CN108918481A - A kind of huve cell system of fluorescence analysis - Google Patents
A kind of huve cell system of fluorescence analysis Download PDFInfo
- Publication number
- CN108918481A CN108918481A CN201810384676.3A CN201810384676A CN108918481A CN 108918481 A CN108918481 A CN 108918481A CN 201810384676 A CN201810384676 A CN 201810384676A CN 108918481 A CN108918481 A CN 108918481A
- Authority
- CN
- China
- Prior art keywords
- cell
- image
- nucleus
- seed point
- fluorescence
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
- G01N21/6428—Measuring fluorescence of fluorescent products of reactions or of fluorochrome labelled reactive substances, e.g. measuring quenching effects, using measuring "optrodes"
Landscapes
- Health & Medical Sciences (AREA)
- Immunology (AREA)
- Chemical & Material Sciences (AREA)
- Physics & Mathematics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Optics & Photonics (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention discloses a kind of huve cell system of fluorescence analysis.Medically in the detection of huve cell, the variation of the micro- sem observation endothelial cell of more doctor, by the present invention in that with the method for image procossing, dim, adhesion cell image, it is partitioned into cell one by one, and cell can be counted, the fluorescence intensity of each cell is measured.Mainly use pretreatment, Methods of Segmentation On Cell Images, the correction of cellular localization label, fluorescence intensity extractive technique, it can quickly realize the segmentation to viscous glutinous endothelial cell and the extraction of cell fluorescence intensity, the workload of artificial observation is greatly reduced, and can effectively avoid the analytical error due to caused by the diagnostic experiences and know-how of doctor.The fluorescence intensity of the different piece of each cell of analysis that can be quantitative, to provide foundation for diagnosis.
Description
Technical field
The present invention relates to a kind of huve cell system of fluorescence analysis, particularly belong to biological cell field.
Background technique
Vascular endothelial cell belongs to the barrier cell of segmentation vascular wall and blood, the exception of function and damage and more middle diseases
The generation of disease is related.Medically doctor mainly passes through microscope, using the changing features of eye-observation cell, then according to the palm
The pathological diagnosis experience and knowledge held carries out cell characteristic variation judgement, to obtain a qualitative diagnostic result.It is this
Method does not require nothing more than doctor's diagnostic experiences with higher and know-how, and result is also consumed vulnerable to extraneous interference and extremely
When.
The watershed algorithm based on range conversion applied in image segmentation of today, in the strip to the inclined rectangle of shape
When the segmentation of shape nucleus figure, there are problems that serious over-segmentation.And to segmentation cell image cell count when, due to according to
Lai Yu segmentation as a result, be easy to cause miscount and correction mistake.
Summary of the invention
Deficiency in view of the above technology, the present invention design a kind of huve cell system of fluorescence analysis, pass through cell
The segmentation to huve cell image is completed in the reference of core and cell image, and to the cell of umbilical vein cell image, carefully
Karyon, the fluorescence intensities such as cytoplasm extract analysis.Wherein, the algorithm being related to mainly has watershed algorithm, cell number
Change to count and be corrected in label, the extraction of fluorescence intensity.
A kind of huve cell system of fluorescence analysis:By being formed with lower module:Image pre-processing module, cell image
Divide module, cellular localization label correction module, fluorescence intensity extraction module, takes seed using nuclei picture and cell image
The accurate segmentation of cell and accurate counting in image are realized in the comparative analysis of image after point and optimization.
A kind of huve cell system of fluorescence analysis:Specifically include following steps:
Gray processing processing is carried out to 24 bit images of original cell and nucleus, the gray level image of generation is utilized into otsu
Algorithmic transformation is bianry image;
Range conversion is carried out using bianry image of the chamfering formwork to obtained nucleus and cell, obtains the gray scale of the two
Image, in gray level image at this moment, the gray level of each pixel is and its distance between nearest background;
After obtaining distance map, the gray value that pixel value on nucleus distance map is greater than or equal to eight pixels of surrounding is non-
0 pixel is denoted as seed point, since nuclear shapes are irregular, so will lead in a nucleus may have multiple kinds
It is sub-, seed point optimization is carried out using neighborhood method;
The gray value of optimized each seed point is set as 255, the gray value of other object pixels is assigned in image
1, background gray levels set 0, then scan seed point diagram, obtain nucleus and cell image completely apart from reconstruct image;
It using improved watershed algorithm, realizes in nucleus and cell image, glues point of glutinous nucleus and cell
It cuts;
The pixel in nucleus figure segmented and the gray value of the pixel at corresponding cell fluorescence grayscale image it
With greater than 255, then it is the position of nucleus, completes all nucleus and positioned in nucleus fluorescence picture, and carry out region mark
Fixed, method, realizes positioning of the cell in cell fluorescence picture, using the label of nucleus as reference, by pair of cell according to this
The label for being changed to nucleus should be marked;
Fluorescence intensity extraction, the fluorescence intensity ratio parameter of extraction are carried out to the cell, nucleus, cytoplasm of blip counting
Value can assist to judge the expression transformation of intracellular NF- kB protein and the occurrence characteristic of nuclear translocation phenomenon.
A kind of huve cell system of fluorescence analysis:Neighborhood method is in the step (3):To the kind initially searched out
Son point certain neighborhood judged, if in the neighborhood there are other seed points if illustrate that the seed point found later is redundancy
Seed point, need to delete them, i.e., their gray value be set to background gray levels 0, if it does not exist redundancy seed point, then
Any operation is not executed, only one seed point in individual cells core can be guaranteed by this method.
A kind of huve cell system of fluorescence analysis:Watershed algorithm is in the step (5):Obtain cell nuclear species
After sub- point image, using the seed point image of the available cell in the position of nucleus seed point in traditional watershed algorithm
On the basis of, setting one and image Mark Array of the same size are used to store the flag bit of image slices vegetarian refreshments, after improving
Watershed algorithm, pass through the segmentation that mark cycle realizes viscous collobast.
Beneficial effects of the present invention:
(1) in the segmentation of Human Umbilical Vein Endothelial Cells, due to dividing using the comparison of nuclei picture and cell image, kind is being taken
When son point, the optimization of seed point can be realized by neighborhood method according to respective positions in nuclei picture and cell image, it can be big
The redundancy of the reduction seed point of amplitude.
(2) using the method in label watershed, the segmentation of the viscous collobast of realization that can be ideal, and utilize nucleus
The comparison of image and cell image counts correction, realizes the accurate metering of cell in cell image, is accurately obtained each cell
Fluorescence intensity level.
(3) and in the prior art medically detection of Human Umbilical Vein Endothelial Cells, compared with the micro- sem observation endothelial cell of doctor
Variation, system of the invention can quickly realize the segmentation to viscous glutinous endothelial cell and the extraction of endothelial cell fluorescence intensity.
Reduce the workload of artificial observation, and the analysis due to caused by the diagnostic experiences and know-how of doctor can effectively be avoided to miss
Difference, the fluorescence intensity of the different piece of each cell of analysis that can be quantitative, to provide foundation for diagnosis.
Detailed description of the invention
Original 24 nucleus of Fig. 1 and cell;
Nucleus and cytological map after Fig. 2 binaryzation;
Seed point diagram after Fig. 3 optimization;
Nucleus and cytological map after Fig. 4 segmentation;
Cell compartment and nuclear area after Fig. 5 correction;
Cell and nucleus after Fig. 6 colorization label;
Nucleus and cell after Fig. 7 digitized markers;
Fig. 8 cell fluorescence intensity;
Fig. 9 nucleus fluorescence intensity;
Figure 10 cytoplasm fluorescence intensity;
Figure 11 nucleus and cytoplasm fluorescence intensity ratio.
Specific embodiment
Two new nucleus and cytological map are opened, system automatically replicates two pictures, presents respectively two thin
Original 24 color images of karyon and cell;See Figure of description 1.
To upper right nucleus and lower-left cell original image after gray processing, using otsu (Otsu algorithm) by image value
Change;See Figure of description 2.
Referring in Figure of description 3, the nuclei picture of the image after the range conversion of chamfering formwork, upper right is taken
Seed point and after seed point optimizes, obtains the seed point image of nucleus, using the seed point image of nucleus, obtains cell
Seed point image.
It reconstructs to obtain from each seed point thin using obtained nucleus and cell seed point diagram referring to Figure of description 4
Image after karyon and cell reconstitution, and watershed algorithm is used, viscous collobast core and cell are split.
Nuclear area and individual cells are positioned using the nucleus and cell image after segmentation referring to Figure of description 5
Region, and being corrected, obtain it is corrected in Figure of description Fig. 5 after lower-left cell compartment figure and bottom right nuclear area
Figure.
See Figure of description 6, using the cell compartment and nuclear area figure after correction, colorization is carried out to each region
Label.
See Figure of description 7, numerical count is carried out to the image after colorization label.
See Figure of description 8, the fluorescence intensity of cell is extracted, the cell after available each numeral mark
Fluorescence intensity.
See Figure of description 9, the fluorescence intensity of nucleus is extracted, the fluorescence of available each nuclear area
The numerical value of intensity.
See Figure of description 10, the fluorescence intensity of nucleus is subtracted using the fluorescence intensity of cell, it is available each thin
The fluorescence intensity number of born of the same parents' endochylema.
See Figure of description 11, calculates the nucleus of each cell and the fluorescence intensity ratio of cytoplasm.
Claims (4)
1. a kind of huve cell system of fluorescence analysis, it is characterised in that:By being formed with lower module:Image preprocessing mould
Block, Methods of Segmentation On Cell Images module, cellular localization mark correction module, fluorescence intensity extraction module, using nuclei picture and carefully
Born of the same parents' image takes the comparative analysis of image after seed point and optimization, realizes the accurate segmentation of cell and accurate counting in image.
2. a kind of huve cell system of fluorescence analysis according to claim 1, it is characterised in that:Specifically include with
Lower step:
(1) gray processing processing is carried out to 24 bit images of original cell and nucleus, the gray level image of generation is utilized into otsu
Algorithmic transformation is bianry image;
(2) range conversion is carried out using bianry image of the chamfering formwork to obtained nucleus and cell, obtains the gray scale of the two
Image, in gray level image at this moment, the gray level of each pixel is and its distance between nearest background;
(3) after obtaining distance map, the gray value that pixel value on nucleus distance map is greater than or equal to eight pixels of surrounding is non-zero
Pixel is denoted as seed point, since nuclear shapes are irregular, so will lead in a nucleus may have multiple kinds
It is sub-, seed point optimization is carried out using neighborhood method;
(4) gray value of optimized each seed point being set as 255, the gray value of other object pixels assigns 1 in image,
Background gray levels set 0, then scan seed point diagram, obtain nucleus and cell image completely apart from reconstruct image;
(5) improved watershed algorithm is used, is realized in nucleus and cell image, point of glutinous nucleus and cell is glued
It cuts;
(6) pixel in nucleus figure segmented and the sum of the gray value of pixel at corresponding cell fluorescence grayscale image
Then it is the position of nucleus greater than 255, completes all nucleus and positioned in nucleus fluorescence picture, and carry out region labeling,
Method according to this realizes positioning of the cell in cell fluorescence picture, and using the label of nucleus as reference, corresponding by cell is marked
Note is changed to the label of nucleus;
(7) fluorescence intensity extraction, the fluorescence intensity ratio parameter of extraction are carried out to the cell of blip counting, nucleus, cytoplasm
Value can assist to judge the expression transformation of intracellular NF- kB protein and the occurrence characteristic of nuclear translocation phenomenon.
3. a kind of huve cell system of fluorescence analysis according to claim 2, it is characterised in that:The step
(3) neighborhood method is in:Certain neighborhood of the seed point initially searched out is judged, if there are other seeds in the neighborhood
Point then illustrates that the seed point found later is the seed point of redundancy, needs to delete them, i.e., their gray value is set to back
Scape gray value 0, redundancy seed point, then do not execute any operation if it does not exist, can be guaranteed in individual cells core by this method
Only one seed point.
4. a kind of huve cell system of fluorescence analysis according to claim 2, it is characterised in that:The step
(5) improved watershed algorithm is in:After obtaining nucleus seed point image, it can be obtained using the position of nucleus seed point
To cell seed point image on the basis of traditional watershed algorithm, setting one with image conventional number of the same size
Group is used to store the flag bit of image slices vegetarian refreshments, by improved watershed algorithm, realizes viscous collobast by mark cycle
Segmentation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810384676.3A CN108918481A (en) | 2018-04-26 | 2018-04-26 | A kind of huve cell system of fluorescence analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810384676.3A CN108918481A (en) | 2018-04-26 | 2018-04-26 | A kind of huve cell system of fluorescence analysis |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108918481A true CN108918481A (en) | 2018-11-30 |
Family
ID=64403092
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810384676.3A Withdrawn CN108918481A (en) | 2018-04-26 | 2018-04-26 | A kind of huve cell system of fluorescence analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108918481A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116403211A (en) * | 2023-03-24 | 2023-07-07 | 无锡市第二人民医院 | Segmentation and clustering method and system based on single-cell pathology image cell nuclei |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104392460A (en) * | 2014-12-12 | 2015-03-04 | 山东大学 | Adherent white blood cell segmentation method based on nucleus-marked watershed transformation |
CN105160675A (en) * | 2015-08-31 | 2015-12-16 | 北京工业大学 | Automatic segmentation method for powdery mildew spore image |
-
2018
- 2018-04-26 CN CN201810384676.3A patent/CN108918481A/en not_active Withdrawn
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104392460A (en) * | 2014-12-12 | 2015-03-04 | 山东大学 | Adherent white blood cell segmentation method based on nucleus-marked watershed transformation |
CN105160675A (en) * | 2015-08-31 | 2015-12-16 | 北京工业大学 | Automatic segmentation method for powdery mildew spore image |
Non-Patent Citations (1)
Title |
---|
徐露: ""医学显微细胞图像分割算法及荧光强度提取研究"", 《中国优秀硕士学位论文全文数据库(电子期刊)》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116403211A (en) * | 2023-03-24 | 2023-07-07 | 无锡市第二人民医院 | Segmentation and clustering method and system based on single-cell pathology image cell nuclei |
CN116403211B (en) * | 2023-03-24 | 2024-04-26 | 无锡市第二人民医院 | Segmentation and clustering method and system based on single-cell pathology image cell nuclei |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106056118B (en) | A kind of identification method of counting for cell | |
JP7496389B2 (en) | Image analysis method, device, program, and method for manufacturing trained deep learning algorithm | |
US10262189B2 (en) | Evaluation of co-registered images of differently stained tissue slices | |
ES2826400T3 (en) | Colony contrast collection | |
CN111091527A (en) | Method and system for automatically detecting pathological change area in pathological tissue section image | |
CN108074243A (en) | A kind of cellular localization method and cell segmentation method | |
CN109492706A (en) | A kind of chromosome classification prediction meanss based on Recognition with Recurrent Neural Network | |
CN110517273B (en) | Cytology image segmentation method based on dynamic gradient threshold | |
CN109147932B (en) | Cancer cell HER2 gene amplification analysis method and system | |
CN115546605A (en) | Training method and device based on image labeling and segmentation model | |
CN113902669A (en) | Method and system for reading urine exfoliative cell fluid-based smear | |
CN115953393A (en) | Intracranial aneurysm detection system, equipment and storage medium based on multitask learning | |
CN110490971B (en) | Method for reconstructing cell dynamic characteristic three-dimensional image under biological microscope | |
Kromp et al. | Deep Learning architectures for generalized immunofluorescence based nuclear image segmentation | |
CN116883341A (en) | Liver tumor CT image automatic segmentation method based on deep learning | |
CN112270684B (en) | Microscopic image immunohistochemical virtual multiple labeling and analyzing method and system | |
CN114235539A (en) | PD-L1 pathological section automatic interpretation method and system based on deep learning | |
CN108918481A (en) | A kind of huve cell system of fluorescence analysis | |
CN114387596A (en) | Automatic interpretation system for cytopathology smear | |
CN109816665A (en) | A kind of fast partition method and device of optical coherence tomographic image | |
Feng et al. | An advanced automated image analysis model for scoring of ER, PR, HER-2 and Ki-67 in breast carcinoma | |
Parpulov et al. | Convolutional neural network application for cells segmentation in immunocytochemical study | |
CN114742803A (en) | Platelet aggregation detection method combining deep learning and digital image processing algorithm | |
Ortiz De Solorzano et al. | Applications of quantitative digital image analysis to breast cancer research | |
CN107977978B (en) | Method and system for determining cell adhesion mode based on elliptical axis relation |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181130 |