CN110415255B - Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system - Google Patents
Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system Download PDFInfo
- Publication number
- CN110415255B CN110415255B CN201910517113.1A CN201910517113A CN110415255B CN 110415255 B CN110415255 B CN 110415255B CN 201910517113 A CN201910517113 A CN 201910517113A CN 110415255 B CN110415255 B CN 110415255B
- Authority
- CN
- China
- Prior art keywords
- image
- segmentation
- cell nucleus
- processed
- immunohistochemical
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for segmenting a CD3 positive cell nucleus of an immunohistochemical pathological image, wherein the method comprises the following steps: performing color deconvolution on the immunohistochemical pathological image, and separating a staining channel; dividing the image into irregular pixel blocks by adopting superpixels, and carrying out kmeans clustering to distinguish and remove the image background; performing image segmentation based on morphological characteristics to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed; performing local threshold bernsen segmentation and morphological feature segmentation on the first image C1 to be processed to obtain a segmented cell nucleus second region image L2 and a second image C2 to be processed; carrying out foreground marking on the second image C2 to be processed, and segmenting a cell nucleus third area image L3 by adopting a watershed algorithm; the cell nucleus first region image L1, the cell nucleus second region image L2, and the cell nucleus third region image L3 constitute a cell nucleus segmentation image. The method has high robustness and accurate segmentation, and can meet the requirements of practical application.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a system for segmenting positive cell nuclei of an immunohistochemical pathological image CD 3.
Background
CD3 is an important leukocyte differentiation antigen, present on the surface of almost all T cells, and is a membrane antigen constituting the T cell antigen receptor (TCR). When an antigen binds to a TCR, it is involved in transmission of its signal into cells, and is an important membrane antigen involved in the discovery of various T cell functions. At present, a medical staff member needs to spend a lot of time and energy on segmenting, marking and counting the cells of the CD3 on pathological images, and therefore, a more accurate pathological image cell segmentation means is needed to reduce the pressure of the medical staff member.
Disclosure of Invention
In order to reduce the burden of medical staff and realize automatic segmentation of CD3 cells aiming at pathological images, the invention provides a method and a system for segmenting immunohistochemical pathological image CD3 positive cell nucleus.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a method for segmenting a CD3 positive cell nucleus of an immunohistochemical pathological image, which comprises the following steps of:
s1: performing color deconvolution on the immunohistochemical pathological image, and separating a staining channel;
s2: converting an original immunohistochemical pathological image from an RGB color space to a Lab color space, dividing the original immunohistochemical pathological image into irregular pixel blocks by adopting superpixels, carrying out kmeans clustering on the irregular pixel blocks to distinguish an image background, and removing the image background;
s3: performing image segmentation based on morphological characteristics to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed;
s4: performing local threshold bernsen segmentation and morphological feature segmentation on the first image C1 to be processed to obtain a segmented cell nucleus second region image L2 and a second image C2 to be processed;
s5: carrying out foreground marking on the second image C2 to be processed, and segmenting a cell nucleus third area image L3 by adopting a watershed algorithm;
s6: and masking the first cell nucleus region image L1, the second cell nucleus region image L2, the third cell nucleus region image L3 and the original immunohistochemical pathology image to obtain a cell nucleus segmentation image.
As a preferred technical solution, in step S1, the immunohistochemical pathology image is subjected to color deconvolution, and the calculation formula is:
C＝M-1[y]
where C denotes the separate H and DAB staining channels, M denotes the different staining parameter matrices, the rows of the parameter matrices denote H, eosin and DAB staining, respectively, the columns of the parameter matrices denote the RGB staining protocol parameter size, and y denotes the optical density of each pixel.
As a preferred technical solution, in step S2, the irregular pixel block performs kmean clustering to distinguish the image background, and the L layer in the Lab color space of the irregular pixel block is used as the object of kmean clustering, where a specific calculation formula is as follows:
wherein E represents the least square error, x represents each irregular pixel block, u represents the centroid, g represents the classified cluster, and kmeans clustering is completed when E is the smallest.
As a preferred technical solution, the image segmentation based on morphological features in step S3 includes the specific steps of:
s31: masking the background-removed image and the DAB dyeing channel to obtain a background-removed DAB dyeing channel image, and segmenting the image by adopting a watershed algorithm;
s32: and extracting the image characteristics segmented by the watershed algorithm, removing the area suspected of dust, and reserving the cell nucleus first area image L1 and the first image C1 to be processed.
As a preferred technical solution, the image features include a gray level average, a contrast, a compactness and a pixel area, the region suspected of being dust is set as an image region with a contrast less than 0.04 or a gray level average greater than 200, the cell nucleus first region image L1 is set as an image region with a pixel area of 40 × magnification less than 2000 and a density greater than 0.93, and the first image to be processed C1 is set as an image region with a pixel area of 40 × magnification greater than 2000.
As a preferred technical solution, in step S4, the local threshold bernsen segmentation and morphological feature segmentation are performed on the first image C1 to be processed, and the specific steps are as follows:
s41: setting an active window of the local threshold bernsen segmentation to be 77 x 77 pixels, obtaining a binary image of the first image C1 to be processed after the local threshold bernsen segmentation, masking the binary image with a DAB staining channel to obtain a cell nucleus area image of the first image C1 to be processed of the DAB staining channel, and performing watershed segmentation after open operation;
s42: and performing image segmentation based on morphological characteristics on the image subjected to the local threshold and watershed segmentation processing, extracting image characteristics, and reserving the cell nucleus second region image L2 and the second image C2 to be processed.
As a preferred technical solution, the cell nucleus second region image L2 is set as an image region with a pixel area of 40 × magnification smaller than 2000 and a density larger than 0.93, and the to-be-processed second image C2 is set as an image region with a pixel area of 40 × magnification larger than 2000.
As a preferred technical solution, in step S5, foreground labeling is performed on the second image C2 to be processed, and a watershed algorithm is used to segment a cell nucleus third region image L3, and the specific steps are as follows:
s51: the second image C2 to be processed is subjected to binarization processing and then is masked with a DAB staining channel to obtain a DAB staining cell segmentation area to be processed, and foreground marking is carried out by adopting opening and closing operation based on reconstruction and taking a local maximum value;
s52: and (4) superposing the image after the foreground mark and the to-be-processed DAB staining cell segmentation area in the step (S51), setting the foreground mark as a local minimum value of the center of the cell nucleus of the image, and performing image segmentation by using a watershed segmentation algorithm to obtain a third area image L3 of the cell nucleus.
Preferably, the method further includes a step of setting boundary lines for cell nuclei, where the boundary lines are set at the edges of the cell nucleus regions in the cell nucleus segmentation image obtained in step S6, and a plurality of boundary lines are connected to generate the boundary lines.
The invention also provides an immunohistochemical pathology image CD3 positive cell nucleus segmentation system, which comprises: the system comprises a dyeing channel separation module, a background removal module, an image preliminary segmentation module, a local threshold bernsen segmentation module and a watershed segmentation module;
the staining channel separation module is provided with a color deconvolution unit, and the color deconvolution unit is used for performing color deconvolution on the immunohistochemical pathological image to separate the staining channels;
the background removing module comprises a super-pixel segmentation unit and a kmeans clustering unit, wherein the super-pixel segmentation unit segments an image into irregular pixel blocks, and the kmeans clustering unit is used for distinguishing background images;
the image preliminary segmentation module is used for segmenting morphological characteristic images to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed;
the local threshold bernsen segmentation module is used for performing local threshold bernsen segmentation on the first image C1 to be processed to obtain a cell nucleus second region image L2 and a second image C2 to be processed;
the watershed segmentation module is used for carrying out foreground marking and watershed algorithm segmentation on the second image C2 to be processed to obtain a cell nucleus third region image L3,
and masking the first cell nucleus region image L1, the second cell nucleus region image L2 and the third cell nucleus region image L3 with an original immunohistochemical pathological image to obtain an immunohistochemical pathological image CD3 positive cell nucleus segmentation result image.
Compared with the prior art, the invention has the following advantages and beneficial effects:
(1) according to the method, when background is removed through Kmeans clustering, the image is divided into irregular pixel blocks through superpixel segmentation, the situation that parts with lighter staining in cells are mistakenly identified as background removal is avoided, and the accuracy of image processing is improved.
(2) According to the method, the local threshold bernsen is adopted to segment the image, the color of the image is darker or lighter due to the fact that dyeing unevenness exists possibly, interference of dyeing difference is eliminated by the local threshold bernsen segmented image, and the precision of image processing is improved.
(3) The method adopts the foreground mark and the watershed algorithm to segment the image, solves the problem that overlapped cells cannot be segmented when the watershed algorithm is used alone for segmentation, and achieves the technical effect of accurate segmentation.
Drawings
FIG. 1 is a schematic flow chart of the method for segmenting a nucleus positive for a immunohistochemical pathology image CD3 according to the present embodiment;
fig. 2 is a schematic diagram of a pathological image cell nucleus segmentation effect of the immunohistochemical pathological image CD3 positive cell nucleus segmentation method in this embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Examples
As shown in fig. 1, the present embodiment provides a method for segmenting a nucleus positive to an immunohistochemical pathology image CD3, comprising the following steps:
s1: performing color deconvolution on an original RGB (red, green and blue) coded immunohistochemical pathological image, separating two staining channels of hematoxylin (Haematoxylin, H) and Diaminobenzidine (3,3' -Diaminobenzidine, DAB), and only segmenting DAB stained CD3 positive cells;
in this embodiment, the color deconvolution algorithm calculates the effect of each color agent on the image based on the specific absorption of the RGB component light of the color agent used in the immunohistochemistry technology for the color information acquired by the RGB camera, and deconvolution refers to a process of calculating an unknown input, where the output is the CD3 staining map and the known input is H ═ 0.6500286,0.704031,0.2860126, and DAB ═ 0.26814753,0.57031375,0.77642715, so as to obtain an H staining part and a DAB staining part, and the formula is as follows:
C＝M-1[y]
wherein C is a separate H and DAB staining channel, M is a different staining parameter matrix, rows of the parameter matrix are respectively H, eosin (eosin) and DAB staining, eosin is only used as a background color in this embodiment, columns of the parameter matrix are RGB staining scheme parameter sizes, and y is an optical density of each pixel;
s2: superpixel segmentation and Kmeans clustering remove background:
in order to extract and segment the characteristics of a cell part, firstly removing a background, converting an original immunohistochemical pathological image from an RGB color space to an Lab color space according to the principle that the background is lightly stained, wherein a region with a larger L (brightness) layer is the background, and simultaneously, in order to avoid identifying a part with lighter staining in a cell as the background, dividing adjacent pixels with similar brightness, texture and color into irregular pixel blocks by using superpixel segmentation;
taking an L (brightness) layer in a Lab color space of an irregular pixel block as an object of kmeans clustering, calculating an average value at the position of an original image where the irregular pixel block is located, performing the kmeans clustering on an average value result, and then dividing the average value result into two types, so that the difference between points of the two types and the type to which the points belong is the minimum, taking the larger one of the L (brightness) layer as a background to be removed, and keeping the smaller cell part of the L (brightness) layer for next segmentation, wherein the calculation formula is as follows:
wherein E represents the minimum square error, x represents each irregular pixel block, u represents the centroid, g represents the classified cluster, and when E is the minimum, the kmeans clustering is completed;
s3: primarily segmenting morphological characteristics, namely segmenting by using basic characteristics of cells;
s31: carrying out mask processing on the image without the background and a DAB staining channel to obtain a DAB staining channel image without the background, and segmenting a cell image which is connected together but has an obvious boundary by using a watershed when cell staining exists, wherein the watershed algorithm is a method for taking the image as a topographic topological graph, the gray scale of a pixel is taken as the altitude of the point, water is overflowed upwards at each local minimum, and the watershed is formed at the junction of different basins and is taken as a part to be segmented;
s32: the watershed segmented image is extracted to obtain the characteristics such as gray level mean value of each area at 40X magnification (0.2520 microns/pixel), contrast, compactness, pixel area, etc. in this embodiment, the sum of the gray levels of the pixel regions is divided by the number of pixels to obtain a gray average, the number of pixels in each disconnected pixel region constitutes the pixel area, setting the gray scale as the contrast between 0 and 255, calculating the distance of each pixel in each disconnected cell nucleus area to obtain the compactness, removing the image area of which the contrast of suspected dust is less than 0.04 or the average gray scale value is more than 200, reserving the image area of which the pixel area of the 40X magnification (0.2520 microns/pixel) is less than 2000 and the density is more than 0.93 as a first cell nucleus area image L1 which is subjected to primary segmentation, and taking the rest part of which the pixel area is more than 2000 as a first image C1 to be processed for next segmentation;
s4: carrying out local threshold bernsen segmentation to remove dyeing difference interference;
s41: the first image C1 to be processed may cause darker or lighter color due to uneven dyeing, the local threshold segmentation is used to eliminate the dyeing difference interference, the local threshold segmentation moves the window at a magnification of 40X (0.2520 microns/pixel) to 77X 77 pixel size, namely, the system automatically determines the threshold value to binarize the image in the window of every 77X 77 pixel size, the local threshold segmentation is carried out to obtain the binarized image of the first image C1 to be processed and the DAB dyeing channel to carry out mask processing, the processed first image C1 cell nucleus region image of the DAB dyeing channel is obtained, the DAB dyeing picture of the rest region of the first image C1 image to be processed after the local threshold bernsen segmentation is reserved, the opening operation is carried out, namely, the operation of corroding and re-expanding the picture by using the 5 pixel size disc, the cell nucleus region of the segmentation is more smooth while the partial noise interference is removed, then carrying out watershed segmentation;
s42: after the dyeing difference interference is eliminated, performing morphological segmentation on the map subjected to the local threshold and watershed segmentation processing again, extracting the characteristics of pixel area, density, contrast, gray level average and the like of each independent region at 40X magnification (0.2520 micrometers/pixel), reserving the image with the pixel area of 40X magnification (0.2520 micrometers/pixel) being less than 2000 and the density being more than 0.93 as a cell nucleus second region image L2, and performing next segmentation on the rest part with the pixel area being more than 2000 as a second image C2 to be processed;
s5: the foreground marks watershed segmentation overlapping cells;
s51: through the segmentation of the steps, a plurality of overlapped cell nuclei remain in the part C2 of the second image to be processed, the part C2 of the second image to be processed is subjected to binarization processing and then is subjected to mask processing with a DAB staining channel, the obtained part is cells with a short distance, and the colors are connected together during staining, so that a DAB staining cell segmentation area to be processed is obtained, a foreground image of the cells is obtained by utilizing opening and closing operations based on reconstruction and taking a local maximum value as a foreground mark, the background part is removed more accurately, and then the overlapped cells are further segmented, wherein the reconstruction opening and closing operations in the embodiment can be realized by adopting a matlab function;
s52: marking foreground objects of the image, superposing the foreground objects with the to-be-processed DAB staining cell segmentation area in the step S51, and setting the foreground marks as a local minimum value for the center of the cell nucleus of the image, so that the watershed algorithm can segment the residual overlapped cells to obtain a third area image L3 of the cell nucleus;
s6: as shown in fig. 2, the first cell nucleus region image L1, the second cell nucleus region image L2, the third cell nucleus region image L3 and the original immunohistochemical pathology image are masked to obtain segmented cell nucleus images, and finally, contour lines are added to the cell nucleus edges to facilitate observation, so that boundary lines are generated for each of the segmented disconnected cell nucleus regions.
The embodiment also provides an immunohistochemical pathology image CD3 positive cell nucleus segmentation system, which comprises: the system comprises a dyeing channel separation module, a background removal module, an image preliminary segmentation module, a local threshold bernsen segmentation module and a watershed segmentation module;
in this embodiment, the staining channel separation module is provided with a color deconvolution unit, and the color deconvolution unit is configured to perform color deconvolution on the immunohistochemical pathological image to separate the staining channels; the background removing module comprises a super-pixel segmentation unit and a kmeans clustering unit, wherein the super-pixel segmentation unit segments the image into irregular pixel blocks, and the kmeans clustering unit is used for distinguishing background images; the image preliminary segmentation module is used for segmenting morphological characteristic images to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed; the local threshold bernsen segmentation module is used for performing local threshold bernsen segmentation on the first image C1 to be processed to obtain a cell nucleus second region image L2 and a second image C2 to be processed; the watershed segmentation module is used for conducting foreground marking on the second image C2 to be processed and segmenting the second image C2 into a cell nucleus third area image L3 through a watershed algorithm, and after the cell nucleus first area image L1, the cell nucleus second area image L2, the cell nucleus third area image L3 and an immunohistochemical pathological image original image are subjected to mask processing, an immunohistochemical pathological image CD3 positive cell nucleus segmentation result image is obtained.
In this embodiment, when processing pathological images, the cpu parallel processing function is turned on, each pathological image is independent of another pathological image, the image processing speed is high, and a save and check folder is established under the data source directory for detecting whether data has been processed while the progress of the previous processing continues.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (10)
1. An immunohistochemical pathology image CD3 positive cell nucleus segmentation method is characterized by comprising the following steps:
s1: performing color deconvolution on the immunohistochemical pathological image, and separating a staining channel;
s2: converting an original immunohistochemical pathological image from an RGB color space to a Lab color space, dividing the original immunohistochemical pathological image into irregular pixel blocks by adopting superpixels, carrying out kmeans clustering on the irregular pixel blocks to distinguish an image background, and removing the image background;
s3: performing image segmentation based on morphological characteristics to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed;
s4: performing local threshold bernsen segmentation and morphological feature segmentation on the first image C1 to be processed to obtain a segmented cell nucleus second region image L2 and a second image C2 to be processed;
s5: carrying out foreground marking on the second image C2 to be processed, and segmenting a cell nucleus third area image L3 by adopting a watershed algorithm;
s6: and masking the first cell nucleus region image L1, the second cell nucleus region image L2, the third cell nucleus region image L3 and the original immunohistochemical pathology image to obtain a cell nucleus segmentation image.
2. The method for segmenting the nucleus of the immunohistochemical pathological image CD3 positive according to claim 1, wherein the color deconvolution is performed on the immunohistochemical pathological image in step S1, which is calculated by the formula:
C＝M-1[y]
where C denotes the separate H and DAB staining channels, M denotes the different staining parameter matrices, the rows of the parameter matrices denote H, eosin and DAB staining, respectively, the columns of the parameter matrices denote the RGB staining protocol parameter size, and y denotes the optical density of each pixel.
3. The method for dividing a nucleus of an immunohistochemical pathology image CD3 positive cell according to claim 1, wherein the irregular pixel block performs a kmeans clustering to distinguish the image background in step S2, and the L layer in the Lab color space of the irregular pixel block is used as the object of the kmeans clustering, and the specific calculation formula is as follows:
wherein E represents the least square error, x represents each irregular pixel block, u represents the centroid, g represents the classified cluster, and kmeans clustering is completed when E is the smallest.
4. The method for segmenting the nucleus of the immunohistochemical pathology image CD3 positive according to claim 1, wherein the image segmentation based on the morphological characteristics in step S3 comprises the following steps:
s31: masking the background-removed image and the DAB dyeing channel to obtain a background-removed DAB dyeing channel image, and segmenting the image by adopting a watershed algorithm;
s32: and extracting the image characteristics segmented by the watershed algorithm, removing the area suspected of dust, and reserving the cell nucleus first area image L1 and the first image C1 to be processed.
5. The immunohistochemical pathology image CD3 positive cell nucleus segmentation method according to claim 4, wherein the image features include a gray level mean, a contrast, a compactness and a pixel area, the suspected dust area is set as an image area with a contrast less than 0.04 or a gray level mean greater than 200, the cell nucleus first area image L1 is set as an image area with a pixel area less than 2000 at 40X magnification and a density greater than 0.93, and the first image C1 to be processed is set as an image area with a pixel area greater than 2000 at 40X magnification.
6. The method for segmenting the nucleus of the immunohistochemical pathology image CD3 positive according to the claim 1, wherein the step S4 is to perform the local threshold bernsen segmentation and the morphological feature segmentation on the first image C1 to be processed, and the specific steps are as follows:
s41: setting an active window of the local threshold bernsen segmentation to be 77 x 77 pixels, obtaining a binary image of the first image C1 to be processed after the local threshold bernsen segmentation, masking the binary image with a DAB staining channel to obtain a cell nucleus area image of the first image C1 to be processed of the DAB staining channel, and performing watershed segmentation after open operation;
s42: and performing image segmentation based on morphological characteristics on the image subjected to the local threshold and watershed segmentation processing, extracting image characteristics, and reserving the cell nucleus second region image L2 and the second image C2 to be processed.
7. The immunohistochemical pathology image CD3 positive cell nucleus segmentation method according to claim 6, wherein the cell nucleus second region image L2 is set to an image region with 40X magnification pixel area less than 2000 and density greater than 0.93, and the second image to be processed C2 is set to an image region with 40X magnification pixel area greater than 2000.
8. The immunohistochemical pathology image CD3 positive cell nucleus segmentation method according to claim 1, wherein in step S5, the second image C2 to be processed is foreground-labeled, and a third region image L3 of the cell nucleus is segmented by using watershed algorithm, and the method comprises the following specific steps:
s51: the second image C2 to be processed is subjected to binarization processing and then is masked with a DAB staining channel to obtain a DAB staining cell segmentation area to be processed, and foreground marking is carried out by adopting opening and closing operation based on reconstruction and taking a local maximum value;
s52: and (4) superposing the image after the foreground mark and the to-be-processed DAB staining cell segmentation area in the step (S51), setting the foreground mark as a local minimum value of the center of the cell nucleus of the image, and performing image segmentation by using a watershed segmentation algorithm to obtain a third area image L3 of the cell nucleus.
9. The immunohistochemical pathology image CD3 positive cell nucleus segmentation method according to claim 1, further comprising a step of setting a boundary line for the cell nucleus, wherein the boundary line is set for the cell nucleus region edge in the cell nucleus segmentation image obtained in step S6, and a plurality of boundary lines are connected to generate the boundary line.
10. An immunohistochemical pathology image CD3 positive nucleus segmentation system, comprising: the system comprises a dyeing channel separation module, a background removal module, an image preliminary segmentation module, a local threshold bernsen segmentation module and a watershed segmentation module;
the staining channel separation module is provided with a color deconvolution unit, and the color deconvolution unit is used for performing color deconvolution on the immunohistochemical pathological image to separate the staining channels;
the background removing module comprises a super-pixel segmentation unit and a kmeans clustering unit, wherein the super-pixel segmentation unit segments an image into irregular pixel blocks, and the kmeans clustering unit is used for distinguishing background images;
the image preliminary segmentation module is used for segmenting morphological characteristic images to obtain a first cell nucleus region image L1 subjected to preliminary segmentation and a first image C1 to be processed;
the local threshold bernsen segmentation module is used for performing local threshold bernsen segmentation on the first image C1 to be processed to obtain a cell nucleus second region image L2 and a second image C2 to be processed;
the watershed segmentation module is used for carrying out foreground marking and watershed algorithm segmentation on the second image C2 to be processed to obtain a cell nucleus third region image L3,
and masking the first cell nucleus region image L1, the second cell nucleus region image L2 and the third cell nucleus region image L3 with an original immunohistochemical pathological image to obtain an immunohistochemical pathological image CD3 positive cell nucleus segmentation result image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910517113.1A CN110415255B (en) | 2019-06-14 | 2019-06-14 | Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910517113.1A CN110415255B (en) | 2019-06-14 | 2019-06-14 | Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110415255A CN110415255A (en) | 2019-11-05 |
| CN110415255B true CN110415255B (en) | 2021-08-20 |
Family
ID=68359139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910517113.1A Active CN110415255B (en) | 2019-06-14 | 2019-06-14 | Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415255B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111275727B (en) * | 2020-01-19 | 2021-09-07 | 北方工业大学 | Cell segmentation method, cell segmentation apparatus, and computer-readable storage medium |
| CN111402267B (en) * | 2020-03-13 | 2023-06-16 | 中山大学孙逸仙纪念医院 | Segmentation method, device and terminal of epithelial cell nuclei in prostate cancer pathological image |
| CN111583185A (en) * | 2020-04-14 | 2020-08-25 | 山东省千佛山医院 | Ki67 cell nucleus counting method and system based on pathological immunohistochemistry |
| CN112614142B (en) * | 2020-12-25 | 2023-05-30 | 华侨大学 | Cell weak label manufacturing method and system based on multichannel image fusion |
| CN113223097A (en) * | 2021-04-29 | 2021-08-06 | 武汉工程大学 | Image preprocessing method for improving density counting precision |
| CN113436206B (en) * | 2021-06-17 | 2022-03-15 | 易普森智慧健康科技（深圳）有限公司 | Pathological tissue section scanning area positioning method based on cluster segmentation |
| CN116645390B (en) * | 2023-07-27 | 2023-10-03 | 吉林省星博医疗器械有限公司 | Fluorescent image cell rapid segmentation method and system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104484877A (en) * | 2014-12-12 | 2015-04-01 | 山东大学 | AML cell segmentation method based on Meanshift cluster and morphological operations |
| CN105741266A (en) * | 2016-01-22 | 2016-07-06 | 北京航空航天大学 | Pathological image cell nucleus quick location method |
-
2019
- 2019-06-14 CN CN201910517113.1A patent/CN110415255B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104484877A (en) * | 2014-12-12 | 2015-04-01 | 山东大学 | AML cell segmentation method based on Meanshift cluster and morphological operations |
| CN105741266A (en) * | 2016-01-22 | 2016-07-06 | 北京航空航天大学 | Pathological image cell nucleus quick location method |
Non-Patent Citations (1)
| Title |
|---|
| "Automated cancer stem cell recognition in H and E stained tissue using convolutional neural networks and color deconvolution";Wolfgang Aichinger等;《PROCEEDINGS OF SPIE》;20171231;第1-7页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110415255A (en) | 2019-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110415255B (en) | Immunohistochemical pathological image CD3 positive cell nucleus segmentation method and system | |
| CN110400318B (en) | Immunohistochemical pathological image CD4 positive cell nucleus segmentation method and system | |
| CN109886896B (en) | Blue license plate segmentation and correction method | |
| Malpica et al. | Applying watershed algorithms to the segmentation of clustered nuclei | |
| CN111145209B (en) | Medical image segmentation method, device, equipment and storage medium | |
| JP4948647B2 (en) | Urine particle image region segmentation method and apparatus | |
| CN109636732A (en) | A kind of empty restorative procedure and image processing apparatus of depth image | |
| US20200193139A1 (en) | Systems and methods for automated cell segmentation and labeling in immunofluorescence microscopy | |
| CN110648330B (en) | Defect detection method for camera glass | |
| AU2019273339B2 (en) | Saturation clustering-based method for positioning bone marrow white blood cells | |
| CN113326846B (en) | Rapid bridge apparent disease detection method based on machine vision | |
| CN116091421A (en) | Method for automatically dividing and calculating area of blastomere image of in-vitro fertilized embryo | |
| CN109378279A (en) | Wafer detection method and wafer detection system | |
| CN104021566A (en) | GrabCut algorithm-based automatic segmentation method of tongue diagnosis image | |
| CN113554638A (en) | Method and system for establishing chip surface defect detection model | |
| KR100903816B1 (en) | System and human face detection system and method in an image using fuzzy color information and multi-neural network | |
| CN113724235B (en) | Semi-automatic Ki67/ER/PR negative and positive cell counting system and method under condition of changing environment under mirror | |
| CN116596899A (en) | Method, device, terminal and medium for identifying circulating tumor cells based on fluorescence image | |
| Lézoray et al. | Histogram and watershed based segmentation of color images | |
| CN111695374A (en) | Method, system, medium, and apparatus for segmenting zebra crossing region in monitoring view | |
| CN112651368B (en) | DNA ploidy analysis method for large-size cell microscopic image | |
| CN115423765A (en) | Grain defect quantitative segmentation method based on template image | |
| SKiran | A framework in shadow detection and compensation of images | |
| CN112949527B (en) | Method and system for extracting road surface area of aerial road image | |
| CN112819795B (en) | Cell number and area statistical method based on multicolor fluorescence picture |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |