luojie companies, it can interpretation cervical cell liquid-based smear, one
Aspect system can not provide final grading diagnosis result, and its function is to select 22 visuals field to supply doctor's interpretation, doctor from smear
Teacher can not carry out interpretation to other visuals field, to improve the efficiency of diagosis；Another aspect system can only judge that the said firm oneself gives birth to
The liquid-based smear of production, can not be judged for the liquid-based smear of other companies, thus cause the versatility of diagosis system.
The blue fourth medical science in domestic Wuhan is absorbed in cervical carcinoma screening, and the automatic cervix uteri cancer for being developed based on the DNA quantitative analyses of image is thin
Born of the same parents' detecting system, but it lacks the feature extraction of each idioblas in terms of data analysis, also fails to provide solving for judged result
The property released.
Existing cervical cell quantitative analysis and aided diagnosis method still suffer from open defect, cervical cell analytical technology
Research is still in the starting stage, is the clinic and research needs in further in-depth cervical cytology field, need to make full use of uterine neck
Cellular prion protein information, space neighborhood information and distribution of color information set up accurate Methods of Segmentation On Cell Images, make full use of palace
Neck cell database and cervical cell clinical diagnosis rule set up cervical cell quantitative analysis and intelligent auxiliary diagnosis framework, are base
Software platform is provided in cervical cell liquid-based smear artificial intelligence auxiliary diagosis system.
The content of the invention
It is an object of the present invention to overcome the above mentioned problem that current cervical cell quantitative analysis and auxiliary diagnosis are present；Base
In existing cervical cytology domain knowledge, learnt by depth learning technology and extract the key feature of cervical cell, divide automatically
The cancerous region and type of cell on liquid-based smear are cut and recognized, is finally reached and is shortened scoring time, reduces wrong diagnosis and escape rate
Purpose, provides solution for cervical cell liquid-based smear artificial intelligence auxiliary diagosis technology, finally realizes that cervical cell is quantified
Change and evaluate and aid in the methodology of intelligent diagnostics to break through.
To achieve these goals, the present invention proposes a kind of cervical cell liquid-based smear artificial intelligence auxiliary diagosis system
System, the system includes：
Cell image acquisition module, for being scanned preservation cell image using smear automatic scanner eclipsed form；
Cell image pretreatment module, for being pre-processed to cell image；
Cell image detection segmentation module, carries out automatic detection, while right for the different cell components to image cell
Nucleus, cytoplasm, background are split automatically in same cell component；Using improved movable contour model and fast area
Convolutional neural networks are detected, cell characteristic is aided with again and multiple dimensioned horizontal scheduling algorithm is finely adjusted and optimized to segmentation result；
Cell rapid classification identification module, for the image after segmentation to be identified, divides into individual cells or cell
Cluster；Individual cells are distinguished using the double-current convolutional neural networks in additional knowledge field and the cell knowledge mapping of structure
Hierarchical identification is carried out, the first classification results and second of classification results are respectively obtained；Using the double-current convolution god of cell cluster
The identification of inseparable cell cluster is realized through network model；With
Interpretation and post-processing module, join for the first classification results and second of classification results to individual cells
Interpretation is closed, clash handle is carried out, obtains the classification results of individual cells；The clash handle is used to solve various features sensing not
With sentence read result when, comprehensive various factors eliminates conflict, makes clearly reliable interpretation；Then knowledge mapping and class are utilized
Active map realizes the readability of cervical cell identification process, the interpretation of cervical cell recognition result.
In above-mentioned technical proposal, the implementation process of the cell image acquisition module is：Using 40 times of amplifications of eyepiece, sweep
Path is retouched for rectangle, scan mode scans for eclipsed form so that scanning range is all covered with liquid-based smear cells location.
In above-mentioned technical proposal, the pretreatment includes：Denoising is carried out to image using two-sided filter, the wave filter is
It is made up of two functions：One function is to determine filter coefficient by geometric space distance, and another determines filter by pixel value difference
Ripple device coefficient；Then the edge of image is repaired using Morphological scale-space, filling cavity simultaneously removes thin connection, finally makes
Increase nucleus and cytoplasmic contrast with histogram equalization.
In above-mentioned technical proposal, the step that implements of the cell image detection segmentation module is：
Step S1) preceding background coarse segmentation is carried out to pretreated cell image, extract the region belonging to cell；
Step S2) to after coarse segmentation cell image carry out cell component detection split, use fast area convolution god
Different types of cell is partitioned into through network；
Step S3) detect and split cervical cell nucleus；
Step S4) nucleus is screened according to the characteristic parameter of nucleus, obtain final candidate cell core；
Step S5) judgment step S2) in obtained cell type whether be cell cluster, if it is not, then using activity
Skeleton pattern and Prior Template carry out the segmentation of cytosolic domain；Otherwise, it is transferred to step S6)；
Step S6) comprehensive nucleus post-processed with cytoplasmic segmentation result and domain knowledge, the whole uterine neck of completion
Effective segmentation of cell.
In above-mentioned technical proposal, network of the fast area convolutional neural networks from the VGG16 of convolutional neural networks
Structure, the size of input picture is 515*512, and the detection classification of final cell composition falls into 5 types：Scale cell, gland cell, neck tube
Cell, metaplasia cell and background quality, the cell component in addition to scale cell and background quality are all defined as inseparable cell
Cluster.
In above-mentioned technical proposal, the step S5) in use movable contour model and Prior Template carry out cytoplasm
The detailed process of the segmentation in region is：
Improved movable contour model is employed, energy function and shape priors is added, is made iteratively profile
Optimization, obtain cytoplasmic exact boundary；
Energy function E (u) is：
E (u)=λ1Es(u)+R(u)
Wherein Es(u) it is shape prior, R (u) is regular terms, the flatness for ensureing partitioning boundary, λ1It can learn
Parameter；Shape prior Es(u) it is：
Wherein, H is Hessian matrix.
In above-mentioned technical proposal, the step that implements of the cell image detection segmentation module is：
Step 1) pretreatment operation is carried out to the cervical cell image after segmentation；
Step 2) whether judge pretreated cell image be individual cells, if it is, being transferred to step 3), otherwise, should
Image is inseparable cell cluster, is transferred to step 6)；
Step 3) computable cell parameters is determined, then calculate cell parameters feature；
Step 4) cell knowledge mapping reasoning and judging model is set up, and cell parameters feature is inputted into the model, obtain list
The first classification results of individual cell；
Step 5) build additional fields knowledge double-current convolutional neural networks model, by cell parameters feature and cell image
The double-current convolutional neural networks model of input, obtains second of classification results of individual cells；
Step 6) the double-current convolutional neural networks model of cell cluster is built, and the model is used to inseparable cell cluster
The hierarchical identification of cell cluster is carried out, the classification results of cell cluster are obtained.
In above-mentioned technical proposal, the input all the way of the double-current convolutional neural networks of the additional fields knowledge is step 3)
The cell parameters feature arrived, input is individual cells image all the way in addition, and size is uniformly normalized to 256*256 pixel values, is passed through
The convolution pond composite module for crossing 5 cascades implicitly extracts the feature of cell image；Most important of which convolution operation
Convolution kernel size uses 7*7 sizes, and step-length selection size is 1, and characteristic pattern number is chosen for 96, and the convolution operation is：
In above formula, M represents the set of the input feature vector figure of selection, wijRepresent weight, bjAdd for the output of each characteristic pattern
On an additional bias, then 1096 dimensional features of extraction are spliced to plus computable 20 dimensional feature of cell field knowledge
Together, the full articulamentum and classification layer of double-current convolutional neural networks are input to.
In above-mentioned technical proposal, the input all the way of the double-current convolutional neural networks of the cell cluster is：Between nucleus
Queueing discipline feature, in addition all the way input be cell cluster corresponding with the cell parameters cervical cell, cervical cell
Input size is uniformly normalized to 512*512 pixel values, and the convolution pond composite module by 8 cascades is implicitly extracted carefully
The feature of born of the same parents' image；The convolution kernel size of most important of which convolution operation uses 5*5 sizes, and step-length selection size is 2, feature
Figure number is chosen for 108.
The beneficial effects of the invention are as follows：
1st, system of the invention has hypersensitivity for lesion cervical cell system, has height for normal cervix cell
Specificity, whole auxiliary diagosis system mitigates the labor intensity of diagosis worker significantly without manually participating in.
Brief description of the drawings
Fig. 1 aids in the composition figure of diagosis system for the cervical cell liquid-based smear artificial intelligence of the present invention；
Fig. 2 splits the schematic diagram of module for the cervical cell image detection of the present invention；
Fig. 3 is the schematic diagram of the cervical cell rapid classification identification module of the present invention.
Embodiment
Presently preferred embodiments of the present invention is described in detail below in conjunction with the accompanying drawings, so that advantages and features of the invention energy
It is easier to be readily appreciated by one skilled in the art, apparent is clearly defined so as to be made to protection scope of the present invention.
As shown in figure 1, a kind of cervical cell liquid-based smear artificial intelligence auxiliary diagosis system, the system includes：
Cell image acquisition module, for being scanned preservation cell image using smear automatic scanner eclipsed form；
Wherein, using 40 times of amplifications of eyepiece, scanning pattern is rectangle, and scan mode scans for eclipsed form so that scanning range and liquid
Base smear cells location can all be covered；
For example, the cervical cell image after 20 scanning can be obtained for the image of a pixel of 20,000 * 20,000.
Cell image pretreatment module, for being pre-processed to cell image, including：Using two-sided filter to image
Denoising is carried out, the wave filter is made up of two functions：One function is to determine filter coefficient by geometric space distance, another
It is individual that filter coefficient is determined by pixel value difference；Then the edge of image is repaired using Morphological scale-space, filling cavity is simultaneously
Thin connection is removed, finally increases nucleus and cytoplasmic contrast using histogram equalization.
Cell image detection segmentation module, carries out automatic detection, while right for the different cell components to image cell
Nucleus, cytoplasm, background are split automatically in same cell component；Using improved movable contour model and fast area
Convolutional neural networks are detected, cell characteristic is aided with again and multiple dimensioned horizontal scheduling algorithm is finely adjusted and optimized to segmentation result；
As shown in Fig. 2 implementing step and being：
Step S1) preceding background coarse segmentation is carried out to pretreated cell image, extract the region belonging to cell；
The removal of the noise of whole cell image is carried out using 9 threshold values；Then SIFT rim detections are used and multiple dimensioned
Watershed algorithm obtains foreground area, and SIFT maintains the invariance to rotation, scaling, brightness change, to visual angle change, affine
Conversion, noise also keep a certain degree of stability；Fine setting finally is optimized for the region being partitioned into, it is poly- using K averages
Class algorithm is by neighbouring potting gum, and the cluster number selection of clustering algorithm of the present invention is 3, that is, is divided into cytoplasm, nucleus and the back of the body
Three classifications of scape.
Step S2) to after coarse segmentation cell image carry out cell component detection split, use fast area convolution god
Different classes of cell and background quality are partitioned into through network；
The cleaning for carrying out data first against training sample is arranged, then using the fast area convolution in deep learning algorithm
Neutral net carries out the detection segmentation of cell component.Its training stage selection training method end to end, model selection convolution
The VGG16 of neutral net network structure, the size of input picture is 515*512, and the detection classification of final cell composition is divided into 5
Class：By the extracellular four kinds of classifications of de-scaling in scale cell, gland cell, neck tube cell, metaplasia cell and background quality, the present invention
Cell component be all defined as inseparable cell cluster.The follow-up hierarchical identification step of diagosis is aided in view of artificial intelligence,
Inseparable cell Cluster zone is divided into gland cell, neck tube cell, metaplasia cell and the class of background quality four in detail herein, its
Middle background quality is not belonging to cell category, is not involved in follow-up hierarchical identification.
Step S3) detect and split cervical cell nucleus；
The algorithm of improved random forest is employed, point of nuclear area is carried out by extracting 5 kinds of features of nucleus
Cut, the number selection 20 set in forest, Characteristic Number selection during selection optimal characteristics is logN, the minimum of leaf node every time
Number selection is 3；
It is aided with the detection that multi-scale watershed algorithm carries out nucleus again to prevent from missing nuclear area, by choosing
5 kinds of different parameters are merged, and image to be split is divided into high and the low different scale of merging degree the cell of merging degree
Image, then both results detected are combined, it is used as the candidate region of nucleus.
Step S4) nucleus is screened according to the characteristic parameter of nucleus, obtain final candidate cell core；
According to the characteristic parameter of nucleus, comprising nucleus size, the circularity of nucleus, the depth of nucleus can be counted
Calculate feature.The parameter of the size of such as nucleus is by directly calculating the summation of the pixel in nuclear area border come table
Show, specific practice such as formula 1.
Wherein f (x, y) is the pixel value of certain point (x, y) on bianry image, and value represents that the pixel belongs to target when being 1
Region, value represents that the pixel belongs to background area when being 0, and its area is exactly to count the number of pixels that f (x, y) is 1.
Step S5) judgment step S2) in obtained cell type whether be cell cluster, if it is not, then using activity
Skeleton pattern and Prior Template carry out the segmentation of cytosolic domain；Otherwise, it is transferred to step S6)；
Improved movable contour model is employed, energy function and shape priors, the carry out profile of iteration are added
Optimization, obtain cytoplasmic exact boundary；
Energy function E (u) is：
E (u)=λ1Es(u)+R(u)
Wherein Es(u) be that the present invention proposes the shape prior that uses, R (u) is regular terms, it is ensured that partitioning boundary it is smooth
Property, λ1For the parameter that can learn；Shape prior Es(u) it is：
Wherein H is Hessian matrix (Hessian Matrix).
Step S6) comprehensive nucleus post-processed with cytoplasmic segmentation result and domain knowledge, the whole uterine neck of completion
Effective segmentation of cell.
Post-processed for the segmentation result of above-mentioned steps, the main morphological operation pair using in Medical Image Processing
The edge of image is repaired, and filling cavity simultaneously removes thin connection, main using the ginseng for opening operation and closed operation, wherein template
Number is set to the methods such as the size of [3 3], and filtering and noise reduction and enters the smooth of row bound, and the wherein setting w of parameter is 2, variance
Sigma uses [2 0.1], so as to obtain final accurate boundary information.
Cell rapid classification identification module, for the image after segmentation to be identified, divides into individual cells or cell
Cluster；Individual cells are classified using the double-current convolutional neural networks in additional knowledge field and the cell knowledge mapping of structure
Identification；Using the identification of the inseparable cell cluster of double-current convolutional neural networks model realization of cell cluster；
As shown in figure 3, implementing step and being：
Step 1) pretreatment operation is carried out to the cervical cell image after segmentation；
Cervical cell region after scanning segmentation, and cell boundaries pixel value filling is carried out, by the pixel outside cell boundaries
Value is filled with 0, and then the cell image after filler pixels value is uniformly normalized to 256*256 pixel value size.
Step 2) whether judge pretreated cell image be individual cells, if it is, being transferred to step 3), otherwise, should
Image is inseparable cell cluster, is transferred to step 7)；
Realized using the watershed algorithm in image procossing, if the cell nuclei in cell image is 1, judgement is single
Individual cell.
Step 3) computable cell parameters is determined, include size, depth, the shape of nucleus, cytoplasmic size, shape
Shape and karyoplasmic ratio, then calculate cell parameters feature；
By taking the size of nucleus as an example, the size parameter of nucleus is by directly calculating the picture in nuclear area border
The summation of element is represented：
Wherein f (x, y) is the pixel value of certain point (x, y) on bianry image, and value represents that the pixel belongs to target when being 1
Region, value represents that the pixel belongs to background area when being 0, and its area is exactly to count the number of pixels that f (x, y) is 1.
Step 4) cell knowledge mapping reasoning and judging model is set up, and cell parameters feature is inputted into the model, obtain thin
The first classification results of born of the same parents；
Step 5) build additional fields knowledge double-current convolutional neural networks model, by cell parameters feature and cell image
The double-current convolutional neural networks model of input, obtains second of classification results of cell；
As shown in Fig. 2 the input all the way of double-current convolutional neural networks is step 3) obtained cell parameters feature, in addition one
Road input is individual cells image, and size is uniformly normalized to 256*256 pixel values, and the convolution pondization by 5 cascades is combined
Module implicitly extracts the feature of cell image.The convolution kernel size of most important of which convolution operation uses 7*7 sizes, step
Long selection size is 1, and characteristic pattern number is chosen for 96, and the convolution operation is：
M represents the set of the input feature vector figure of selection, wijRepresent weight, bjOne added is exported for each characteristic pattern
1096 dimensional features of extraction, are then spliced together by additional bias plus computable 20 dimensional feature of cell field knowledge, defeated
Enter the full articulamentum and classification layer to double-current convolutional neural networks.According to TBS standard diagnostics, by the hierarchical identification knot of different cells
It is combined, one is divided into 9 classes.
As shown in figure 3, cytologic characteristic is completely according to TBS criterions, extraction is the language using criterion.Cell
A variety of classification can be carried out on different abstraction hierarchies, from the level with the presence or absence of lesion, cell is broadly divided into normal cell
With the major class of abnormal cell two, normal cell, including columnar cell, middle layer cells, cells of superficial layer；Abnormal cell includes slight squamous
Intraepithelial lesions cell, moderate SIL cell, severe SIL cell, epidermoid carcinoma cell.
The interpretation process and thought of diagnostic rule storehouse also completely according to diagosis doctor, such as cytoplasmic color characteristic reflect
Penetrate, for the color that cervical lesionses have interpretation meaning have blueness, pink, crocus, cell technics be commonly referred to as basophilla,
Acidophilia, the cell matter of thermophilic crocus, i.e. basophilla show as blueness, acidophic cell cytoplasm show as pink,
Thermophilic crocus cell matter shows as crocus.
Step 6) the double-current convolutional neural networks model of cell cluster is built, and the model is used to inseparable cell cluster
Carry out the hierarchical identification of cluster cell；
The input all the way of the double-current convolutional neural networks of cell cluster is：The feature of queueing discipline between nucleus, separately
Input is the cervical cell of cell cluster corresponding with the cell parameters all the way outside, and the present invention is unified by cervical cell input size
512*512 pixel values are normalized to, the convolution pond composite module by 8 cascades implicitly extracts the spy of cell image
Levy.The convolution kernel size of most important of which convolution operation uses 5*5 sizes, and step-length selection size is 2, and characteristic pattern number is chosen
For 108.
Interpretation and post-processing module, join for the first classification results and second of classification results to individual cells
Interpretation is closed, clash handle is carried out, obtains the classification results of individual cells；Utilize knowledge mapping and CAM (Class
Activation Mapping, class active map) method realize the readability of cervical cell identification process, cervical cell identification
As a result interpretation.
The first classification results and second of classification results of such as some cell are all the result of squamous cell carcinoma, then
Then the interpretation cell is squamous cell carcinoma.
When clash handle mainly solves various features sensing different sentence read results, comprehensive various factors eliminates conflict, done
Go out clearly reliable interpretation.The result that such as some cell is obtained by the first classification (double-current convolutional neural networks model) is squama
Shape cell cancer, and the result obtained by the second classification (knowledge mapping model) is low level SIL, such a feelings
Condition belongs to outcome conflict, and the processing present invention for conflict is to mark out the cell come final result can by the way of
Determined by diagosis doctor.
Above-described embodiment, has been carried out further to the purpose of the present invention, technical scheme and beneficial effect
Describe in detail, should be understood that the embodiment that the foregoing is only the present invention, be not intended to limit the present invention
Protection domain, within the spirit and principles of the invention, any modification, equivalent substitution and improvements done etc. all should be included
Within protection scope of the present invention.
Claims (9)
1. a kind of cervical cell liquid-based smear artificial intelligence aids in diagosis system, the system includes：
Cell image acquisition module, for being scanned preservation cell image using smear automatic scanner eclipsed form；
Cell image pretreatment module, for being pre-processed to cell image；
Cell image detection segmentation module, carries out automatic detection, while to same for the different cell components to image cell
Nucleus, cytoplasm, background are split automatically in cell component；Using improved movable contour model and fast area convolution
Neutral net is detected, cell characteristic is aided with again and multiple dimensioned horizontal scheduling algorithm is finely adjusted and optimized to segmentation result；
Cell rapid classification identification module, for the image after segmentation to be identified, divides into individual cells or cell cluster；
Individual cells are carried out respectively using the double-current convolutional neural networks in additional knowledge field and the cell knowledge mapping of structure
Hierarchical identification, respectively obtains the first classification results and second of classification results；Using the double-current convolutional Neural net of cell cluster
The identification of the inseparable cell cluster of network model realization；With
Interpretation and post-processing module, combine sentencing for the first classification results to individual cells and second of classification results
Read, carry out clash handle, obtain the classification results of individual cells；The clash handle is different for solving various features sensing
During sentence read result, comprehensive various factors eliminates conflict, makes clearly reliable interpretation；Then knowledge mapping and class activity are utilized
Readability, the interpretation of cervical cell recognition result of Mapping implementation cervical cell identification process.
2. cervical cell liquid-based smear artificial intelligence according to claim 1 aids in diagosis system, it is characterised in that described
Cell image acquisition module, using 40 times of amplifications of eyepiece, scanning pattern is rectangle, and scan mode scans for eclipsed form so that
Scanning range is all covered with liquid-based smear cells location.
3. cervical cell liquid-based smear artificial intelligence according to claim 1 aids in diagosis system, it is characterised in that described
Pretreatment includes：Denoising is carried out to image using two-sided filter, the wave filter is made up of two functions：One function be by
Geometric space distance determines filter coefficient, and another determines filter coefficient by pixel value difference；Then Morphological scale-space is used
The edge of image is repaired, filling cavity simultaneously removes thin connection, finally increases nucleus using histogram equalization
With cytoplasmic contrast.
4. cervical cell liquid-based smear artificial intelligence according to claim 1 aids in diagosis system, it is characterised in that described
Cell image detects that the step that implements of segmentation module is：
Step S1) preceding background coarse segmentation is carried out to pretreated cell image, extract the region belonging to cell；
Step S2) to after coarse segmentation cell image carry out cell component detection split, use fast area convolutional Neural net
Network is partitioned into different types of cell；
Step S3) detect and split cervical cell nucleus；
Step S4) nucleus is screened according to the characteristic parameter of nucleus, obtain final candidate cell core；
Step S5) judgment step S2) in obtained cell type whether be cell cluster, if it is not, then using active contour
Model and Prior Template carry out the segmentation of cytosolic domain；Otherwise, it is transferred to step S6)；
Step S6) comprehensive nucleus post-processed with cytoplasmic segmentation result and domain knowledge, the whole cervical cell of completion
Effective segmentation.
5. cervical cell liquid-based smear artificial intelligence according to claim 4 aids in diagosis system, it is characterised in that described
Fast area convolutional neural networks are from the VGG16 of convolutional neural networks network structure, and the size of input picture is 515*
512, the detection classification of final cell composition falls into 5 types：Scale cell, gland cell, neck tube cell, metaplasia cell and background element
Matter, the cell component in addition to scale cell and background quality is all defined as inseparable cell cluster.
6. cervical cell liquid-based smear artificial intelligence according to claim 4 aids in diagosis system, it is characterised in that described
Step S5) in use movable contour model and Prior Template carry out the detailed process of segmentation of cytosolic domain and be：
Improved movable contour model is employed, energy function and shape priors are added, the excellent of profile is made iteratively
Change, obtain cytoplasmic exact boundary；
Energy function E (u) is：
E (u)=λ1Es(u)+R(u)
Wherein Es(u) it is shape prior, R (u) is regular terms, the flatness for ensureing partitioning boundary, λ1For the ginseng that can learn
Number；Shape prior Es(u) it is：
<mrow>
<msub>
<mi>E</mi>
<mi>s</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Integral;</mo>
<mi>&Pi;</mi>
</munder>
<mi>H</mi>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>(</mo>
<mi>p</mi>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>p</mi>
</mrow>
Wherein, H is Hessian matrix.
7. cervical cell liquid-based smear artificial intelligence according to claim 1 aids in diagosis system, it is characterised in that described
Cell image detects that the step that implements of segmentation module is：
Step 1) pretreatment operation is carried out to the cervical cell image after segmentation；
Step 2) whether judge pretreated cell image be individual cells, if it is, being transferred to step 3), otherwise, the image
For inseparable cell cluster, step 6 is transferred to)；
Step 3) computable cell parameters is determined, then calculate cell parameters feature；
Step 4) cell knowledge mapping reasoning and judging model is set up, and cell parameters feature is inputted into the model, obtain single thin
The first classification results of born of the same parents；
Step 5) the double-current convolutional neural networks model of additional fields knowledge is built, cell parameters feature and cell image are inputted
Double-current convolutional neural networks model, obtains second of classification results of individual cells；
Step 6) the double-current convolutional neural networks model of cell cluster is built, and inseparable cell cluster is carried out using the model
The hierarchical identification of cell cluster, obtains the classification results of cell cluster.
8. cervical cell liquid-based smear artificial intelligence according to claim 7 aids in diagosis system, it is characterised in that described
The input all the way of the double-current convolutional neural networks of additional fields knowledge is step 3) obtained cell parameters feature, it is defeated all the way in addition
Enter for individual cells image, size is uniformly normalized to 256*256 pixel values, by the convolution pond composite modules of 5 cascades
Implicitly extract the feature of cell image；The convolution kernel size of most important of which convolution operation uses 7*7 sizes, step-length choosing
It is 1 to select size, and characteristic pattern number is chosen for 96, and the convolution operation is：
<mrow>
<mi>x</mi>
<mo>=</mo>
<mi>f</mi>
<mrow>
<mo>(</mo>
<munder>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>M</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>*</mo>
<msub>
<mi>w</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>b</mi>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
In above formula, M represents the set of the input feature vector figure of selection, wijRepresent weight, bjExport what is added for each characteristic pattern
Then 1096 dimensional features of extraction are spliced to one by one additional bias plus computable 20 dimensional feature of cell field knowledge
Rise, be input to the full articulamentum and classification layer of double-current convolutional neural networks.
9. cervical cell liquid-based smear artificial intelligence according to claim 7 aids in diagosis system, it is characterised in that described
The input all the way of the double-current convolutional neural networks of cell cluster is：The feature of queueing discipline between nucleus, it is defeated all the way in addition
Enter be cell cluster corresponding with the cell parameters cervical cell, cervical cell input size be uniformly normalized to 512*512
Pixel value, the convolution pond composite module by 8 cascades implicitly extracts the feature of cell image；Wherein, it is most important
The convolution kernel size of convolution operation uses 5*5 sizes, and step-length selection size is 2, and characteristic pattern number is chosen for 108.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710351064.XA CN107274386B (en) | 2017-05-18 | 2017-05-18 | artificial intelligent auxiliary cervical cell fluid-based smear reading system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710351064.XA CN107274386B (en) | 2017-05-18 | 2017-05-18 | artificial intelligent auxiliary cervical cell fluid-based smear reading system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107274386A true CN107274386A (en) | 2017-10-20 |
| CN107274386B CN107274386B (en) | 2019-12-17 |
Family
ID=60064694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710351064.XA Active CN107274386B (en) | 2017-05-18 | 2017-05-18 | artificial intelligent auxiliary cervical cell fluid-based smear reading system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107274386B (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107995521A (en) * | 2017-12-07 | 2018-05-04 | 南京熊猫电子制造有限公司 | A kind of medical display all-in-one machine control system for being used for intelligent diagosis |
| CN108334860A (en) * | 2018-03-01 | 2018-07-27 | 北京航空航天大学 | The treating method and apparatus of cell image |
| CN108389198A (en) * | 2018-02-27 | 2018-08-10 | 深思考人工智能机器人科技（北京）有限公司 | The recognition methods of atypia exception gland cell in a kind of Cervical smear |
| CN108982500A (en) * | 2018-07-03 | 2018-12-11 | 怀光智能科技（武汉）有限公司 | A kind of cervical liquid-based cells intelligence auxiliary diagosis method and system |
| CN109190441A (en) * | 2018-06-21 | 2019-01-11 | 丁彦青 | Female genital tract cells pathology intelligent method for classifying, diagnostic equipment and storage medium |
| CN109272492A (en) * | 2018-08-24 | 2019-01-25 | 深思考人工智能机器人科技（北京）有限公司 | A kind of processing method and system of cell pathology smear |
| CN109690562A (en) * | 2018-05-18 | 2019-04-26 | 香港应用科技研究院有限公司 | Accelerate the image preprocessing of cytology image classification by full convolutional neural networks |
| CN109815888A (en) * | 2019-01-21 | 2019-05-28 | 武汉兰丁医学高科技有限公司 | A kind of novel Papanicolau staining process and abnormal cervical cells automatic identifying method |
| CN110189310A (en) * | 2019-05-24 | 2019-08-30 | 上海联影智能医疗科技有限公司 | Acquisition methods, computer equipment and the storage medium of image feature value |
| CN110231259A (en) * | 2019-05-28 | 2019-09-13 | 怀光智能科技（武汉）有限公司 | A kind of cervical cell slide numerical dialing system |
| WO2019218393A1 (en) * | 2018-05-18 | 2019-11-21 | Hong Kong Applied Science and Technology Research Institute Company Limited | Image pre-processing for accelerating cytological image classification by fully convolutional neural networks |
| CN110736747A (en) * | 2019-09-03 | 2020-01-31 | 深思考人工智能机器人科技（北京）有限公司 | cell liquid based smear under-mirror positioning method and system |
| CN111429761A (en) * | 2020-02-28 | 2020-07-17 | 中国人民解放军陆军军医大学第二附属医院 | Artificial intelligent simulation teaching system and method for bone marrow cell morphology |
| CN111785364A (en) * | 2020-06-15 | 2020-10-16 | 杭州思柏信息技术有限公司 | Internet and cervical image intelligent auxiliary film reading method and auxiliary film reading system |
| CN112088394A (en) * | 2018-07-24 | 2020-12-15 | 狄希斯医药有限公司 | Computerized classification of biological tissue |
| CN112102277A (en) * | 2020-09-10 | 2020-12-18 | 深圳市森盈生物科技有限公司 | Device and method for detecting tumor cells in pleural fluid fluorescence image |
| CN112257681A (en) * | 2019-12-06 | 2021-01-22 | 珠海圣美生物诊断技术有限公司 | Cell interpretation method and system |
| CN112396583A (en) * | 2020-11-18 | 2021-02-23 | 深思考人工智能机器人科技（北京）有限公司 | Method and system for DNA (deoxyribonucleic acid) ploid quantitative analysis based on Papanicolaou staining mode |
| CN112884663A (en) * | 2021-01-18 | 2021-06-01 | 北京晶科瑞医学检验实验室有限公司 | Method for identifying and dividing cell boundaries aiming at tissue mass spectrum imaging result |
| CN113052806A (en) * | 2021-03-15 | 2021-06-29 | 黑龙江机智通智能科技有限公司 | Canceration degree grading system |
| CN113111926A (en) * | 2021-03-31 | 2021-07-13 | 南京华晟医学检验实验室有限公司 | Abnormal cervical blood cell screening method based on TCT (TCT) slide |
| CN113222911A (en) * | 2021-04-26 | 2021-08-06 | 清华大学深圳国际研究生院 | Cervical cell image screening method, cervical cell image screening system, computer equipment and storage medium |
| CN113256577A (en) * | 2021-05-18 | 2021-08-13 | 湖南医药学院 | Cancer auxiliary analysis system and device based on HE staining pathological image |
| CN113420653A (en) * | 2021-06-22 | 2021-09-21 | 深圳跃美生物医学科技有限公司 | Method, device and readable medium for comprehensive detection and analysis of bacterial cell morphology |
| CN113902669A (en) * | 2021-08-24 | 2022-01-07 | 苏州深思考人工智能科技有限公司 | Method and system for reading urine exfoliative cell fluid-based smear |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1106246A (en) * | 1993-11-01 | 1995-08-09 | 波拉技术有限公司 | Method and apparatus for tissue type recognition |
| WO2007059119A2 (en) * | 2005-11-11 | 2007-05-24 | Ivan Borozan | Systems and methods for identifying diagnostic indicators |
| CN101460090A (en) * | 2006-06-05 | 2009-06-17 | 福思光子学有限公司 | Methods for characterizing tissues |
| CN105894490A (en) * | 2015-11-05 | 2016-08-24 | 广西师范大学 | Fuzzy integration multiple classifier integration-based uterine neck cell image identification method and device |
-
2017
- 2017-05-18 CN CN201710351064.XA patent/CN107274386B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1106246A (en) * | 1993-11-01 | 1995-08-09 | 波拉技术有限公司 | Method and apparatus for tissue type recognition |
| WO2007059119A2 (en) * | 2005-11-11 | 2007-05-24 | Ivan Borozan | Systems and methods for identifying diagnostic indicators |
| CN101460090A (en) * | 2006-06-05 | 2009-06-17 | 福思光子学有限公司 | Methods for characterizing tissues |
| CN105894490A (en) * | 2015-11-05 | 2016-08-24 | 广西师范大学 | Fuzzy integration multiple classifier integration-based uterine neck cell image identification method and device |
Cited By (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107995521A (en) * | 2017-12-07 | 2018-05-04 | 南京熊猫电子制造有限公司 | A kind of medical display all-in-one machine control system for being used for intelligent diagosis |
| CN108389198A (en) * | 2018-02-27 | 2018-08-10 | 深思考人工智能机器人科技（北京）有限公司 | The recognition methods of atypia exception gland cell in a kind of Cervical smear |
| CN108334860A (en) * | 2018-03-01 | 2018-07-27 | 北京航空航天大学 | The treating method and apparatus of cell image |
| WO2019218393A1 (en) * | 2018-05-18 | 2019-11-21 | Hong Kong Applied Science and Technology Research Institute Company Limited | Image pre-processing for accelerating cytological image classification by fully convolutional neural networks |
| CN109690562B (en) * | 2018-05-18 | 2022-09-13 | 香港应用科技研究院有限公司 | Image pre-processing to accelerate cytological image classification by full convolution neural network |
| US10586336B2 (en) | 2018-05-18 | 2020-03-10 | Hong Kong Applied Science and Technology Research Institute Company Limited | Image pre-processing for accelerating cytological image classification by fully convolutional neural networks |
| CN109690562A (en) * | 2018-05-18 | 2019-04-26 | 香港应用科技研究院有限公司 | Accelerate the image preprocessing of cytology image classification by full convolutional neural networks |
| CN109190441A (en) * | 2018-06-21 | 2019-01-11 | 丁彦青 | Female genital tract cells pathology intelligent method for classifying, diagnostic equipment and storage medium |
| CN109190441B (en) * | 2018-06-21 | 2022-11-08 | 丁彦青 | Intelligent classification method, diagnostic instrument and storage medium for female genital tract cell pathology |
| CN108982500B (en) * | 2018-07-03 | 2020-07-14 | 怀光智能科技（武汉）有限公司 | Intelligent auxiliary cervical fluid-based cytology reading method and system |
| CN108982500A (en) * | 2018-07-03 | 2018-12-11 | 怀光智能科技（武汉）有限公司 | A kind of cervical liquid-based cells intelligence auxiliary diagosis method and system |
| US12002573B2 (en) | 2018-07-24 | 2024-06-04 | Dysis Medical Limited | Computer classification of biological tissue |
| US11562820B2 (en) | 2018-07-24 | 2023-01-24 | Dysis Medical Limited | Computer classification of biological tissue |
| CN112088394A (en) * | 2018-07-24 | 2020-12-15 | 狄希斯医药有限公司 | Computerized classification of biological tissue |
| CN109272492A (en) * | 2018-08-24 | 2019-01-25 | 深思考人工智能机器人科技（北京）有限公司 | A kind of processing method and system of cell pathology smear |
| CN109272492B (en) * | 2018-08-24 | 2022-02-15 | 深思考人工智能机器人科技（北京）有限公司 | Method and system for processing cytopathology smear |
| CN109815888A (en) * | 2019-01-21 | 2019-05-28 | 武汉兰丁医学高科技有限公司 | A kind of novel Papanicolau staining process and abnormal cervical cells automatic identifying method |
| CN110189310B (en) * | 2019-05-24 | 2022-03-25 | 上海联影智能医疗科技有限公司 | Image characteristic value acquisition method, computer device and storage medium |
| CN110189310A (en) * | 2019-05-24 | 2019-08-30 | 上海联影智能医疗科技有限公司 | Acquisition methods, computer equipment and the storage medium of image feature value |
| CN110231259A (en) * | 2019-05-28 | 2019-09-13 | 怀光智能科技（武汉）有限公司 | A kind of cervical cell slide numerical dialing system |
| CN110231259B (en) * | 2019-05-28 | 2022-02-18 | 怀光智能科技（武汉）有限公司 | Digital diagnosis system for cervical cell slide |
| CN110736747A (en) * | 2019-09-03 | 2020-01-31 | 深思考人工智能机器人科技（北京）有限公司 | cell liquid based smear under-mirror positioning method and system |
| CN110736747B (en) * | 2019-09-03 | 2022-08-19 | 深思考人工智能机器人科技（北京）有限公司 | Method and system for positioning under cell liquid-based smear mirror |
| CN112257681A (en) * | 2019-12-06 | 2021-01-22 | 珠海圣美生物诊断技术有限公司 | Cell interpretation method and system |
| WO2021110143A1 (en) * | 2019-12-06 | 2021-06-10 | 珠海圣美生物诊断技术有限公司 | Cell interpretation method and system |
| CN111429761A (en) * | 2020-02-28 | 2020-07-17 | 中国人民解放军陆军军医大学第二附属医院 | Artificial intelligent simulation teaching system and method for bone marrow cell morphology |
| CN111785364A (en) * | 2020-06-15 | 2020-10-16 | 杭州思柏信息技术有限公司 | Internet and cervical image intelligent auxiliary film reading method and auxiliary film reading system |
| CN112102277A (en) * | 2020-09-10 | 2020-12-18 | 深圳市森盈生物科技有限公司 | Device and method for detecting tumor cells in pleural fluid fluorescence image |
| CN112396583B (en) * | 2020-11-18 | 2024-01-26 | 深思考人工智能机器人科技（北京）有限公司 | DNA ploidy quantitative analysis method and system based on Papanicolaou staining mode |
| CN112396583A (en) * | 2020-11-18 | 2021-02-23 | 深思考人工智能机器人科技（北京）有限公司 | Method and system for DNA (deoxyribonucleic acid) ploid quantitative analysis based on Papanicolaou staining mode |
| CN112884663A (en) * | 2021-01-18 | 2021-06-01 | 北京晶科瑞医学检验实验室有限公司 | Method for identifying and dividing cell boundaries aiming at tissue mass spectrum imaging result |
| CN112884663B (en) * | 2021-01-18 | 2023-11-21 | 北京晶科瑞医学检验实验室有限公司 | Method for identifying and dividing cell boundaries aiming at tissue mass spectrum imaging result |
| CN113052806A (en) * | 2021-03-15 | 2021-06-29 | 黑龙江机智通智能科技有限公司 | Canceration degree grading system |
| CN113052806B (en) * | 2021-03-15 | 2023-02-28 | 黑龙江机智通智能科技有限公司 | Canceration degree grading system |
| CN113111926A (en) * | 2021-03-31 | 2021-07-13 | 南京华晟医学检验实验室有限公司 | Abnormal cervical blood cell screening method based on TCT (TCT) slide |
| CN113222911A (en) * | 2021-04-26 | 2021-08-06 | 清华大学深圳国际研究生院 | Cervical cell image screening method, cervical cell image screening system, computer equipment and storage medium |
| CN113256577A (en) * | 2021-05-18 | 2021-08-13 | 湖南医药学院 | Cancer auxiliary analysis system and device based on HE staining pathological image |
| CN113256577B (en) * | 2021-05-18 | 2022-06-28 | 湖南医药学院 | Cancer auxiliary analysis system and device based on HE staining pathological image |
| CN113420653A (en) * | 2021-06-22 | 2021-09-21 | 深圳跃美生物医学科技有限公司 | Method, device and readable medium for comprehensive detection and analysis of bacterial cell morphology |
| CN113902669A (en) * | 2021-08-24 | 2022-01-07 | 苏州深思考人工智能科技有限公司 | Method and system for reading urine exfoliative cell fluid-based smear |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107274386B (en) | 2019-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107274386A (en) | A kind of cervical cell liquid-based smear artificial intelligence aids in diagosis system | |
| Jia et al. | Detection of cervical cancer cells based on strong feature CNN-SVM network | |
| CN107886514B (en) | Mammary gland molybdenum target image lump semantic segmentation method based on depth residual error network | |
| CN107256558A (en) | The cervical cell image automatic segmentation method and system of a kind of unsupervised formula | |
| CN106780460B (en) | A kind of Lung neoplasm automatic checkout system for chest CT images | |
| Man et al. | Classification of breast cancer histopathological images using discriminative patches screened by generative adversarial networks | |
| Joshi et al. | Classification of brain cancer using artificial neural network | |
| Albayrak et al. | Mitosis detection using convolutional neural network based features | |
| Beevi et al. | Automatic mitosis detection in breast histopathology images using convolutional neural network based deep transfer learning | |
| Dundar et al. | Computerized classification of intraductal breast lesions using histopathological images | |
| Jiang et al. | A novel white blood cell segmentation scheme using scale-space filtering and watershed clustering | |
| CN109903284B (en) | HER2 immunohistochemical image automatic discrimination method and system | |
| CN108334909A (en) | Cervical carcinoma TCT digital slices data analysing methods based on ResNet | |
| Pan et al. | Mitosis detection techniques in H&E stained breast cancer pathological images: A comprehensive review | |
| CN109389129A (en) | A kind of image processing method, electronic equipment and storage medium | |
| CN109635846A (en) | A kind of multiclass medical image judgment method and system | |
| CN105894490A (en) | Fuzzy integration multiple classifier integration-based uterine neck cell image identification method and device | |
| CN110288582A (en) | A kind of accurate dividing method of nucleus | |
| CN110264454B (en) | Cervical cancer histopathological image diagnosis method based on multi-hidden-layer conditional random field | |
| Jia et al. | Multi-layer segmentation framework for cell nuclei using improved GVF Snake model, Watershed, and ellipse fitting | |
| Cao et al. | An automatic breast cancer grading method in histopathological images based on pixel-, object-, and semantic-level features | |
| Apou et al. | Detection of lobular structures in normal breast tissue | |
| Anari et al. | Computer-aided detection of proliferative cells and mitosis index in immunohistichemically images of meningioma | |
| Zhang et al. | Automatic detection of invasive ductal carcinoma based on the fusion of multi-scale residual convolutional neural network and SVM | |
| Raman et al. | Geometric approach to segmentation and protein localization in cell culture assays |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20171020 Assignee: Suzhou deep thinking Artificial Intelligence Technology Co.,Ltd. Assignor: IDEEPWISE ARTIFICIAL INTELLIGENCE ROBOT TECHNOLOGY (BEIJING) CO.,LTD. Contract record no.: X2022980003658 Denomination of invention: An artificial intelligence assisted film reading system for cervical cell liquid based smear Granted publication date: 20191217 License type: Common License Record date: 20220401 |
|
| EE01 | Entry into force of recordation of patent licensing contract |