CN109360198A - Bone marrwo cell sorting method and sorter based on deep learning - Google Patents
Bone marrwo cell sorting method and sorter based on deep learning Download PDFInfo
- Publication number
- CN109360198A CN109360198A CN201811168824.4A CN201811168824A CN109360198A CN 109360198 A CN109360198 A CN 109360198A CN 201811168824 A CN201811168824 A CN 201811168824A CN 109360198 A CN109360198 A CN 109360198A
- Authority
- CN
- China
- Prior art keywords
- cell
- bone marrow
- bone
- marrwo
- image
- 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
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
-
- 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
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Abstract
The present invention provides a kind of bone marrwo cell sorting method and sorter based on deep learning, wherein this method comprises: to the bone marrow cell mark cell position and its tag along sort in bone marrow cell sample image；The image block sample with single tag along sort of pre-set dimension is extracted from bone marrow cell sample image；The convolutional neural networks for constructing bone marrwo cell sorting task, then utilize the training set being made of image block sample to be trained, obtain bone marrwo cell sorting model；Bone marrow cell testing image is cut into multiple test image blocks of pre-set dimension, multiple test image blocks are inputted into bone marrwo cell sorting model with traversing, it detects the bone marrow cell edge in multiple test image blocks, and exports the corresponding tag along sort of bone marrow cell and classification fiducial probability.
Description
Technical field
The present invention relates to a kind of computer and software technology fields, and in particular to a kind of bone marrow cell based on deep learning
Classification method and sorter.
Background technique
Leukaemia is initiated by the malignant tumour of hemopoietic system, and disease incidence ranked sixth position in various tumours.Cell shape
State is using a kind of most, most extensive, most direct and most economical important diagnostic means in Diagnosis of Acute Leukemia, is form
, immunology, cytogenetics, the important component of Molecular strain typing diagnosis.Morphological method is mainly by patient
Bone marrow smear and blood film carry out Rui Shi-Giemsa staining analysis respectively, and further give other cytochemical stainings, according to
FAB (French, American, Britain) standard determines acute leukemia type.
However, bone marrow cell includes multiple systems, it is respectively that cell is divided into original, inmature, mature three phases again.Although adhering to separately
Different is, but due to being differentiated by marrow hemopoietic stem cells, there is no so big in the imagination, ginsengs for intercellular difference
Examine Fig. 1.How various bone marrow cells are precisely identified in bone marrow smear and statistical distribution ratio becomes technological difficulties.
At present in actual operation, using manually operated alternative, heavy workload is examined, repeatability is poor, not only time-consuming consumption
Power, doctor's continuous work easily because of fatigue or careless initiation wrong identification, influence condition-inference, and objective to morphologic description shortage
Quantitative criterion.Meanwhile diagnostic level depends on the experience of doctor to a certain extent.Therefore, need to develop automation,
By the bone marrwo cell sorting method and sorter of computer image processing technology, this is to the whole water for improving leukemia diagnosis
It is flat to be of great significance.
Summary of the invention
In view of this, the present invention provides a kind of bone marrwo cell sorting method and sorter based on deep learning, it can
The technical problem that low efficiency, the subjectivity of the solution prior art are strong, error is big, fidelity factor is bad.
To achieve the above object, according to an aspect of the invention, there is provided a kind of bone marrow cell based on deep learning
Classification method, comprising: to the bone marrow cell mark cell position and its tag along sort in the bone marrow cell sample image；From
The image block sample with single tag along sort of pre-set dimension is extracted in the bone marrow cell sample image；Construct bone marrow cell
Then the convolutional neural networks of classification task utilize the training set being made of described image block sample to be trained, obtain marrow
Cell classification model；Bone marrow cell testing image is cut into multiple test image blocks of the pre-set dimension, it will be the multiple
Test image block traversal ground inputs the bone marrwo cell sorting model, detects the bone marrow cell in the multiple test image block
Edge, and export the corresponding tag along sort of bone marrow cell and classification fiducial probability.
Optionally, further includes: the bone marrow cell sample image and the bone marrow cell testing image are normalized
Pretreatment.
Optionally, extracted from the bone marrow cell sample image described pre-set dimension with single tag along sort
After the step of image block sample, further includes: carry out the processing of data augmentation to described image block sample, wherein the data increase
Wide processing includes: rotation processing, mirror image processing, scaling processing.
Optionally, in the step of convolutional neural networks of the building bone marrwo cell sorting task, the convolutional Neural net
Network is Retinanet+Focal loss.
Optionally, the tag along sort include: rubricyte, metarubricyte, other erythroid cells, initial cell,
Mature lymphocyte, other leaching are cell, monokaryon system cell, progranulocyte, myelocyte, metamylocyte, band form nucleus
Cell, segmented cell and other myeloid cells.
To achieve the above object, according to another aspect of the present invention, it is also proposed that a kind of marrow based on deep learning is thin
Born of the same parents' sorter, comprising: labeling module, for in the bone marrow cell sample image bone marrow cell mark cell position with
And its tag along sort；Sampling module, for from the bone marrow cell sample image extract pre-set dimension have single classification
The image block sample of label；Then modeling module is utilized for constructing the convolutional neural networks of bone marrwo cell sorting task by institute
The training set for stating image block sample composition is trained, and obtains bone marrwo cell sorting model；Categorization module is used for bone marrow cell
Testing image is cut into multiple test image blocks of the pre-set dimension, the multiple test image block is inputted with traversing described
Bone marrwo cell sorting model detects the bone marrow cell edge in the multiple test image block, and it is corresponding to export bone marrow cell
Tag along sort and classification fiducial probability.
It optionally, further include preprocessing module, for be measured to the bone marrow cell sample image and the bone marrow cell
Pretreatment is normalized in image.
It optionally, further include augmentation module, for carrying out the processing of data augmentation to described image block sample, wherein described
The processing of data augmentation includes: rotation processing, mirror image processing, scaling processing.
Optionally, in the modeling module, the convolutional neural networks are Retinanet+Focal loss.
Optionally, the tag along sort include: rubricyte, metarubricyte, other erythroid cells, initial cell,
Mature lymphocyte, other leaching are cell, monokaryon system cell, progranulocyte, myelocyte, metamylocyte, band form nucleus
Cell, segmented cell and other myeloid cells.
According to the technique and scheme of the present invention, detection classification can be carried out to bone marrow cell based on deep learning, at least had
As follows the utility model has the advantages that (1) can detect bone marrow cell from image and judge cell category, high-efficient, favorable reproducibility；
(2) intelligence degree is high, has self-learning characteristics, determines characteristic factor without artificial；(3) algorithm simplicity is apparent, model instruction
It is low to hardware requirement in the test application stage after the completion of white silk.
Detailed description of the invention
Attached drawing for a better understanding of the present invention, does not constitute an undue limitation on the present invention.Wherein:
Fig. 1 is the schematic diagram of candidate stem cell differentiation；
Fig. 2 is the flow diagram of the bone marrwo cell sorting method according to an embodiment of the present invention based on deep learning；
Fig. 3 is the structural schematic diagram of the bone marrwo cell sorting device according to an embodiment of the present invention based on deep learning；
Fig. 4 is the schematic diagram for the neural network that the training pattern process of a specific embodiment of the invention uses.
Specific embodiment
Below in conjunction with attached drawing, an exemplary embodiment of the present invention will be described, including the various of the embodiment of the present invention
Details should think them only exemplary to help understanding.Therefore, those of ordinary skill in the art should recognize
It arrives, it can be with various changes and modifications are made to the embodiments described herein, without departing from scope and spirit of the present invention.Together
Sample, for clarity and conciseness, descriptions of well-known functions and structures are omitted from the following description.
Fig. 2 is the flow diagram of the bone marrwo cell sorting method according to an embodiment of the present invention based on deep learning.Such as
Shown in Fig. 2, this method may include following step S1 to step S4.
S1: to the bone marrow cell mark cell position and its tag along sort in the bone marrow cell sample image.Wherein,
Tag along sort may include: rubricyte, metarubricyte, other erythroid cells, initial cell, mature lymphocyte, its
It is cell, monokaryon system cell, progranulocyte, myelocyte, metamylocyte, band-cell, segmented cell that he, which drenches,
With other myeloid cells.
S2: the image block sample with single tag along sort of pre-set dimension is extracted from the bone marrow cell sample image
This.Wherein, pre-set dimension can be 1024*1024.Pre-set dimension is excessive will improve it is (such as aobvious to hardware in calculating process
Deposit) demand, pre-set dimension Guo is small, may be switched to entire cell in multiple images block sample, be unfavorable for identifying.
S3: then the convolutional neural networks of building bone marrwo cell sorting task are utilized and are made of described image block sample
Training set is trained, and obtains bone marrwo cell sorting model.Wherein, convolutional neural networks can use Retinanet+Focal
loss.Retinanet can be more than the precision of two step detectors, and speed is similar with single step detector.Focal loss is just
It is a loss layer for solving class imbalance in classification problem, difficulty difference of classifying.
S4: bone marrow cell testing image is cut into multiple test image blocks of the pre-set dimension, by the multiple survey
It tries image block traversal ground and inputs the bone marrwo cell sorting model, detect the bone marrow cell side in the multiple test image block
Edge, and export the corresponding tag along sort of bone marrow cell and classification fiducial probability.
Optionally, the bone marrwo cell sorting method based on deep learning of the embodiment of the present invention further include: to the marrow
Pretreatment is normalized in cell sample image and the bone marrow cell testing image.It is as follows to normalize pretreated process: first
First, the mean value and variance of RGB triple channel are calculated separately using samples all in sample set.When training and test, sample image is needed
The mean value of RGB triple channel is subtracted, and divided by variance.This preprocessing process can make the strong of each pixel of sample
Degree, which complies with standard, to be just distributed very much, accelerates convergence when training.
Optionally, the bone marrwo cell sorting method based on deep learning of the embodiment of the present invention upon step s 2, is also wrapped
It includes: the processing of data augmentation is carried out to described image block sample, wherein data augmentation processing includes: rotation processing, at mirror image
Reason, scaling processing.The processing of data augmentation is specific as follows: flip horizontal, vertical to overturn, and is rotated by 90 °, 180 degree, 270 degree and contracting
It puts, then cuts and obtain sample.Using lossless data enhancing and slight scaling, it can effectively keep marrow thin here
The original characteristic of born of the same parents, cell type changes after avoiding enhancing.
Fig. 3 is the structural schematic diagram of the bone marrwo cell sorting device according to an embodiment of the present invention based on deep learning.Such as
Shown in Fig. 3, which includes: labeling module 100, sampling module 200, modeling module 300 and categorization module 400.
Labeling module 100 is used for the bone marrow cell mark cell position and its point in the bone marrow cell sample image
Class label.Wherein, tag along sort includes: rubricyte, metarubricyte, other erythroid cells, initial cell, mature lymph
Cell, other leaching are cell, monokaryon system cell, progranulocyte, myelocyte, metamylocyte, band-cell, leaflet
Nucleus and other myeloid cells.
Sampling module 200 for from the bone marrow cell sample image extract pre-set dimension have single tag along sort
Image block sample.Wherein, pre-set dimension can be 1024*1024.Pre-set dimension is excessive will to be improved in calculating process to hard
The demand of part (such as video memory), pre-set dimension Guo is small, may be switched to entire cell in multiple images block sample, be unfavorable for knowing
Not.
Modeling module 300 is used to construct the convolutional neural networks of bone marrwo cell sorting task, then utilizes by described image
The training set of block sample composition is trained, and obtains bone marrwo cell sorting model.
Categorization module 400 is used to for bone marrow cell testing image being cut into multiple test image blocks of the pre-set dimension,
The multiple test image block is inputted into the bone marrwo cell sorting model with traversing, is detected in the multiple test image block
Bone marrow cell edge, and export the corresponding tag along sort of bone marrow cell and classification fiducial probability.
Optionally, the bone marrwo cell sorting device based on deep learning of the embodiment of the present invention further includes preprocessing module,
For pretreatment to be normalized to the bone marrow cell sample image and the bone marrow cell testing image.Normalization pretreatment
Process it is as follows: firstly, calculating separately the mean value and variance of RGB triple channel using samples all in sample set.Training and test
When, sample image needs to subtract the mean value of RGB triple channel, and divided by variance.This preprocessing process can make the every of sample
The intensity of a pixel, which complies with standard, to be just distributed very much, accelerates convergence when training.
Optionally, the bone marrwo cell sorting device based on deep learning of the embodiment of the present invention further includes augmentation module, is used
In carrying out the processing of data augmentation to described image block sample, wherein data augmentation processing includes: rotation processing, at mirror image
Reason, scaling processing.The processing of data augmentation is specific as follows: flip horizontal, vertical to overturn, and is rotated by 90 °, 180 degree, 270 degree and contracting
It puts, then cuts and obtain sample.Using lossless data enhancing and slight scaling, it can effectively keep marrow thin here
The original characteristic of born of the same parents, cell type changes after avoiding enhancing.
Bone marrwo cell sorting method and device according to an embodiment of the present invention based on deep learning, can be based on depth
Habit detection classification is carried out to bone marrow cell, at least have the following beneficial effects: (1) can be detected from image bone marrow cell with
And judge cell category, high-efficient, favorable reproducibility；(2) intelligence degree is high, has self-learning characteristics, determines without artificial
Characteristic factor；(3) algorithm simplicity is apparent, low to hardware requirement in the test application stage after the completion of model training.
To more fully understand those skilled in the art, Retinanet and focal loss is done in detail below with reference to Fig. 4
It introduces.
Retinanet comes from paper " Focal Loss for Dense Object Detection ", this paper obtains
Iccv2017 best student papers.The method of the Retinanet+focal loss mentioned in paper is (and used herein
Method), using resnet as feature extraction network, multi-scale information effectively is obtained using fpn (feature pyramid), is lost
It calculates and uses focal loss, effectively inhibit imbalance problem between class, this method is trained on public data collection, test
As a result, map approaches two stages the network model of (two stage), and because itself it is the net of a stage (one stage)
Network, so having a clear superiority in speed.
In Retinanet, sorter network Resnet is removed into global average pondization and full connection, as its feature extraction
Network, the rear pyramidal structure of introduced feature, the feature that effective use feature extraction network extracts, and generate multiple dimensioned defeated
Out, prediction result has better robustness to various sizes of object.The costing bio disturbance of Retinanet, using focal
loss.The main contribution of Focal loss is the problem of effectively inhibiting positive and negative imbalanced training sets, and contributes difficult sample
More losses.The biasing initialization for sub-network of classifying uses b=-log ((1- π)=π), solves to be likely to occur when training starts
Instability problem.
Because in real world, although up to more than 70 kinds of bone marrow cell, often occur in the actually collected visual field
Include mature monocyte, neutral paging core, neutral band form nucleus, myelocyte, rubricyte, blood platelet, mature leaching
Bar cell, metarubricyte and metamylocyte.Our algorithm only trains this 9 kinds of cells, the bone of other classifications in training
We label it as -1 to myelocyte, do not calculate its loss when training.The recall rate that final algorithm predicts this 9 class
Up to 82%.
Describe basic principle of the invention in conjunction with specific embodiments above, in the apparatus and method of the present invention, it is clear that
Each component or each step can be decomposed and/or be reconfigured.These decompose and/or reconfigure should be regarded as it is of the invention etc.
Efficacious prescriptions case.Also, the step of executing above-mentioned series of processes can execute according to the sequence of explanation in chronological order naturally, still
It does not need centainly to execute sequentially in time.Certain steps can execute parallel or independently of one another.
Above-mentioned specific embodiment, does not constitute a limitation on the scope of protection of the present invention.Those skilled in the art should be bright
It is white, design requirement and other factors are depended on, various modifications, combination, sub-portfolio and substitution can occur.It is any
Made modifications, equivalent substitutions and improvements etc. within the spirit and principles in the present invention, should be included in the scope of the present invention
Within.
Claims (10)
1. a kind of bone marrwo cell sorting method based on deep learning characterized by comprising
To the bone marrow cell mark cell position and its tag along sort in the bone marrow cell sample image；
The image block sample with single tag along sort of pre-set dimension is extracted from the bone marrow cell sample image；
The convolutional neural networks for constructing bone marrwo cell sorting task, then utilize the training set that is made of described image block sample into
Row training, obtains bone marrwo cell sorting model；
Bone marrow cell testing image is cut into multiple test image blocks of the pre-set dimension, by the multiple test image block
Traversal ground inputs the bone marrwo cell sorting model, detects the bone marrow cell edge in the multiple test image block, and defeated
The corresponding tag along sort of bone marrow cell and classification fiducial probability out.
2. the method according to claim 1, wherein further include: to the bone marrow cell sample image and described
Pretreatment is normalized in bone marrow cell testing image.
3. the method according to claim 1, wherein being extracted in advance from the bone marrow cell sample image described
If after the step of image block sample with single tag along sort of size, further includes: counted to described image block sample
It is handled according to augmentation, wherein the data augmentation processing includes: rotation processing, mirror image processing, scaling processing.
4. the method according to claim 1, wherein the convolutional Neural net of the building bone marrwo cell sorting task
In the step of network, the convolutional neural networks are Retinanet+Focal loss.
5. the method according to claim 1, wherein the tag along sort includes: that rubricyte, evening children are red thin
Born of the same parents, other erythroid cells, initial cell, mature lymphocyte, other leaching are cell, monokaryon system cell, progranulocyte, middle children
Granulocyte, metamylocyte, band-cell, segmented cell and other myeloid cells.
6. a kind of bone marrwo cell sorting device based on deep learning characterized by comprising
Labeling module, for the bone marrow cell mark cell position and its contingency table in the bone marrow cell sample image
Label；
Sampling module, for extracting the image with single tag along sort of pre-set dimension from the bone marrow cell sample image
Block sample；
Then modeling module is utilized for constructing the convolutional neural networks of bone marrwo cell sorting task by described image block sample
The training set of composition is trained, and obtains bone marrwo cell sorting model；
Categorization module will be described for bone marrow cell testing image to be cut into multiple test image blocks of the pre-set dimension
Multiple test image block traversals ground inputs the bone marrwo cell sorting model, detects the marrow in the multiple test image block
Cell edges, and export the corresponding tag along sort of bone marrow cell and classification fiducial probability.
7. device according to claim 6, which is characterized in that further include preprocessing module, for the bone marrow cell
Pretreatment is normalized in sample image and the bone marrow cell testing image.
8. device according to claim 6, which is characterized in that further include augmentation module, for described image block sample
Carry out the processing of data augmentation, wherein the data augmentation processing includes: rotation processing, mirror image processing, scaling processing.
9. device according to claim 6, which is characterized in that in the modeling module, the convolutional neural networks are
Retinanet+Focal loss。
10. device according to claim 6, which is characterized in that the tag along sort include: rubricyte, evening children it is red
Cell, other erythroid cells, initial cell, mature lymphocyte, other leaching be cell, monokaryon system cell, progranulocyte, in
Myelocyte, metamylocyte, band-cell, segmented cell and other myeloid cells.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811168824.4A CN109360198A (en) | 2018-10-08 | 2018-10-08 | Bone marrwo cell sorting method and sorter based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811168824.4A CN109360198A (en) | 2018-10-08 | 2018-10-08 | Bone marrwo cell sorting method and sorter based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109360198A true CN109360198A (en) | 2019-02-19 |
Family
ID=65348548
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811168824.4A Pending CN109360198A (en) | 2018-10-08 | 2018-10-08 | Bone marrwo cell sorting method and sorter based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109360198A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109961107A (en) * | 2019-04-18 | 2019-07-02 | 北京迈格威科技有限公司 | Training method, device, electronic equipment and the storage medium of target detection model |
| CN110223287A (en) * | 2019-06-13 | 2019-09-10 | 首都医科大学北京友谊医院 | A method of early diagnosing mammary cancer rate can be improved |
| CN110334565A (en) * | 2019-03-21 | 2019-10-15 | 江苏迪赛特医疗科技有限公司 | A kind of uterine neck neoplastic lesions categorizing system of microscope pathological photograph |
| CN111105416A (en) * | 2019-12-31 | 2020-05-05 | 北京理工大学重庆创新中心 | Automatic grading method and system for bone marrow cell proliferation degree |
| CN111310568A (en) * | 2020-01-16 | 2020-06-19 | 武汉大学 | Multiple myeloma bone marrow elephant identification equipment based on computer vision |
| CN111476754A (en) * | 2020-02-28 | 2020-07-31 | 中国人民解放军陆军军医大学第二附属医院 | Artificial intelligence auxiliary grading diagnosis system and method for bone marrow cell image |
| CN111488846A (en) * | 2020-04-16 | 2020-08-04 | 上海芯翌智能科技有限公司 | Method and equipment for identifying water level |
| CN111667472A (en) * | 2020-06-08 | 2020-09-15 | 江西卫生职业学院 | Artificial intelligent analysis method for bone marrow cell morphology |
| RU2732895C1 (en) * | 2019-05-27 | 2020-09-24 | Общество с ограниченной ответственностью "ПЛАТФОРМА ТРЕТЬЕ МНЕНИЕ" | Method for isolating and classifying blood cell types using deep convolution neural networks |
| CN112419262A (en) * | 2020-11-19 | 2021-02-26 | 上海联影智能医疗科技有限公司 | Bone marrow smear microscopic image analysis method and device, electronic equipment and storage medium |
| CN112504947A (en) * | 2020-12-03 | 2021-03-16 | 中国人民解放军陆军军医大学第二附属医院 | Morphological analysis and counting method for peripheral blood cells |
| CN113011306A (en) * | 2021-03-15 | 2021-06-22 | 中南大学 | Method, system and medium for automatic identification of bone marrow cell images in continuous maturation stage |
| CN113408480A (en) * | 2021-07-13 | 2021-09-17 | 上海交通大学医学院附属瑞金医院 | Artificial intelligent auxiliary diagnosis system for blood diseases based on bone marrow cell morphology |
| CN113604544A (en) * | 2021-08-03 | 2021-11-05 | 北京大学口腔医学院 | Biological material function prediction evaluation method |
| CN114187480A (en) * | 2022-01-05 | 2022-03-15 | 北京瀚远医药科技有限公司 | Bone marrow image cell image detection and classification method and system based on deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107729932A (en) * | 2017-10-10 | 2018-02-23 | 李强 | Bone marrow cell labeling method and system |
| CN107977684A (en) * | 2017-12-20 | 2018-05-01 | 杭州智微信息科技有限公司 | A kind of quick exchange method for correcting bone marrow nucleated cell classification |
| CN108021903A (en) * | 2017-12-19 | 2018-05-11 | 南京大学 | The error calibrating method and device of artificial mark leucocyte based on neutral net |
| CN108376400A (en) * | 2018-02-12 | 2018-08-07 | 华南理工大学 | A kind of bone marrow cell automatic classification method |
-
2018
- 2018-10-08 CN CN201811168824.4A patent/CN109360198A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107729932A (en) * | 2017-10-10 | 2018-02-23 | 李强 | Bone marrow cell labeling method and system |
| CN108021903A (en) * | 2017-12-19 | 2018-05-11 | 南京大学 | The error calibrating method and device of artificial mark leucocyte based on neutral net |
| CN107977684A (en) * | 2017-12-20 | 2018-05-01 | 杭州智微信息科技有限公司 | A kind of quick exchange method for correcting bone marrow nucleated cell classification |
| CN108376400A (en) * | 2018-02-12 | 2018-08-07 | 华南理工大学 | A kind of bone marrow cell automatic classification method |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110334565A (en) * | 2019-03-21 | 2019-10-15 | 江苏迪赛特医疗科技有限公司 | A kind of uterine neck neoplastic lesions categorizing system of microscope pathological photograph |
| CN109961107A (en) * | 2019-04-18 | 2019-07-02 | 北京迈格威科技有限公司 | Training method, device, electronic equipment and the storage medium of target detection model |
| CN109961107B (en) * | 2019-04-18 | 2022-07-19 | 北京迈格威科技有限公司 | Training method and device for target detection model, electronic equipment and storage medium |
| RU2732895C1 (en) * | 2019-05-27 | 2020-09-24 | Общество с ограниченной ответственностью "ПЛАТФОРМА ТРЕТЬЕ МНЕНИЕ" | Method for isolating and classifying blood cell types using deep convolution neural networks |
| CN110223287A (en) * | 2019-06-13 | 2019-09-10 | 首都医科大学北京友谊医院 | A method of early diagnosing mammary cancer rate can be improved |
| CN111105416A (en) * | 2019-12-31 | 2020-05-05 | 北京理工大学重庆创新中心 | Automatic grading method and system for bone marrow cell proliferation degree |
| CN111105416B (en) * | 2019-12-31 | 2022-09-09 | 北京理工大学重庆创新中心 | Automatic grading method and system for bone marrow cell proliferation degree |
| CN111310568A (en) * | 2020-01-16 | 2020-06-19 | 武汉大学 | Multiple myeloma bone marrow elephant identification equipment based on computer vision |
| CN111476754A (en) * | 2020-02-28 | 2020-07-31 | 中国人民解放军陆军军医大学第二附属医院 | Artificial intelligence auxiliary grading diagnosis system and method for bone marrow cell image |
| CN111476754B (en) * | 2020-02-28 | 2022-12-09 | 中国人民解放军陆军军医大学第二附属医院 | Bone marrow cell image artificial intelligence auxiliary grading diagnosis system and method |
| CN111488846A (en) * | 2020-04-16 | 2020-08-04 | 上海芯翌智能科技有限公司 | Method and equipment for identifying water level |
| CN111667472A (en) * | 2020-06-08 | 2020-09-15 | 江西卫生职业学院 | Artificial intelligent analysis method for bone marrow cell morphology |
| CN112419262A (en) * | 2020-11-19 | 2021-02-26 | 上海联影智能医疗科技有限公司 | Bone marrow smear microscopic image analysis method and device, electronic equipment and storage medium |
| CN112504947A (en) * | 2020-12-03 | 2021-03-16 | 中国人民解放军陆军军医大学第二附属医院 | Morphological analysis and counting method for peripheral blood cells |
| CN113011306A (en) * | 2021-03-15 | 2021-06-22 | 中南大学 | Method, system and medium for automatic identification of bone marrow cell images in continuous maturation stage |
| CN113408480A (en) * | 2021-07-13 | 2021-09-17 | 上海交通大学医学院附属瑞金医院 | Artificial intelligent auxiliary diagnosis system for blood diseases based on bone marrow cell morphology |
| CN113408480B (en) * | 2021-07-13 | 2022-09-30 | 上海交通大学医学院附属瑞金医院 | Artificial intelligent auxiliary diagnosis system for blood diseases based on bone marrow cell morphology |
| CN113604544A (en) * | 2021-08-03 | 2021-11-05 | 北京大学口腔医学院 | Biological material function prediction evaluation method |
| CN113604544B (en) * | 2021-08-03 | 2023-03-10 | 北京大学口腔医学院 | Biological material function prediction and evaluation method |
| CN114187480A (en) * | 2022-01-05 | 2022-03-15 | 北京瀚远医药科技有限公司 | Bone marrow image cell image detection and classification method and system based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109360198A (en) | Bone marrwo cell sorting method and sorter based on deep learning | |
| CN106056118B (en) | A kind of identification method of counting for cell | |
| CN106248559B (en) | A kind of five sorting technique of leucocyte based on deep learning | |
| CN108376400B (en) | Automatic bone marrow cell classification method | |
| Chang et al. | Nuclear segmentation in H&E sections via multi-reference graph cut (MRGC) | |
| Tomari et al. | Computer aided system for red blood cell classification in blood smear image | |
| Ramesh et al. | Isolation and two-step classification of normal white blood cells in peripheral blood smears | |
| Savkare et al. | Automatic detection of malaria parasites for estimating parasitemia | |
| CN108021903B (en) | Error calibration method and device for artificially labeling leucocytes based on neural network | |
| CN104331710B (en) | On off state identifying system | |
| CN108596038B (en) | Method for identifying red blood cells in excrement by combining morphological segmentation and neural network | |
| CN103984939B (en) | A kind of sample visible component sorting technique and system | |
| CN103323455A (en) | Tobacco leaf grading method based on reflection, perspective and microscopic images | |
| CN103679184A (en) | Method for leukocyte automatic identification based on relevant vector machine | |
| CN101526994B (en) | Fingerprint image segmentation method irrelevant to collecting device | |
| CN103020585A (en) | Method for identifying positive cells and negative cells of immunologic tissue | |
| CN112580748B (en) | Method for counting classified cells of stain image | |
| CN109636824A (en) | A kind of multiple target method of counting based on image recognition technology | |
| CN107730499A (en) | A kind of leucocyte classification method based on nu SVMs | |
| CN109815974A (en) | A kind of cell pathology slide classification method, system, equipment, storage medium | |
| CN116309368B (en) | Lung cancer pathological diagnosis system based on deep migration learning | |
| CN105303169A (en) | Slow characteristic based cell division recognition method and recognition device thereof | |
| US20230273188A1 (en) | Method and system for detecting cell-killing efficacy and/or immune activity, and application thereof | |
| CN108090928A (en) | A kind of method and system detected with screening similar round cell compartment | |
| Angulo et al. | Ontology-based lymphocyte population description using mathematical morphology on colour blood images |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190219 |