CN113295692B - Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium - Google Patents
Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium Download PDFInfo
- Publication number
- CN113295692B CN113295692B CN202110570957.XA CN202110570957A CN113295692B CN 113295692 B CN113295692 B CN 113295692B CN 202110570957 A CN202110570957 A CN 202110570957A CN 113295692 B CN113295692 B CN 113295692B
- Authority
- CN
- China
- Prior art keywords
- cells
- abnormal
- cell
- image
- threshold
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 55
- 108091093105 Nuclear DNA Proteins 0.000 title claims abstract description 9
- 230000002159 abnormal effect Effects 0.000 claims abstract description 133
- 230000035755 proliferation Effects 0.000 claims abstract description 69
- 238000000034 method Methods 0.000 claims abstract description 24
- 239000012472 biological sample Substances 0.000 claims abstract description 17
- 230000002062 proliferating effect Effects 0.000 claims abstract description 10
- 210000004027 cell Anatomy 0.000 claims description 230
- 108020004414 DNA Proteins 0.000 claims description 41
- 238000013528 artificial neural network Methods 0.000 claims description 31
- 210000003855 cell nucleus Anatomy 0.000 claims description 21
- 239000002131 composite material Substances 0.000 claims description 14
- 230000001086 cytosolic effect Effects 0.000 claims description 11
- 238000011176 pooling Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000007477 logistic regression Methods 0.000 claims description 5
- 230000009977 dual effect Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 3
- 238000000338 in vitro Methods 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 150000001875 compounds Chemical class 0.000 claims 3
- 238000003556 assay Methods 0.000 claims 2
- 230000035945 sensitivity Effects 0.000 abstract description 13
- 230000000877 morphologic effect Effects 0.000 abstract description 2
- 230000005856 abnormality Effects 0.000 abstract 1
- 239000010410 layer Substances 0.000 description 41
- 238000001514 detection method Methods 0.000 description 16
- 230000006870 function Effects 0.000 description 13
- 238000004891 communication Methods 0.000 description 8
- 238000003745 diagnosis Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 6
- 239000000523 sample Substances 0.000 description 5
- 241000193830 Bacillus <bacterium> Species 0.000 description 4
- 210000001519 tissue Anatomy 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 244000005700 microbiome Species 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000001413 cellular effect Effects 0.000 description 2
- 238000013329 compounding Methods 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 239000002346 layers by function Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010186 staining Methods 0.000 description 2
- 238000007447 staining method Methods 0.000 description 2
- FFRBMBIXVSCUFS-UHFFFAOYSA-N 2,4-dinitro-1-naphthol Chemical compound C1=CC=C2C(O)=C([N+]([O-])=O)C=C([N+]([O-])=O)C2=C1 FFRBMBIXVSCUFS-UHFFFAOYSA-N 0.000 description 1
- 241000894006 Bacteria Species 0.000 description 1
- 241000606161 Chlamydia Species 0.000 description 1
- 208000035473 Communicable disease Diseases 0.000 description 1
- 241000233866 Fungi Species 0.000 description 1
- 241000238631 Hexapoda Species 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 241000204031 Mycoplasma Species 0.000 description 1
- 238000005481 NMR spectroscopy Methods 0.000 description 1
- 102000029797 Prion Human genes 0.000 description 1
- 108091000054 Prion Proteins 0.000 description 1
- 241000606701 Rickettsia Species 0.000 description 1
- 241000589970 Spirochaetales Species 0.000 description 1
- 241000700605 Viruses Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000001574 biopsy Methods 0.000 description 1
- 238000004195 computer-aided diagnosis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 235000013601 eggs Nutrition 0.000 description 1
- 210000003527 eukaryotic cell Anatomy 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- -1 gold amine Chemical class 0.000 description 1
- 244000000013 helminth Species 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 230000007762 localization of cell Effects 0.000 description 1
- 244000000010 microbial pathogen Species 0.000 description 1
- 238000009206 nuclear medicine Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 244000045947 parasite Species 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 210000001236 prokaryotic cell Anatomy 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
- 230000009469 supplementation Effects 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
- 210000004916 vomit Anatomy 0.000 description 1
- 230000008673 vomiting Effects 0.000 description 1
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
- G01N21/84—Systems specially adapted for particular applications
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a cell analysis method based on a cell nuclear DNA and TBS double analysis method, a computer device and a storage medium, wherein the cell analysis method comprises the following steps: s1: obtaining an image of an ex vivo biological sample; s2: judging the first appearance form, and dividing the cells into four types: normal cells, suspected abnormal cells, highly suspected abnormal cells, and positive abnormal proliferating cells; s3: performing second appearance morphological judgment on the positive abnormal proliferation cells or the highly suspicious abnormal proliferation cells obtained in the step S2, and classifying the cells into three types: suspicious proliferating cells, highly suspicious abnormal proliferating cells and positive abnormal proliferating cells; step S4: and (3) obtaining a comprehensive cell analysis result by utilizing the analysis results of the step S2 and the step S3, and displaying the abnormality indexes of the cells in sequence from high to low when outputting the result. The method can give consideration to the sensitivity and the specificity in cell analysis, and effectively improves the analysis efficiency of staff.
Description
Technical Field
The invention relates to a cell analysis method, in particular to a cell analysis method, computer equipment and storage medium based on a nuclear DNA and TBS double analysis method.
Background
Many lesions in organisms cause normal cellular changes, and thus it is often clinically necessary to determine whether a tissue or organism is diseased by microscopic examination of a biological sample.
With the development of AI technology, the cell analysis work is increasingly performed by using AI technology, but the existing analysis work of biological samples by using AI technology generally has the following problems: the sample analysis error is large, and false detection and missing detection often occur; the sample data volume is big, and sample analysis work load is big, and is inconvenient for medical staff to carry out the consulting of biological sample image, and the analysis efficiency to the sample is low.
Disclosure of Invention
In order to solve the technical problems, the invention provides a cell analysis method, computer equipment and a storage medium of a cell nuclear DNA and TBS double analysis method.
The cell analysis method based on the dual analysis method of the cell nuclear DNA and the TBS comprises a first appearance form judgment step and a second appearance form judgment step, wherein the cells are divided into four types through the first appearance form judgment step: normal cells, suspected abnormal cells, highly suspected abnormal cells, and positive abnormal proliferating cells; the second appearance form judging step is to calculate and analyze the nuclear plasma ratio of the positive abnormal proliferation cells or the highly suspicious abnormal proliferation cells obtained in the first appearance form judging step to realize cell analysis, define the nuclear plasma ratio as a DNA parameter 2, respectively preset a threshold D and a threshold E for the DNA parameter 2, and divide the DNA parameter 2 into three numerical ranges, thereby dividing the cells to be analyzed into three types: ① DNA parameter 2 is less than or equal to threshold value D, and is suspicious proliferation cells; ② The threshold value D is less than the DNA parameter 2 and is less than or equal to the threshold value E, and is a highly suspicious abnormal proliferation cell; ③ DNA parameter 2 > threshold E, positive abnormal proliferation cells; judging the abnormal proliferation cells judged to be positive by the first appearance form judging step and the second appearance form judging step as first positive abnormal cells; the first appearance form judging step judges positive abnormal proliferation cells, and the second appearance form judging step judges highly suspicious abnormal proliferation cells to be second positive abnormal cells; the first appearance form judging step judges the cells to be highly suspicious abnormal cells, the second appearance form judging step judges the cells to be positive abnormal proliferation cells to be third positive abnormal cells; judging the cell which is judged to be highly suspicious by the first appearance form judging step and the cell which is judged to be highly suspicious by the second appearance form judging step as the fourth positive abnormal cell; the first appearance form judging step judges positive abnormal proliferation cells or highly suspicious abnormal cells, and the second appearance form judging step judges suspicious proliferation cells to be fifth positive abnormal cells; and (3) judging the cells which are judged to be suspicious abnormal proliferation by the first appearance form judging step, and judging the cells which are judged to be suspicious proliferation or positive abnormal proliferation by the second appearance form judging step to be fifth suspicious cells.
Further, the first appearance form judging step is implemented by acquiring a DNA parameter 1 of the size of the cell nucleus area, defining the DNA parameter 1 as a DNA parameter 1, respectively presetting a threshold a, a threshold B and a threshold C for the DNA parameter 1, and dividing the parameter 1 into four numerical ranges, thereby dividing the cells to be analyzed into four types: ① Abnormal cells, parameter 1 is less than or equal to threshold A; ② Suspicious abnormal cells, wherein the threshold A is less than the parameter 1 and less than or equal to the threshold B; ③ Highly suspicious abnormal cells, wherein the threshold B is less than the parameter 1 and less than or equal to the threshold C; ④ Positive abnormal proliferation cells, wherein the parameter 1 is larger than the threshold C, and the calculation and analysis of the parameter 1 are realized through a pre-trained neural network;
the neural network comprises an input layer, an implicit layer and an output layer, wherein the output layer of the neural network outputs scores corresponding to the isolated biological samples, and the highest score is used as an analysis index of the first appearance form judging step;
The neural network comprises a first neural network module, a second neural network module and a classifier, wherein the first neural network module is a Darknet-53 network, the second neural network module comprises 5 convolution layers, a ReLU layer, a pooling layer, a BN+ LeakyReLU layer and a convolution layer which alternately appear, and the classifier adopts an LR logistic regression classifier.
Further, before the first appearance form judging step, the method further comprises the steps of acquiring and preprocessing an image of the cell to be analyzed, and specifically comprises the following steps:
(1) Amplifying cells under a microscope by different times, automatically focusing by the microscope, and acquiring a target picture to be analyzed; dividing each picture of a group of pictures shot by a cell under the same view of a microscope into three channels R (red), G (green) and B (blue), compositing any two channels to obtain a first composite image, respectively calculating variances of the pictures relative to the first composite image to obtain variances Variance_value, respectively comparing the variances of each picture, and selecting the picture with the largest Variance value relative to the first composite image as a first target picture of the group of pictures;
(2) And fusing a plurality of first target pictures of different levels in the same visual field to obtain an image of the cell to be analyzed.
Further, the step (2) is implemented according to the following steps: defining a first target picture as A (X, y), defining the last picture as A_Pre (X, y), defining the Next picture as A_Next (X, y), defining the fused image as Fuse_image (X, y), setting the pixel values of the three images at points (X, y) as A (X, y), A_Pre (X, y), and A_Next (X, y), comparing the three images by using 3X3 regions, calculating the sum of the pixel values of the 3X3 regions of the three first target pictures to obtain A_add_value, A_Pre_add_value and A_Ne_add_value, respectively, giving the pixel values of the photo with the largest value at the points (X, y) to the fused image as the fused image.
Furthermore, the cell nucleus of the image of the cell to be analyzed is positioned and identified, and the area of the cell nucleus is calculated.
Further, the second appearance form judgment step includes a step of calculating the size of the cytoplasmic area of the positive or highly suspected abnormal proliferation cells located by the first appearance form judgment step.
Further, the calculation of the cytoplasmic area is to obtain a red channel image R (x, y), a green channel image G (x, y) and a blue channel image B (x, y) by separating the positioned and segmented positive abnormal proliferation cells or highly suspicious abnormal proliferation cells, obtain a second composite image by using the images of any two channels, and calculate the Mean value mean_value of the second composite image; each pixel of the second composite image is differenced with mean_value 0.8, the new pixel value is equal to the value-mean_value 0.8 of each pixel of the second composite image, and when the new pixel value is larger than 0, the new pixel value is assigned to 255; when the new pixel value is smaller than 0, the new pixel value is assigned to 0, a binarized image Threshold (x, y) is obtained according to the new pixel value, a convex hull is obtained for the Threshold (x, y) image, the area of the convex hull is obtained, and finally the largest area is found out to be the cytoplasmic area of the cell.
Furthermore, the second appearance form judging step further comprises dividing the DNA parameter 1 of the positive abnormal proliferation cells or the highly suspicious abnormal proliferation cells obtained by the first appearance form judging step by the cytoplasmic area of the corresponding cells to obtain a nuclear plasma ratio, and calculating to obtain the DNA parameter 2.
The invention also provides a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of the aforementioned method.
The invention also provides a storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of the aforementioned method.
The invention has the following beneficial effects:
① According to the invention, firstly, through a first appearance form judging step, the size of the cell nucleus area is analyzed, so that abnormal cells can be obtained through rapid analysis, then, through a second appearance form judging step, highly suspicious abnormal cells and positive abnormal proliferation cells in the abnormal cells are further analyzed, and the cells to be analyzed are further separated into suspicious proliferation cells, highly suspicious abnormal proliferation cells and positive abnormal proliferation cells through DNA parameters 2, after the abnormal cells are separated through the first appearance form judging step, the abnormal cells are further analyzed through a nuclear plasma to TBS relationship, so that diagnosis and confirmation are further carried out on the abnormal cells, the diagnosis specificity of the abnormal cells is further improved, and the analysis and the mutual complementation of two methodologies are combined, so that the sensitivity and the specificity of the falling cytology are provided.
② By using the method, the cell nuclear area is changed from large to small, and the cell nuclear plasma is positive to suspicious cells analyzed by TBS, so that the corresponding fields of view are arranged from front to back, an analyst can conveniently and quickly check the field of view of abnormal cells and locate the abnormal cells, the analysis efficiency of the cells is improved, and the burden of the staff is reduced.
③ The method can also be used for pointedly adjusting the threshold value (such as a threshold value A or a threshold value B or a threshold value C) and the nuclear plasma ratio TBS analysis threshold value (such as a threshold value D or a threshold value E) set by the DNA parameter 1 of the area of the cell nucleuses of different types of the cell nucleuses according to the characteristics of the cell nucleuses of different types of the cell nucleuses, so that the cell nucleuses of different types of the cell nucleuses are pointedly analyzed, and the detection sensitivity and the detection specificity of different ranges are provided.
④ The method of the invention can find a researched balance point on the specificity and sensitivity of detection, thereby improving the sensitivity of analysis and the specificity of analysis diagnosis, solving the problems of sensitivity and specificity of different types of abscissas, ensuring the negative discharge rate with high specificity and high sensitivity, providing positive results with high accuracy, and effectively reducing the workload of cell analysis.
Detailed Description
The technical scheme of the invention is described below with reference to the specific embodiments.
In the present invention, "ex vivo" refers to the separation of a target organism from an animal or human body to an in vitro environment, whether by surgery or sampling or vomit, etc.; "biological sample" refers broadly to a pathogenic microorganism or tissue cell, including but not limited to: non-cellular microorganisms such as: mainly comprises viruses, prion and the like; prokaryotic cell type microorganisms, such as: bacteria, rickettsia, chlamydia, mycoplasma, spirochete, etc.; eukaryotic cell type microorganisms such as: including fungi, protozoa, parasites (protozoa, helminths, medical insects), and eggs thereof. Tissue cells include, but are not limited to: tissue sections and various cells.
The TBS in the present invention is an abbreviation for descriptive diagnosis (TheBethesdasystem). The dual analysis of nuclear DNA and TBS of the present invention refers to the analysis of cells by combining the analysis of nuclear size with TBS of the cells.
In the present invention, the image may be acquired by various medical image acquisition apparatuses. For example, for detection of various infectious diseases, microscopic images of biological samples may be obtained by microscopic imaging methods. The image may be obtained by means of an ultrasound device, an X-ray device, a nuclear magnetic resonance device, a nuclear medicine device, a medical optical device, a thermal imaging device, or the like.
In the embodiment of the invention, the image can be a two-dimensional image or a three-dimensional image. The image may be a gray scale image or a binarized image, or may be a color image.
The image of the present invention may be an image of an isolated biological sample obtained by observation with a biological microscope.
Embodiment one:
the cell analysis method based on the dual analysis method of the cell nuclear DNA and the TBS comprises the following steps:
S1: obtaining an image of an ex vivo biological sample;
S2: judging the first appearance form, and dividing the cells into four types: normal cells, suspected abnormal cells, highly suspected abnormal cells, and positive abnormal proliferating cells;
S3: performing second appearance morphological judgment on the positive abnormal proliferation cells or the highly suspicious abnormal proliferation cells obtained in the step S2, and classifying the cells into three types: suspicious proliferating cells, highly suspicious abnormal proliferating cells and positive abnormal proliferating cells;
Step S4: and (3) obtaining a comprehensive cell analysis result by using the analysis results of the steps S2 and S3: determining that the cells determined to be positive abnormal proliferation in both the step S2 and the step S3 are first positive abnormal cells; the positive abnormal proliferation cells are judged in the step S2, and the second positive abnormal cells are judged in the step S3; the highly suspicious abnormal cells determined in the step S2 are determined to be positive abnormal proliferation cells determined in the step S3 to be third positive abnormal cells; determining the cells determined to be highly suspicious abnormal cells in the step S2, and determining the cells determined to be highly suspicious abnormal proliferation cells in the step S3 to be fourth positive abnormal cells; the cells judged to be positive abnormal proliferation cells or highly suspicious abnormal cells in the step S2 are judged to be fifth positive abnormal cells in the step S3; and (3) judging the cells judged to be suspicious abnormal proliferation cells in the step S2, and judging the cells judged to be suspicious proliferation cells or positive abnormal proliferation cells in the step S3 to be fifth suspicious cells. The display sequence of each cell is sequentially from front to back, namely a first positive abnormal cell, a second positive abnormal cell, a third positive abnormal cell, a fourth positive abnormal cell, a fifth positive abnormal cell and a fifth suspicious cell, so that the cell image with the highest comprehensive abnormal risk coefficient is displayed in front of a worker at first, and the worker can analyze the cells conveniently.
The image in step S1 may be an image of the original undyed biological sample or an image of the dyed biological sample.
In the present invention, the staining method for biological samples may be any staining method in the prior art, such as, for example, wilt-ni acid-fast staining (in which tubercle bacillus cells are stained red under a microscope, non-tubercle bacillus is stained blue), gold amine O fluorescent acid-fast staining (in which tubercle bacillus is stained golden yellow under a microscope, and other non-tubercle bacillus is colorless or black background), so as to have better identifiability in an image.
The resolution of the image of the biological sample may be suitably processed as desired, for example, the resolution may be 416 x 416 or 256 x 256, etc.
The image of the ex-vivo biological sample of step S1 of the present embodiment is obtained by the steps of: dividing each picture of a group of pictures taken by an isolated biological sample under the same view of a microscope into three channels of R (red), G (green) and B (blue), obtaining a new image B_G by using B-G, respectively carrying out Variance calculation on each picture relative to the image B_G to obtain a Variance value, respectively comparing the Variance value of each picture, and selecting the picture with the largest Variance value relative to the image B_G as a first target picture of the group of pictures. The cell image to be analyzed is obtained by fusing a plurality of first target pictures with different levels in the same visual field, and is obtained according to the following steps: defining a first target picture as A (X, y), defining the last picture as A_Pre (X, y), defining the Next picture as A_Next (X, y), defining the fused image as Fuse_image (X, y), setting the pixel values of the three images at points (X, y) as A (X, y), A_Pre (X, y), value and A_Next (X, y), comparing the three images by using a 3X3 region, calculating the sum of the pixel values of the 3X3 region of the three first target pictures to obtain A_add_value, A_Pre_add_value and A_Ne_add_value, respectively, assigning the pixel values of the photo with the largest value at the points (X, y) to the fused image, namely, the fused image is obtained by traversing the fused image. The cell image to be analyzed according to the present invention may be obtained by a variety of different magnification factors, including but not limited to: 4-fold, 10-fold, 20-fold, 40-fold and 100-fold.
Step S2 of the present embodiment includes: the image obtained in step S1 is input into a pre-trained neural network comprising an input layer, an hidden layer and an output layer.
The neural network of the present invention may be a connection in one or more functional layers of the prior art, convolution, pooling, full connection, residual, excitation, regularization, tensor stitching, and the functions to be performed by each functional layer are known to those skilled in the art. The convolution layer may be used to perform a convolution operation to extract feature information of an input image (e.g., 227 x 227 in size) to obtain a feature map featuremap (e.g., 13 x 13 in size); the pooling layer may perform pooling operations on the input image, such as a maximum value combining (max-pooling) method, a mean-pooling (mean-pooling) method, and the like; the activation layer introduces non-linear factors through the activation function, such as a sexual modification unit (ReLU, leak-ReLU, P-ReLU, R-ReLU) function, a Sigmoid function, or a hyperbolic tangent function (tanh function), etc. The full connection layer is used for converting the feature map of the convolution output into a one-dimensional vector. The loss function is used for evaluating the inconsistency degree of the predicted value f (x) and the true value Y in the training of the neural network, and can be, for example, a log-log loss function, a square loss function, an exponential loss function, a range loss function and the like.
Depending on the specific function to be performed, various types of neural networks may be used, for example, deep convolutional neural networks CNN, such as Lenet, alexnet, VGG, may be employed to extract features in the image and output classification results of physical or pathological properties of the image through a softmax, svm classifier or the like.
The pretrained neural network of the present embodiment includes a first neural network module, a second neural network module, and a classifier, where the first neural network module is a Darknet-53 network, the second neural network module includes 5 convolution layers, a ReLU layer, a pooling layer, a bn+ LeakyReLU layer, and a convolution layer that alternately occur, and the classifier adopts an LR logistic regression classifier.
The Darknet-53 network obtains three feature images with different sizes corresponding to each image by adopting a multi-scale fusion method, performs dimension clustering on the marked data set by adopting a K-means clustering algorithm to obtain a plurality of prior frames (anchorbox) with different sizes, performs bbox prediction by adopting the K-means clustering, and performs a target scoring (finding out the part with the highest score, namely the final prediction category) on the part surrounded by the anchor.
The first neural network module comprises a convolution layer and five residual error modules which are sequentially connected, the first residual error module comprises a convolution layer and a residual error unit, the second residual error module comprises a convolution layer and two residual error units, the third residual error module comprises a convolution layer and eight residual error units, the fourth residual error module comprises a convolution layer and eight residual error units, the fifth residual error module comprises a convolution layer and four residual error units, each residual error unit comprises two convolution layers and a residual error connecting layer, and the five residual error modules are connected in a ResNet-jump layer connecting mode. The output ends of the fourth residual error module and the fifth residual error module are input to the same first tensor splicing layer concat, and the output end of the third residual error module is input to another second tensor splicing layer concat; the output end of the fourth residual error module obtains a second characteristic diagram after the output of the second tensor splicing layer and the output of the fifth residual error module are subjected to up-sampling processing and combined to output convolution processing (Conv 2D); the output end of the third residual error module outputs convolution processing after the outputs of the tensor splicing layer and the first tensor splicing layer are combined through up-sampling processing
(Conv 2D) a third profile is obtained.
The second neural network module includes 5 alternately occurring convolutional layers Conv, reLU layer, pooling layer Pooling, DBL layer, and convolutional layer Conv.
The classifier adopts an LR logistic regression classifier, and uses logistic regression to score the content surrounded by each anchor of the first neural network module and the feature map output by the second neural network module (objectnessscore) in a targeting manner, and the highest score is the final prediction category.
The invention positions the cell nucleus through yolov algorithm and divides the cell nucleus, the separated cell nucleus image is separated to obtain red channel image R (x, y), green channel image G (x, y), blue channel image B (x, y), and the first composite image is obtained by compounding any two channel images, the invention is used for compounding blue channel image B (x, y) and green channel image G (x, y) to obtain new image-first B_G (x, y), and the average value of the first B_G (x, y) image is obtained to obtain first mean_value; each pixel of the first B_G (x, y) is differenced from the first mean_value, the new pixel value is equal to each pixel value of the first B_G (x, y) -the first mean_value, 255 is assigned to the new pixel value when the new pixel value is larger than 0,0 is assigned to the new pixel value when the new pixel value is smaller than 0, thereby obtaining a first B_G_threshold (x, y) of the binarized image, the convex hull is obtained from the first B_G_threshold (x, y) image, the convex hull area is obtained, and finally the largest area is found, namely the cell nucleus area, thereby obtaining the DNA parameter 1.
DNA parameter 1 was analyzed by step S2 and cells were divided into four categories: ① The DNA parameter 1 is less than or equal to the threshold A and is a normal cell; ② The threshold A is less than DNA parameter 1 and less than or equal to threshold B, and is suspicious abnormal cells; ③ The threshold B is less than DNA parameter 1 and less than or equal to threshold C, and is a highly suspicious abnormal cell; ④ DNA parameter 1 > threshold C is positive abnormal proliferation cell.
As a preferred implementation of the present embodiment, step S2 further includes a step of forming the pre-trained neural network by training. Sample images for training are classified into positive and suspicious two types; the output layer of the neural network is configured to classify scores within a preset scoring range, samples with scores more than or equal to 95% as the optimal solution are considered positive, samples with scores less than or equal to 75% as the optimal solution are considered highly suspicious, and samples with scores less than or equal to 50% as the optimal solution are considered suspicious; samples with a score of < 50% were considered negative.
And the scoring is more than or equal to 80% of the optimal solution, the optimal solution is more than or equal to 60% and less than 80% of the optimal solution is highly suspicious, and the scoring is more than or equal to 35% and less than 55% of the optimal solution is suspicious.
Therefore, through the training mode, the image of the biological sample can be subjected to four-class by using the detection method, and compared with the traditional AI detection method for cell analysis, the AI detection method can only output positive results, and suspicious cells cannot be further analyzed, so that the cell analysis method has higher accuracy, is convenient for doctors to carry out cell analysis, and improves the working efficiency.
In step S3 of the present embodiment, this is achieved by further analyzing the positive or highly suspected abnormal proliferation cells located and identified in the first appearance morphology determining step.
In step S3, the calculation of the DNA parameter 2 is performed by performing cell localization and independent segmentation on the positive abnormal proliferation cells or the highly suspected abnormal proliferation cells obtained in step S2, and performing cell analysis by calculating the nuclear-plasma ratio thereof, defining the nuclear-plasma ratio as the DNA parameter 2, respectively presetting a threshold D and a threshold E for the DNA parameter 2, and dividing the DNA parameter 2 into three numerical ranges, thereby dividing the cells to be analyzed into three categories: ① DNA parameter 2 is less than or equal to threshold value D, and is suspicious proliferation cells; ② The threshold value D is less than the DNA parameter 2 and is less than or equal to the threshold value E, and is a highly suspicious abnormal proliferation cell; ③ DNA parameter 2 > threshold E, positive abnormal proliferation cells. As a preferred implementation of the present embodiment, the threshold D may be set to 1, the threshold E may be set to 1.5, and the threshold D and the threshold E may be adjusted as needed.
The calculation of the cytoplasmic area is to obtain a red channel image R (x, y), a green channel image G (x, y) and a blue channel image B (x, y) by separating the positioned and segmented positive abnormal proliferation cells or highly suspicious abnormal proliferation cells, and obtain a third composite image by compositing any two channels, wherein the embodiment takes the compositing of the B (x, y) channel and the G (x, y) channel images as an example to explain, and after compositing to obtain a second B_G (x, y) image, a second mean_value of the second mean_value is obtained; each pixel of the second b_g (x, y) is differenced from the second mean_value by 0.8, the new pixel value is equal to each pixel of the second b_g (x, y) -the second mean_value, and when the new pixel value is greater than 0, the new pixel value is assigned 255; when the new pixel value is less than 0, the value is assigned 0. And obtaining a second B_G_threshold (x, y) of the binarized image, obtaining a convex hull of the second B_G_threshold (x, y) image, obtaining the area of the convex hull, and finally finding out the maximum area, namely the cytoplasmic area of the cell.
In step S3, the numerical value of the size of the nuclear area of the positive abnormal proliferation cells or the highly suspected abnormal proliferation cells calculated in step S2 is divided by the cytoplasmic area of the corresponding cells to obtain a nuclear-plasma ratio, and the DNA parameter 2 is calculated.
The invention has the following beneficial effects:
① Firstly, through a first appearance form judging step, analyzing a DNA parameter 1 (cell nucleus area size), and dividing cells to be analyzed into normal cells, suspicious abnormal cells, highly suspicious abnormal cells and positive abnormal proliferation cells; the method comprises the steps of judging the first appearance form, analyzing the highly suspicious abnormal cells and the positive abnormal proliferation cells, dividing the cells to be analyzed into the suspicious proliferation cells, the highly suspicious abnormal proliferation cells and the positive abnormal proliferation cells through the DNA parameter 2, separating the abnormal cells through the first appearance form judging step, and further analyzing the abnormal cells by adopting a relation of nuclear plasma to TBS (tunnel boring system), so that diagnosis and confirmation of the abnormal cells are further carried out, the specificity of diagnosis of the abnormal cells is further improved, and the analysis and the mutual supplementation of two methodologies are combined, so that the sensitivity and the specificity of cell analysis are ensured.
② By using the method, the corresponding fields of view are arranged from front to back by the sequence from large to small nuclear area and positive nuclear plasma analysis to suspicious cells compared with TBS, so that an analyst can conveniently and quickly find the field of view of abnormal cells and quickly review the field of view, and the analysis efficiency of the abnormal cells is improved.
③ The method can also be used for pointedly adjusting the threshold value (such as a threshold value A or a threshold value B or a threshold value C) and the nuclear plasma ratio TBS analysis threshold value (such as a threshold value D or a threshold value E) set by the DNA parameter 1 of the area of the cell nucleuses of different types of the exfoliated cytology cells according to the characteristics of the cell nucleuses of different types of the exfoliated cytology cells, thereby pointedly analyzing the cell nucleuses of different types of the exfoliated cytology cells, providing detection sensitivity and detection specificity in different ranges and providing an effective research method for analyzing the isolated cells of different types of the large-scale artificial intelligence.
④ The method of the invention can find a researched balance point on the specificity and sensitivity of detection, thereby improving the sensitivity of analysis and the specificity of analysis diagnosis, solving the problems of the sensitivity and the specificity of different types of exfoliated cells, ensuring the negative discharge rate with high specificity and high sensitivity, providing positive results with high accuracy, and effectively reducing the workload of cell analysis.
Embodiment two:
The present embodiment provides a computer device including: a processor; a memory for storing a computer program; the computer program, when executed by a processor, causes the processor to perform the steps of the detection method described in embodiment one.
The processor may be a logic operation device with data processing capability and/or program execution capability, such as a Central Processing Unit (CPU), a field programmable logic array (FPGA), a single chip Microcomputer (MCU), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), or the like. The one or more processors may be configured to perform the above-described detection methods simultaneously in parallel computing processor groups, or may be configured to perform some of the steps of the above-described detection methods in some of the processors, some of the processors perform other some of the steps of the above-described detection methods, etc. Computer instructions comprise one or more processor operations defined by an instruction set architecture corresponding to a processor, and may be logically contained and represented by one or more computer programs.
The computer program of the present embodiment may be stored on a local memory or downloaded and installed from a network through a communication component.
The memory of the present embodiment may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, a removable storage device, or any suitable combination of the foregoing.
In order to realize the operation of the computer device, it is easy to understand that there are also input/output interfaces, communication interfaces and the like in common, and the input/output interfaces may be configured in the computer device as components or may be externally connected to the device to provide corresponding functions. Wherein the input devices may include a keyboard, mouse, touch screen, microphone, various types of sensors, etc., and the output devices may include a display, speaker, vibrator, indicator lights, etc. The communication interface is used for realizing information communication between the electronic equipment and other devices so as to realize communication interaction between the electronic equipment and other equipment. The communication module may implement communication through a wired manner (such as USB, network cable, etc.), or may implement communication through a wireless manner (such as mobile network, WIFI, bluetooth, etc.).
The computer device may include only the components necessary for implementing the embodiments of the present specification, and not necessarily include all the components shown in the drawings.
The computer equipment provided by the embodiment of the invention can be used as a computer aided diagnosis device (CAD) in medical application, can be used as a computer system for aided interpretation of images, and can be used for assisting professional medical staff and the like to diagnose diseases of patients based on parameter or state description output by the CAD, data obtained by combining clinical examination, biopsy and the like and personal medical experience.
Embodiment III:
The present embodiment is a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method of embodiment one.
The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and to implement the same, but are not intended to limit the scope of the present invention, and all equivalent changes or modifications made according to the spirit of the present invention should be included in the scope of the present invention.
Claims (8)
1. A cell assay method based on a dual assay of nuclear DNA and TBS comprising: a first appearance form judging step and a second appearance form judging step,
The cells were classified into four categories by the first appearance morphology judgment step: normal cells, suspected abnormal cells, highly suspected abnormal cells, and positive abnormal proliferating cells;
The second appearance form judging step is to calculate and analyze the nuclear plasma ratio of the positive abnormal proliferation cells or the highly suspicious abnormal cells obtained in the first appearance form judging step to realize cell analysis, define the nuclear plasma ratio as a DNA parameter 2, respectively preset a threshold D and a threshold E for the DNA parameter 2, and divide the DNA parameter 2 into three numerical ranges, thereby dividing the cells to be analyzed into three types: ① DNA parameter 2 is less than or equal to threshold value D, and is suspicious proliferation cells; ② The threshold value D is less than the DNA parameter 2 and is less than or equal to the threshold value E, and is a highly suspicious abnormal proliferation cell; ③ DNA parameter 2 > threshold E, positive abnormal proliferation cells;
Judging the abnormal proliferation cells judged to be positive by the first appearance form judging step and the second appearance form judging step as first positive abnormal cells; the first appearance form judging step judges positive abnormal proliferation cells, and the second appearance form judging step judges highly suspicious abnormal proliferation cells to be second positive abnormal cells; the first appearance form judging step judges the cells to be highly suspicious abnormal cells, the second appearance form judging step judges the cells to be positive abnormal proliferation cells to be third positive abnormal cells; judging the cell which is judged to be highly suspicious by the first appearance form judging step and the cell which is judged to be highly suspicious by the second appearance form judging step as the fourth positive abnormal cell; the first appearance form judging step judges positive abnormal proliferation cells or highly suspicious abnormal cells, and the second appearance form judging step judges suspicious proliferation cells to be fifth positive abnormal cells; the suspicious abnormal proliferation cells are judged by the first appearance form judgment step, the suspicious proliferation cells or the positive abnormal proliferation cells are judged by the second appearance form judgment step to be the fifth suspicious cells,
The first appearance form judging step is realized by acquiring parameters of the size of the cell nucleus area, defining the parameters as DNA parameters 1, respectively presetting a threshold A, a threshold B and a threshold C for the DNA parameters 1, and dividing the parameters 1 into four numerical ranges, so that the cells to be analyzed are divided into four types: ① Abnormal cells, parameter 1 is less than or equal to threshold A; ② Suspicious abnormal cells, wherein the threshold A is less than the parameter 1 and less than or equal to the threshold B; ③ Highly suspicious abnormal cells, wherein the threshold B is less than the parameter 1 and less than or equal to the threshold C; ④ Positive abnormal proliferation cells, wherein the parameter 1 is larger than the threshold C, and the calculation and analysis of the parameter 1 are realized through a pre-trained neural network;
the second appearance morphology judging step includes a step of calculating the size of the cytoplasmic area of the positive abnormal proliferation cells or the highly suspected abnormal cells located by the first appearance morphology judging step,
The calculation of the cytoplasmic area is to separate the positioned and segmented positive abnormal proliferation cells or highly suspicious abnormal cells to obtain a red channel image R (x, y), a green channel image G (x, y) and a blue channel image B (x, y), and compound the images of any two channels to obtain a second compound image, and calculate the Mean value mean_value of the second compound image; subtracting the mean_value from each pixel of the second composite image by 0.8, and assigning 255 to each new pixel value when the new pixel value is greater than 0, wherein the new pixel value is equal to each pixel value of the second composite image minus the mean_value by 0.8; when the new pixel value is smaller than 0, the new pixel value is assigned to 0, a binarized image Threshold (x, y) is obtained according to the new pixel value, a convex hull is obtained for the Threshold (x, y) image, the area of the convex hull is obtained, and finally the largest area is found out to be the cytoplasmic area of the cell.
2. The method of claim 1, wherein the neural network comprises an input layer, an implied layer, and an output layer, the output layer of the neural network outputting a corresponding in vitro biological sample score, the highest score being used as an analysis index for the first appearance morphology determining step;
The neural network comprises a first neural network module, a second neural network module and a classifier, wherein the first neural network module is a Darknet-53 network, the second neural network module comprises 5 convolution layers, a ReLU layer, a pooling layer, a BN+ LeakyReLU layer and a convolution layer which alternately appear, and the classifier adopts an LR logistic regression classifier.
3. The method according to claim 1, further comprising, before the first appearance morphology determining step, a step of acquiring and preprocessing an image of the cells to be analyzed, comprising in particular the steps of:
(1) Amplifying cells under a microscope by different times, automatically focusing by the microscope, and acquiring a target picture to be analyzed; dividing each picture of a group of pictures shot by a cell under the same view of a microscope into three channels R (red), G (green) and B (blue), compositing any two channels to obtain a first composite image, respectively calculating variances of the pictures relative to the first composite image to obtain Variance values variance_value, respectively comparing the Variance values of each picture, and selecting the picture with the largest Variance value relative to the first composite image as a first target picture of the group of pictures;
(2) And fusing a plurality of first target pictures of different levels in the same visual field to obtain an image of the cell to be analyzed.
4. A method according to claim 3, wherein step (2) is performed as follows
The steps are as follows: defining one first target picture as A (X, y), defining the last picture as A_Pre (X, y), defining the Next picture as A_Next (X, y), defining the fused image as Fuse_image (X, y), setting the pixel values of the three images at the points (X, y) as A (X, y), respectively, setting the pixel values of the three images as A_pre (X, y), and A_Next (X, y), respectively, comparing the three images in 3X3 areas, respectively calculating the sum of the pixel values of the 3X3 areas of the three first target pictures to obtain A_add_value, A_Pre_add_value and A_Next_add_value, and assigning the picture with the largest value to the pixel value at the point (X, y).
Value, traversing the whole image to obtain the fused image.
5. The method according to claim 4, wherein the image of the cells to be analyzed
And (5) positioning and identifying the cell nucleus, and calculating to obtain the area of the cell nucleus.
6. The method according to claim 5, wherein the second appearance judging step further comprises dividing the DNA parameter 1 of the positive abnormal proliferation cells or the highly suspected abnormal cells obtained by the first appearance judging step by the cytoplasmic area of the corresponding cells to obtain a nuclear plasma ratio, and calculating the DNA parameter 2.
7. A computer device comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method of any of claims 1-6.
8. A storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110570957.XA CN113295692B (en) | 2021-05-25 | 2021-05-25 | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110570957.XA CN113295692B (en) | 2021-05-25 | 2021-05-25 | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113295692A CN113295692A (en) | 2021-08-24 |
| CN113295692B true CN113295692B (en) | 2024-06-21 |
Family
ID=77324731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110570957.XA Active CN113295692B (en) | 2021-05-25 | 2021-05-25 | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113295692B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190567A (en) * | 2018-09-10 | 2019-01-11 | 哈尔滨理工大学 | Abnormal cervical cells automatic testing method based on depth convolutional neural networks |
| CN111797786A (en) * | 2020-07-09 | 2020-10-20 | 郑州中普医疗器械有限公司 | Detection method for in vitro biological samples, four-classification, computer device and computer-readable storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5482057B2 (en) * | 2009-09-29 | 2014-04-23 | オリンパス株式会社 | Method for analyzing structure constituting cell nucleus |
| CN106875404A (en) * | 2017-01-18 | 2017-06-20 | 宁波摩视光电科技有限公司 | The intelligent identification Method of epithelial cell in a kind of leukorrhea micro-image |
| CN109145941B (en) * | 2018-07-03 | 2021-03-09 | 怀光智能科技（武汉）有限公司 | Irregular cervical cell mass image classification method and system |
| CN110120040B (en) * | 2019-05-13 | 2021-09-14 | 广州锟元方青医疗科技有限公司 | Slice image processing method, slice image processing device, computer equipment and storage medium |
| CN110705583B (en) * | 2019-08-15 | 2024-03-15 | 平安科技（深圳）有限公司 | Cell detection model training method, device, computer equipment and storage medium |
-
2021
- 2021-05-25 CN CN202110570957.XA patent/CN113295692B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190567A (en) * | 2018-09-10 | 2019-01-11 | 哈尔滨理工大学 | Abnormal cervical cells automatic testing method based on depth convolutional neural networks |
| CN111797786A (en) * | 2020-07-09 | 2020-10-20 | 郑州中普医疗器械有限公司 | Detection method for in vitro biological samples, four-classification, computer device and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113295692A (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ortega et al. | Hyperspectral imaging and deep learning for the detection of breast cancer cells in digitized histological images | |
| AU2014230824A1 (en) | Tissue object-based machine learning system for automated scoring of digital whole slides | |
| CN112784856A (en) | Channel attention feature extraction method and identification method of chest X-ray image | |
| WO2012041333A1 (en) | Automated imaging, detection and grading of objects in cytological samples | |
| CN114332572B (en) | Method for extracting breast lesion ultrasonic image multi-scale fusion characteristic parameters based on saliency map-guided hierarchical dense characteristic fusion network | |
| CN112215799A (en) | Automatic classification method and system for grinded glass lung nodules | |
| JP2022547722A (en) | Weakly Supervised Multitask Learning for Cell Detection and Segmentation | |
| Cabrera et al. | HeMatic: An automated leukemia detector with separation of overlapping blood cells through Image Processing and Genetic Algorithm | |
| Radha | Analysis of COVID-19 and pneumonia detection in chest X-ray images using deep learning | |
| Matias et al. | Segmentation, detection, and classification of cell nuclei on oral cytology samples stained with papanicolaou | |
| CN115908255A (en) | Improved light-weight YOLOX-nano model for target detection and detection method | |
| Hu et al. | Automatic detection of tuberculosis bacilli in sputum smear scans based on subgraph classification | |
| CN111797786B (en) | Detection method for in vitro biological sample, four-classification, computer device and computer-readable storage medium | |
| Yan et al. | HLDnet: Novel deep learning based artificial intelligence tool fuses acetic acid and Lugol’s iodine cervicograms for accurate pre-cancer screening | |
| Hao et al. | DBM-ViT: A multiscale features fusion algorithm for health status detection in CXR/CT lungs images | |
| CN113178228B (en) | Cell analysis method based on nuclear DNA analysis, computer device, and storage medium | |
| CN113295692B (en) | Cell analysis method based on cell nuclear DNA and TBS double analysis method, computer equipment and storage medium | |
| Cao et al. | A deep learning-based method for cervical transformation zone classification in colposcopy images | |
| Durkee et al. | Pseudo-spectral angle mapping for automated pixel-level analysis of highly multiplexed tissue image data | |
| Gao et al. | Evaluation of GAN architectures for visualisation of HPV viruses from microscopic images | |
| Momenzadeh et al. | Computer‐aided diagnosis software for vulvovaginal candidiasis detection from Pap smear images | |
| Tahaa et al. | Superpixel Segmentation-Enabled Transmission Electron Microscopy Images for Rapid and Accurate Detection of Coronavirus | |
| Sergeev et al. | Development of automated computer vision methods for cell counting and endometrial gland detection for medical images processing | |
| Sabata | Digital pathology imaging-The next frontier in medical imaging | |
| Sheikhzadeh et al. | Automatic labeling of molecular biomarkers on a cell-by-cell basis in immunohistochemistry images using convolutional neural networks |
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 |