CN109190567A - Abnormal cervical cells automatic testing method based on depth convolutional neural networks - Google Patents
Abnormal cervical cells automatic testing method based on depth convolutional neural networks Download PDFInfo
- Publication number
- CN109190567A CN109190567A CN201811048234.8A CN201811048234A CN109190567A CN 109190567 A CN109190567 A CN 109190567A CN 201811048234 A CN201811048234 A CN 201811048234A CN 109190567 A CN109190567 A CN 109190567A
- Authority
- CN
- China
- Prior art keywords
- cell
- cervical
- abnormal
- neural networks
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/695—Preprocessing, e.g. image segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/698—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Multimedia (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of abnormal cervical cells automatic testing method based on depth convolutional neural networks, mainly comprises the steps that the artificial mark of (1) TCT cervical cell picture;(2) segmentation of cell picture has been marked;(3) more classification of cell after dividing;(4) training of the cell picture of all categories in migration models;(5) to segmentation, identification, the positioning of cytological map piece to be detected;(6) for the dividing processing again and identification of abnormal cell group.Cell after segmentation is divided into 9 classes by abnormal cervical cells automatic testing method of the present invention, is trained using transfer learning mode, obtains being fitted preferable more disaggregated models.The model can be identified, screened and be accurately positioned to unlabelled source cervical cell picture, and do dividing processing again and identification to the abnormal cell detected, improve the accuracy rate of detection.The present invention has preferable auxiliaring effect in cervical cell pathological diagnosis field.
Description
Technical field
The invention belongs to cervical cell pathological diagnosis fields, and in particular to a kind of based on the different of depth convolutional neural networks
Normal cervical cell automatic testing method.
Background technique
Cervical carcinoma is high-incidence in recent years, has become the social concern for threatening women life.Current effective diagnosis of cervical cancer
Method is cervical smear pathologic finding, and this method is paid a home visit after needing pathologist to find sick cell by micro- sem observation
It is disconnected.This aspect needs to expend a large amount of manpower and material resources, on the other hand the accuracy of diagnosis be easy by doctor's subjective factor or
The influence of visual fatigue.Therefore the cervical cell pathological diagnosis technology of automation becomes more and more important.
Existing cervical cell identification technology is usually based on fine segmentation and feature extraction, however due to adopting in microscope
The overlapping of the cell image of collection, presence that is irregular, dyeing the problems such as uneven bring very big be stranded to the fine segmentation of cell
Difficulty, while can there are problems that validity feature can not be extracted during feature extraction or excessively introduce invalid feature, Wu Faqu
Obtain good effect.
For existing abnormal cervical cells automatic measurement technique, negative, two classification of the positive is mostly used to do identification inspection greatly
It surveys, since the similitude and cell mass of abnormal cervical cells and basal cell are difficult to discrimination property, classifying quality is caused to be paid no attention to
Think.And current cell classification is mainly all directed to individual cells, and excessive processing is not done for cell mass, leads to cell mass
Detection effect it is undesirable.Meanwhile existing abnormal cervical cells automatic identifying method mainly all with classification based on, it is desirable that input
It is the picture after dividing processing, the initial acquisition sample of cervical cell cannot be handled directly, abnormal palace cannot be reached
The direct purpose of the identification and positioning of neck cell.
Summary of the invention
In view of the deficiencies of the prior art, the present invention proposes a kind of abnormal cervical cells based on depth convolutional neural networks certainly
Dynamic detection method.
The technical solution adopted by the present invention to solve the technical problems mainly includes the following steps:
Step 1: the source cervical cell image of acquisition manually being marked, obtaining includes abnormal cervical cells labeling position information
File.
Step 2: processing is split according to mark gained file and source cervical cell image;
1) cell mass is divided using threshold method first: being carried out pair according to the location information in the location information of segmentation and mark file
Than, judge whether to belong to abnormal cervical cells group, after be partitioned into the location information 299 ' 299 cervical cell group, threshold value point
It is as follows to cut formula:
, (1)
Wherein,Indicate the gray value of pixel in target image,Indicate the gray value of pixel in source images,Indicate taken threshold value.
2) discrete cellular is partitioned into using Adaptive Thresholding: determines the position of nucleus first, cell nuclear location is believed
Location information in breath and mark file compares, and judges whether to belong to abnormal cell, then be divided centered on nucleus
The cell picture of 299 ' 299 sizes out.
Step 3: combination cell pathology and computer technology relevant knowledge are classified the cell after segmentation, carefully
Born of the same parents group mainly includes that negative cervical cell mass and positive cervical cell roll into a ball two parts;Discrete cellular is broadly divided into negative single squamous
Cell, negative single basal cell, neutral grain, negative overlapping squamous cell, negative overlapping basal cell, impurity and positive uterine neck
Cell.
Step 4: various types of cells picture obtained above being overturn, the processing such as contrast variation, expand the sample of data
Amount, balance it is each it is different classes of in cell image quantity.
Step 5: the convolutional neural networks model trained in advance on ImageNet being modified, institute in reserving model
There is the parameter of convolutional layer, modify last full articulamentum, using cross entropy loss function, formula is expressed as follows shown
, (2)
Wherein,Indicate sample size,Expression parameter set,Indicate required loss function,Indicate sampleLabel,Indicate sampleHypothesis function.
Using stochastic gradient descent optimization algorithm, specific formula is expressed as follows:
, (3)
Wherein,Indicate required parameter,Indicate loss functionGradient,Indicate learning rate, neural network
Using dropout mode, for neural network unit, it is temporarily abandoned from network according to certain probability, it can be effective
Prevent over-fitting.Transfer learning mode is used simultaneously, and the various types of cells picture that above-mentioned steps are obtained inputs modified convolution
Neural network model is trained, and is stored to preferable depth convolutional neural networks model is finally fitted.
Step 6: in the convolutional neural networks model after any one source cervical cell picture not marked input is finely tuned,
It can correctly identify the position of positive cervical cell and be marked, next to the positive cervical cell region marked
Screening is compared according to location information, processing is filtered for the cell for having repetition to cover, guarantees source cervical cell picture
The accuracy of middle positive cervical cell positioning.
Step 7: the abnormal cell recognized for convolutional network model or abnormal cell group, for further processing.First
Determine whether for cell mass, it is then current without any processing if discrete cellular;Adaptive Thresholding is then used if cell mass
It processes, it is first determined then the position of nucleus in cell mass is carried out centered on nucleus using watershed algorithm slender
The segmentation of born of the same parents.It is unicellular according to what is obtained, the calculating of the parameters such as nucleus size, shape, color and karyoplasmic ratio is carried out, into one
It walks and determines cell generic.
Beneficial effects of the present invention:
The present invention has abandoned traditional feminine gender, positive two classification methods, combination cell pathology and computer technology relevant knowledge,
Cell after segmentation cell is done into more classification processings, has higher discrimination compared to two classification methods.Due to medical data sheet
Body is difficult to obtain, and profound neural network framework is trained to need a large amount of tape label data, so the present invention is using migration
Mode of learning, being one for the model modification after ImageNet is trained in advance can satisfy the polytypic network of cervical cell
Model, and network model according to available data can train the preferably more disaggregated models of degree of fitting according to this.More classification moulds
Type can directly identify the source cervical cell picture not marked, detect wherein positive cell and positioned.For volume
The abnormal cell group that product network detects, makees further dividing processing, and calculate relevant parameter and identified, is further promoted
The accuracy of testing result.The present invention has preferable auxiliaring effect in cervical cell pathological diagnosis field.
Detailed description of the invention
Fig. 1 is source cervical cell image labeling cutting procedure figure;
Fig. 2 is cell classification image, is always divided into 9 classes, is respectively: negative cervical cell mass, positive cervical cell group, negative single
A squamous cell, negative single basal cell, feminine gender overlapping squamous cell, feminine gender overlapping basal cell, neutrophil leucocyte are miscellaneous
Matter, positive cervical cell;
Fig. 3 is the training detection schematic diagram of depth convolutional neural networks model.
Specific embodiment
Below in conjunction with attached drawing, the present invention is further described.
As shown, the cervical cell automatic testing method based on depth convolutional neural networks, the specific implementation steps are as follows:
Step 1: as shown in Figure 1, carrying out artificial mark abnormal cervical cells to the source cervical cell image of acquisition, obtaining comprising different
The file of normal cell labeling position information.
Step 2: as shown in Figure 1, being split processing according to mark gained file and source cervical cell image;
1) cell mass is divided using threshold method first, is carried out pair according to the location information in the location information of segmentation and mark file
Than, judge whether to belong to abnormal cell, after be partitioned into the location information 299 ' 299 cervical cell group, Threshold segmentation formula
It is as follows:
, (4)
Wherein,Indicate the gray value of pixel in target image,Indicate the gray value of pixel in source images,Indicate taken threshold value.
2) discrete cellular is divided using Adaptive Thresholding, it is first determined the location information of nucleus, by cell nuclear location
Location information in information and mark file compares, and judges whether to belong to abnormal cell, is then divided centered on nucleus
Cut out the cell picture of 299 ' 299 sizes.
Step 3: the cell after segmentation being classified, cell mass mainly includes negative cervical cell mass and positive uterine neck
Cell mass two parts;Discrete cellular is broadly divided into negative single squamous cell, negative single basal cell, neutral grain, negative weight
Folded squamous cell, negative overlapping basal cell, impurity and positive cervical cell.Cell classification image is as shown in Figure 2.
Step 4: various types of cells picture obtained above being overturn, the processing such as contrast variation, expand the sample of data
Amount, balance it is each it is different classes of in cell image quantity.
Step 5: as shown in the pre-training process of Fig. 3, convolutional neural networks mould that will be trained in advance on ImageNet
Type is modified, and the parameter of all convolutional layers in reserving model modifies last full articulamentum, using cross entropy loss function,
Shown in formula is expressed as follows
, (5)
Wherein,Indicate sample size,Expression parameter set,Indicate required loss function,Indicate sampleLabel,Indicate sampleHypothesis function.
Using stochastic gradient descent optimization algorithm, specific formula is expressed as follows:
, (6)
Wherein,Indicate required parameter,Indicate loss functionGradient,Indicate learning rate, neural network
Using dropout mode, for neural network unit, it is temporarily abandoned from network according to certain probability, it can be effective
Prevent over-fitting.As shown in the training process of Fig. 3, while using transfer learning mode, all kinds of uterine neck that above-mentioned steps are obtained
Cell picture inputs modified convolutional neural networks model and is trained, and to being finally fitted preferable convolutional neural networks mould
Type is stored.
Step 6: as shown in Fig. 3 detection process, after any one source cervical cell picture input fine tuning to be detected
In convolutional neural networks model, it can correctly identify the position of abnormal cervical cells and be marked, next to label
Screening is compared according to location information in abnormal cervical cells region out, and the cell for having repetition to cover is filtered, and protects
The accuracy that abnormal cervical cells position in the cervical cell picture of card source.
Step 7: the abnormal cell recognized for convolutional network model or abnormal cell group, for further processing.First
Determine whether for cell mass, it is then current without any processing if discrete cellular;Adaptive Thresholding is then used if cell mass
It processes, it is first determined then the position of nucleus in cell mass is carried out centered on nucleus using watershed algorithm slender
The segmentation of born of the same parents.Unicellular, the calculating of the parameters such as progress nucleus size, shape, color and karyoplasmic ratio is obtained according to segmentation,
Further determine that cell generic.
Claims (6)
1. the abnormal cervical cells automatic testing method based on depth convolutional neural networks, it is characterised in that include the following steps:
Step 1: the source cervical cell image of acquisition is manually marked;
Step 2: source cervical cell image being split according to mark gained information;
Step 3: classifying to the cell after segmentation;
Step 4: various types of cells picture obtained above is pre-processed;
Step 5: using transfer learning mode, the pre-training model after the input change of treated cell is trained;
Step 6: a cervical cell picture input model in source to be detected can be correctly detected into out abnormal cell position;
Step 7: to abnormal cell, group makees further dividing processing, unicellular carry out nucleus size, the shape obtained to segmentation
The calculating of the parameters such as shape, color and karyoplasmic ratio further determines that cell generic.
2. the abnormal cervical cells automatic testing method according to claim 1 based on depth convolutional neural networks, special
Sign is: for source cervical cell picture, obtaining cell mass using Threshold segmentation, Adaptive Thresholding obtains discrete cellular, tool
Body formula is as follows:
, (1)
Wherein,Indicate the gray value of pixel in target image,Indicate the gray value of pixel in source images,Indicate taken threshold value.
3. the abnormal cervical cells automatic testing method according to claim 1 based on depth convolutional neural networks, special
Sign is: combination cell pathology and computer technology relevant knowledge do more classification processings, cell to the cell image after segmentation
Group is mainly cervical cell group, including negative and positive two parts;Discrete cellular be broadly divided into negative single squamous cell,
Negative single basal cell, neutral grain, negative overlapping squamous cell, negative overlapping basal cell, impurity and positive cervical cell.
4. the abnormal cervical cells automatic testing method according to claim 1 based on depth convolutional neural networks, special
Sign is: in the training process of convolutional neural networks, using dropout method, and although there is certain extension in the training time,
Effectively prevent over-fitting.
5. the abnormal cervical cells automatic testing method according to claim 1 based on depth convolutional neural networks, special
Sign is: by inputting any one unlabelled source cervical cell picture, which can correctly identify different
The band of position of normal cervical cell is simultaneously marked, and the cell compartment for having repetition to cover is filtered, and guarantees abnormal palace
The correct identification and accurate positionin of neck cell.
6. the abnormal cervical cells automatic testing method according to claim 1 based on depth convolutional neural networks, special
Sign is: further dividing processing is done, by joining to obtained unicellular do by the abnormal cervical cells group obtained for detection
Number calculates, and further improves the accuracy of testing result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811048234.8A CN109190567A (en) | 2018-09-10 | 2018-09-10 | Abnormal cervical cells automatic testing method based on depth convolutional neural networks |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811048234.8A CN109190567A (en) | 2018-09-10 | 2018-09-10 | Abnormal cervical cells automatic testing method based on depth convolutional neural networks |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109190567A true CN109190567A (en) | 2019-01-11 |
Family
ID=64915532
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811048234.8A Pending CN109190567A (en) | 2018-09-10 | 2018-09-10 | Abnormal cervical cells automatic testing method based on depth convolutional neural networks |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109190567A (en) |
Cited By (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109815945A (en) * | 2019-04-01 | 2019-05-28 | 上海徒数科技有限公司 | A kind of respiratory tract inspection result interpreting system and method based on image recognition |
CN109948429A (en) * | 2019-01-28 | 2019-06-28 | 上海依智医疗技术有限公司 | Image analysis method, device, electronic equipment and computer-readable medium |
CN110009050A (en) * | 2019-04-10 | 2019-07-12 | 杭州智团信息技术有限公司 | A kind of classification method and device of cell |
CN110060244A (en) * | 2019-04-15 | 2019-07-26 | 深圳市麦迪普科技有限公司 | The system and method for cell detection and segmentation based on deep learning neural network |
CN110378313A (en) * | 2019-07-26 | 2019-10-25 | 玖壹叁陆零医学科技南京有限公司 | Cell mass recognition methods, device and electronic equipment |
CN110415212A (en) * | 2019-06-18 | 2019-11-05 | 平安科技(深圳)有限公司 | Abnormal cell detection method, device and computer readable storage medium |
CN110598561A (en) * | 2019-08-15 | 2019-12-20 | 平安科技(深圳)有限公司 | Cell slide analysis method and device based on machine learning and storage medium |
CN110675411A (en) * | 2019-09-26 | 2020-01-10 | 重庆大学 | Cervical squamous intraepithelial lesion recognition algorithm based on deep learning |
CN110765855A (en) * | 2019-09-12 | 2020-02-07 | 杭州迪英加科技有限公司 | Pathological image processing method and system |
CN111126162A (en) * | 2019-11-28 | 2020-05-08 | 东软集团股份有限公司 | Method, device and storage medium for identifying inflammatory cells in image |
CN111461165A (en) * | 2020-02-26 | 2020-07-28 | 上海商汤智能科技有限公司 | Image recognition method, recognition model training method, related device and equipment |
CN111612027A (en) * | 2019-02-26 | 2020-09-01 | 沛智生医科技股份有限公司 | Cell classification method, system and medical analysis platform |
CN111783571A (en) * | 2020-06-17 | 2020-10-16 | 陕西中医药大学 | Cervical cell automatic classification model establishment and cervical cell automatic classification method |
CN111833354A (en) * | 2020-07-21 | 2020-10-27 | 哈尔滨理工大学 | Cervical cell image segmentation system based on deep learning |
CN112215807A (en) * | 2020-09-24 | 2021-01-12 | 北京航空航天大学 | Cell image automatic classification method and system based on deep learning |
CN112365471A (en) * | 2020-11-12 | 2021-02-12 | 哈尔滨理工大学 | Cervical cancer cell intelligent detection method based on deep learning |
CN112884725A (en) * | 2021-02-02 | 2021-06-01 | 杭州迪英加科技有限公司 | Correction method for neural network model output result for cell discrimination |
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 |
CN113178228A (en) * | 2021-05-25 | 2021-07-27 | 郑州中普医疗器械有限公司 | Cell analysis method based on nuclear DNA analysis, computer device, and storage medium |
CN113295692A (en) * | 2021-05-25 | 2021-08-24 | 郑州中普医疗器械有限公司 | Cell analysis method based on cell nucleus DNA and TBS double analysis method, computer equipment and storage medium |
CN113378967A (en) * | 2021-06-28 | 2021-09-10 | 哈尔滨工业大学 | Structural health monitoring multivariate data anomaly diagnosis method based on convolutional neural network and transfer learning |
WO2021227708A1 (en) * | 2020-05-09 | 2021-11-18 | 腾讯科技(深圳)有限公司 | Method and device for identifying abnormal cell in sample to be tested, and storage medium |
CN114037868A (en) * | 2021-11-04 | 2022-02-11 | 杭州医策科技有限公司 | Image recognition model generation method and device |
CN115100646A (en) * | 2022-06-27 | 2022-09-23 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
CN116309280A (en) * | 2022-12-16 | 2023-06-23 | 上海药明康德新药开发有限公司 | Lymphocyte labeling method and system |
CN113295692B (en) * | 2021-05-25 | 2024-06-21 | 郑州中普医疗器械有限公司 | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium |
-
2018
- 2018-09-10 CN CN201811048234.8A patent/CN109190567A/en active Pending
Cited By (33)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109948429A (en) * | 2019-01-28 | 2019-06-28 | 上海依智医疗技术有限公司 | Image analysis method, device, electronic equipment and computer-readable medium |
CN111612027A (en) * | 2019-02-26 | 2020-09-01 | 沛智生医科技股份有限公司 | Cell classification method, system and medical analysis platform |
CN109815945B (en) * | 2019-04-01 | 2024-04-30 | 上海徒数科技有限公司 | Respiratory tract examination result interpretation system and method based on image recognition |
CN109815945A (en) * | 2019-04-01 | 2019-05-28 | 上海徒数科技有限公司 | A kind of respiratory tract inspection result interpreting system and method based on image recognition |
CN110009050A (en) * | 2019-04-10 | 2019-07-12 | 杭州智团信息技术有限公司 | A kind of classification method and device of cell |
CN110060244A (en) * | 2019-04-15 | 2019-07-26 | 深圳市麦迪普科技有限公司 | The system and method for cell detection and segmentation based on deep learning neural network |
CN110060244B (en) * | 2019-04-15 | 2023-07-14 | 深圳市麦迪普科技有限公司 | Cell detection and segmentation system and method based on deep learning neural network |
CN110415212A (en) * | 2019-06-18 | 2019-11-05 | 平安科技(深圳)有限公司 | Abnormal cell detection method, device and computer readable storage medium |
CN110378313A (en) * | 2019-07-26 | 2019-10-25 | 玖壹叁陆零医学科技南京有限公司 | Cell mass recognition methods, device and electronic equipment |
CN110378313B (en) * | 2019-07-26 | 2021-05-18 | 玖壹叁陆零医学科技南京有限公司 | Cell cluster identification method and device and electronic equipment |
CN110598561A (en) * | 2019-08-15 | 2019-12-20 | 平安科技(深圳)有限公司 | Cell slide analysis method and device based on machine learning and storage medium |
CN110765855A (en) * | 2019-09-12 | 2020-02-07 | 杭州迪英加科技有限公司 | Pathological image processing method and system |
CN110675411A (en) * | 2019-09-26 | 2020-01-10 | 重庆大学 | Cervical squamous intraepithelial lesion recognition algorithm based on deep learning |
CN110675411B (en) * | 2019-09-26 | 2023-05-16 | 重庆大学 | Cervical squamous intraepithelial lesion recognition algorithm based on deep learning |
CN111126162A (en) * | 2019-11-28 | 2020-05-08 | 东软集团股份有限公司 | Method, device and storage medium for identifying inflammatory cells in image |
CN111461165A (en) * | 2020-02-26 | 2020-07-28 | 上海商汤智能科技有限公司 | Image recognition method, recognition model training method, related device and equipment |
WO2021227708A1 (en) * | 2020-05-09 | 2021-11-18 | 腾讯科技(深圳)有限公司 | Method and device for identifying abnormal cell in sample to be tested, and storage medium |
CN111783571A (en) * | 2020-06-17 | 2020-10-16 | 陕西中医药大学 | Cervical cell automatic classification model establishment and cervical cell automatic classification method |
CN111833354A (en) * | 2020-07-21 | 2020-10-27 | 哈尔滨理工大学 | Cervical cell image segmentation system based on deep learning |
CN112215807A (en) * | 2020-09-24 | 2021-01-12 | 北京航空航天大学 | Cell image automatic classification method and system based on deep learning |
CN112365471A (en) * | 2020-11-12 | 2021-02-12 | 哈尔滨理工大学 | Cervical cancer cell intelligent detection method based on deep learning |
CN112365471B (en) * | 2020-11-12 | 2022-06-24 | 哈尔滨理工大学 | Cervical cancer cell intelligent detection method based on deep learning |
CN112884725A (en) * | 2021-02-02 | 2021-06-01 | 杭州迪英加科技有限公司 | Correction method for neural network model output result for cell discrimination |
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 |
CN113295692A (en) * | 2021-05-25 | 2021-08-24 | 郑州中普医疗器械有限公司 | Cell analysis method based on cell nucleus DNA and TBS double analysis method, computer equipment and storage medium |
CN113178228A (en) * | 2021-05-25 | 2021-07-27 | 郑州中普医疗器械有限公司 | Cell analysis method based on nuclear DNA analysis, computer device, and storage medium |
CN113295692B (en) * | 2021-05-25 | 2024-06-21 | 郑州中普医疗器械有限公司 | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium |
CN113378967A (en) * | 2021-06-28 | 2021-09-10 | 哈尔滨工业大学 | Structural health monitoring multivariate data anomaly diagnosis method based on convolutional neural network and transfer learning |
CN113378967B (en) * | 2021-06-28 | 2022-11-08 | 哈尔滨工业大学 | Structural health monitoring multivariate data anomaly diagnosis method based on convolutional neural network and transfer learning |
CN114037868A (en) * | 2021-11-04 | 2022-02-11 | 杭州医策科技有限公司 | Image recognition model generation method and device |
CN115100646A (en) * | 2022-06-27 | 2022-09-23 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
CN115100646B (en) * | 2022-06-27 | 2023-01-31 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
CN116309280A (en) * | 2022-12-16 | 2023-06-23 | 上海药明康德新药开发有限公司 | Lymphocyte labeling method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109190567A (en) | Abnormal cervical cells automatic testing method based on depth convolutional neural networks | |
CN106056118B (en) | A kind of identification method of counting for cell | |
Quelhas et al. | Cell nuclei and cytoplasm joint segmentation using the sliding band filter | |
Song et al. | A deep learning based framework for accurate segmentation of cervical cytoplasm and nuclei | |
CN113723573B (en) | Tumor tissue pathological classification system and method based on adaptive proportion learning | |
CN107274386A (en) | A kind of cervical cell liquid-based smear artificial intelligence aids in diagosis system | |
CN104598885B (en) | The detection of word label and localization method in street view image | |
CN112819821B (en) | Cell nucleus image detection method | |
CN108334860A (en) | The treating method and apparatus of cell image | |
CN111079620B (en) | White blood cell image detection and identification model construction method and application based on transfer learning | |
CN108765412A (en) | A kind of steel strip surface defect sorting technique | |
CN108898595A (en) | A kind of construction method of thoracopathy detection model and application | |
CN107610087A (en) | A kind of tongue fur automatic division method based on deep learning | |
CN110827310A (en) | CT image automatic detection method and system | |
CN109635846A (en) | A kind of multiclass medical image judgment method and system | |
CN110705403A (en) | Cell sorting method, cell sorting device, cell sorting medium, and electronic apparatus | |
CN109670489B (en) | Weak supervision type early senile macular degeneration classification method based on multi-instance learning | |
CN105069818A (en) | Image-analysis-based skin pore identification method | |
CN108961242A (en) | A kind of fluorescent staining image CTC intelligent identification Method | |
CN106340000A (en) | Bone age assessment method | |
CN106875404A (en) | The intelligent identification Method of epithelial cell in a kind of leukorrhea micro-image | |
CN114240978B (en) | Cell edge segmentation method and device based on adaptive morphology | |
CN109002831A (en) | A kind of breast density classification method, system and device based on convolutional neural networks | |
CN114240874A (en) | Bone age assessment method and device based on deep convolutional neural network and feature fusion and computer readable storage medium | |
JP7326316B2 (en) | Method for detecting cells with at least one abnormality in a cytological sample |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190111 |
|
WD01 | Invention patent application deemed withdrawn after publication |