CN116823823A - Artificial intelligence cerebrospinal fluid cell automatic analysis method - Google Patents
Artificial intelligence cerebrospinal fluid cell automatic analysis method Download PDFInfo
- Publication number
- CN116823823A CN116823823A CN202311091847.0A CN202311091847A CN116823823A CN 116823823 A CN116823823 A CN 116823823A CN 202311091847 A CN202311091847 A CN 202311091847A CN 116823823 A CN116823823 A CN 116823823A
- Authority
- CN
- China
- Prior art keywords
- image
- cell
- cerebrospinal fluid
- value
- cell image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 210000001175 cerebrospinal fluid Anatomy 0.000 title claims abstract description 51
- 238000004458 analytical method Methods 0.000 title claims abstract description 13
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 8
- 210000004027 cell Anatomy 0.000 claims abstract description 86
- 238000013507 mapping Methods 0.000 claims abstract description 21
- 238000001914 filtration Methods 0.000 claims abstract description 16
- 230000011218 segmentation Effects 0.000 claims abstract description 15
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 23
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 210000004460 N cell Anatomy 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 239000003550 marker Substances 0.000 claims description 3
- 230000006870 function Effects 0.000 description 38
- 210000002798 bone marrow cell Anatomy 0.000 description 10
- 238000001514 detection method Methods 0.000 description 6
- 239000000523 sample Substances 0.000 description 6
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 210000004698 lymphocyte Anatomy 0.000 description 3
- 210000001616 monocyte Anatomy 0.000 description 3
- 244000052769 pathogen Species 0.000 description 3
- 238000010186 staining Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 210000001539 phagocyte Anatomy 0.000 description 2
- 210000004180 plasmocyte Anatomy 0.000 description 2
- 210000000278 spinal cord Anatomy 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 210000000576 arachnoid Anatomy 0.000 description 1
- 210000003651 basophil Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 210000003169 central nervous system Anatomy 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 210000004289 cerebral ventricle Anatomy 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000012470 diluted sample Substances 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 210000003979 eosinophil Anatomy 0.000 description 1
- 210000003743 erythrocyte Anatomy 0.000 description 1
- 210000003958 hematopoietic stem cell Anatomy 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000009593 lumbar puncture Methods 0.000 description 1
- 210000004324 lymphatic system Anatomy 0.000 description 1
- 239000012528 membrane Substances 0.000 description 1
- 210000002418 meninge Anatomy 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 210000000440 neutrophil Anatomy 0.000 description 1
- 230000000242 pagocytic effect Effects 0.000 description 1
- 230000001717 pathogenic effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 238000004062 sedimentation Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 210000002330 subarachnoid space Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 210000004881 tumor cell Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/695—Preprocessing, e.g. image segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/698—Matching; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Radiology & Medical Imaging (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Image Analysis (AREA)
Abstract
The application provides an artificial intelligence cerebrospinal fluid cell automatic analysis method, which relates to the technical field of cerebrospinal fluid cell automatic analysis, and comprises the steps of obtaining a cerebrospinal fluid cell image, and preprocessing the cerebrospinal fluid cell image through filtering; establishing an identification window, and performing cell morphology segmentation on the filtered cerebrospinal fluid cell image; calculating a density function value of the image after cell morphology segmentation by adopting a Gaussian kernel function; and (3) mapping the density function value of the input cell image to the distribution value based on the feature image convolution network to obtain the cerebrospinal fluid cell classification.
Description
Technical Field
The application relates to the technical field of cerebrospinal fluid cell automatic analysis, in particular to an artificial intelligence cerebrospinal fluid cell automatic analysis method.
Background
The cell morphology examination is a necessary basis for diagnosing the disease of the cell system, and in clinical examination work, the cerebrospinal fluid cells are observed through a microscope, and the hematopoietic cells of different development stages and normal or abnormal forms are classified, counted and observed in morphology to determine the diagnosis, curative effect and prognosis of the disease of the cerebrospinal fluid system. Cerebrospinal fluid is a colorless transparent viscous liquid located between the arachnoid membrane of the meninges and in the spinal cord. Cerebrospinal fluid occupies the subarachnoid space and the ventricular system around and within the brain and spinal cord. Therefore, a cerebrospinal fluid sample can be obtained through lumbar puncture in medicine, and the sample is analyzed to obtain a relevant diagnosis result. The cells in the cerebrospinal fluid are mainly lymphocytes, and also comprise other monocyte and other types, and the detection and the technology of the cells can be used for diagnosing diseases related to the central nervous system, the cerebral blood vessels and the lymphatic system.
Currently, detection reports of cerebrospinal fluid cells and pathogens are based on manual work. Firstly, performing centrifugal sedimentation smear on cerebrospinal fluid and performing corresponding staining treatment, and then, manually observing the staining condition and morphology of cells and pathogens on the staining specimen under an optical microscope to perform cell classification statistics and pathogen detection. This requires a significant amount of expertise and a great deal of experience for the inspector to ensure the objectivity and accuracy of the inspection results. In actual work, with the increase of the inspection samples, the working intensity of the detection personnel is increased greatly, and the accuracy of result judgment is reduced greatly; the detection period is long, and it takes two days to receive a report from a sample.
Patent application number CN201811168824.4 discloses a bone marrow cell classification method and classification device based on deep learning, wherein the method comprises the following steps: labeling cell positions and classification labels of the bone marrow cells in the bone marrow cell sample image; extracting an image block sample with a single classification label and a preset size from a bone marrow cell sample image; constructing a convolutional neural network of a bone marrow cell classification task, and then training by utilizing a training set consisting of image block samples to obtain a bone marrow cell classification model; cutting the bone marrow cell image to be measured into a plurality of test image blocks with preset sizes, inputting the test image blocks into a bone marrow cell classification model in a traversing way, detecting bone marrow cell edges in the test image blocks, and outputting classification labels and classification confidence probabilities corresponding to the bone marrow cells. The disadvantage is that the existing cell identification and counting methods mostly use manual means. The method for automatically identifying the cells only performs frame cutting on the cells, and the method can lead the identification result to be influenced by cell background and surrounding cells, so that errors are more likely to occur when the cells are dense.
Disclosure of Invention
In order to solve the technical problems, the application provides an artificial intelligence cerebrospinal fluid cell automatic analysis method, which comprises the following steps:
s1, acquiring a cerebrospinal fluid cell image, and preprocessing the cerebrospinal fluid cell image through filtering;
s2, establishing an identification window, dividing and identifying cerebrospinal fluid cells, and counting the cells by using a deep learning detection model;
s3, adopting a Gaussian kernel to represent a deconvolution regression network of single cells to separate overlapped cells, and counting the cells;
and S4, mapping the density function value of the input image to the distribution value based on the feature image convolution network.
Further, in step S3: is provided with N cell images I 1 ,I 2 ,…,I i ,…,I N For arbitrary cell image I i With 2D point set P i Marking P i ={P 1(i) ,…P j(i) ,…P C(i) C (I) is cell image I i The number of labelled cells, j (I) is the cell image I i Number of cells at j-th marker, cell image I i Density function value F of (2) i (p) is calculated by the following formula:
;
where p represents a pixel and,representing a 2D Gaussian kernel function at p, the Gaussian kernel variance being。
Further, in step S4,
the mapping of the density function value of the input cell image to the distribution value is realized according to the following formula:
;
wherein D (I i ) Represents the value of the distribution,representative will Density function value F i (p) input mapping function as variable, c ’ Is a mapping function parameter.
Further, the mapping function is trained by a gradient descent algorithm that uses the training datasetThe gradient, loss function J, was calculated as follows:
;
wherein the method comprises the steps ofIs the predicted value, y j’ Is a true value for ∈>Is +.>The gradient descent method is used for minimizing the loss value of the loss function J, and m is the iteration number.
Further, the method comprises the steps of,;
wherein the method comprises the steps ofIs a loss function vs. parameter->A is the step size of the update.
Further, in step S1, the cell image f is subjected to a first filtering and a second filtering to obtain a filtered imageThe post-image f dg =f*(G 1 -G 2 ) -representing a convolution operation;
first filter function G 1 For preserving low frequency information in the cell image f, a second filter function G 2 For filtering noise in the cell image f;
for the filtered image f dg And performing self-adaptive thresholding to obtain a cell region, and performing hole filling and area constraint on the cell region to obtain a preprocessed image.
Further, in step S2, the identification window is moved pixel by pixel, and coordinates of the upper left corner and the lower right corner of the identification window are obtained once each time the identification window is moved, and the coordinates of the upper left corner and the lower right corner are converted into the radius and the center of the circular frame, and the single cell morphology is segmented according to the circular frame.
Further, the coordinates of the upper left corner and the lower right corner of the recognition window are set to be (x) 1 ,y 1 ) And (x) 2 ,y 2 ) The center C and the radius R of the circular frame are calculated by the following formula,
;
。
compared with the prior art, the application has the following beneficial technical effects:
acquiring a cerebrospinal fluid cell image, and preprocessing the cerebrospinal fluid cell image through filtering; establishing an identification window, and performing cell morphology segmentation on the filtered cerebrospinal fluid cell image; calculating a density function value of the image after cell morphology segmentation by adopting a Gaussian kernel function; the method realizes the mapping from the density function value of the input cell image to the distribution value based on the feature image convolution network, obtains the cerebrospinal fluid cell classification, improves the segmentation accuracy, solves the technical problems of unbalanced brightness and low contrast, does not consume a large amount of calculation resources, and is suitable for large-scale rapid analysis of the cell image.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.
FIG. 1 is a flow chart of an artificial intelligence method for automatically analyzing cerebrospinal fluid cells according to the application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
In the drawings of the specific embodiments of the present application, in order to better and more clearly describe the working principle of each element in the system, the connection relationship of each part in the device is represented, but only the relative positional relationship between each element is clearly distinguished, and the limitations on the signal transmission direction, connection sequence and the structure size, dimension and shape of each part in the element or structure cannot be constructed.
As shown in FIG. 1, the flow chart of the artificial intelligence cerebrospinal fluid cell automatic analysis method according to the application comprises the following steps:
s1, acquiring a cerebrospinal fluid cell image, and preprocessing the cerebrospinal fluid cell image through filtering to eliminate noise interference in the image.
Constructing a first filter and a second filter, and performing first filtering and second filtering on the cell image f to obtain a filtered image f dg =f*(G 1 -G 2 ) And represents a convolution operation.
First filter function G 1 For preserving cell imagesf, a second filter function G 2 For filtering noise in the cell image f.
For the filtered image f dg Performing self-adaptive threshold processing to obtain a cell area; and (5) hole filling and area constraint are carried out on the cell area, and a preprocessed image is obtained.
And establishing an identification window, and performing cell morphology segmentation on the filtered cerebrospinal fluid cell image.
And moving the identification window pixel by pixel, and acquiring coordinates of the upper left corner and the lower right corner of the identification window once, so as to generate a rectangular boundary box.
And converting the coordinates of the upper left corner and the lower right corner into the radius and the circle center of the circular frame, and dividing the single cell form according to the circular frame.
Let the coordinates of the upper left corner and the lower right corner of the recognition window be (x) 1 ,y 1 ) And (x) 2 ,y 2 ) The center C and radius R of the circular frame are calculated by the following formula.
;
。
S3, calculating a density function value of the image after cell morphology segmentation by using a Gaussian kernel function.
Image I after segmentation of N cell morphologies 1 ,I 2 ,…,I i ,…,I N For arbitrary cell image I i With 2D point set P i Marking P i ={P 1(i) ,…P j(i) ,…P C(i) C (I) is cell image I i The number of labelled cells, j (I) is the cell image I i Number of cells at j-th marker, cell image I i Density function value F of (2) i (p) is calculated by the following formula:
;
where p represents a pixel and,representing a 2D Gaussian kernel function at p, the Gaussian kernel variance being。
And S4, mapping the density function value of the input cell image to the distribution value based on the feature image convolution network to obtain the cerebrospinal fluid cell classification.
The mapping of the density function value of the input cell image to the distribution value is realized according to the following formula:
;
wherein D (I i ) Represents the value of the distribution,representative will Density function value F i (p) an input mapping function as a variable, c' being a mapping function parameter.
The mapping from the density function value of the input cell image to the distribution value is realized based on the characteristic mapping convolution network, finally, the cell classification corresponding to the original image of the single cerebrospinal fluid is obtained, and the cell classification is preserved according to cell types, wherein the cell types comprise lymphocyte, erythrocyte, neutrophil, monocyte, basophil, eosinophil, plasma cell, binuclear cell, activated monocyte, phagocyte, phagocytic ferruginous yellow phagocyte, tumor-like cell, activated lymphocyte, binuclear plasma cell and tumor cell, and the total is 15 types of cells.
In a preferred embodiment, the mapping of the density function values of the input cell image to the distribution values is achieved by a eigen-image convolution network, the mapping function of which is trained end-to-end by a gradient descent algorithm.
The gradient descent algorithm is used for obtaining the corresponding value of the independent variable when the minimum value of the loss function is obtained, and the gradient descent algorithm uses a training data setTo calculate the gradient assuming the loss function J is as follows:
;
wherein the method comprises the steps ofIs the predicted value, y j’ Is a true value +.>The gradient descent method is used for minimizing the loss value of the loss function J, and m is the iteration number.
;
Wherein the method comprises the steps ofIs a loss function vs. parameter->A is the step size of the update.
The gray level of a cell area in a cerebrospinal fluid main cell counting image is extremely low, the brightness of a background area occupying most area is high, and the whole gray level of the image is bright; the gray level of the platelet area in the platelet counting block is improved, the brightness of the background area occupying most area is reduced, and the whole gray level of the image is dark.
The meaning of image classification according to the distribution value is: depending on the distribution density of cerebrospinal fluid cells in the actual diluted sample, more than A images containing cerebrospinal fluid cells are distributed, i.e. more than A significant areas are considered to be prone to be a first type of cell count image of cerebrospinal fluid, otherwise platelet count images.
The distribution of the image containing cerebrospinal fluid cells is more than B and not more than A, namely, the image is considered to be a second type of cell count image of cerebrospinal fluid, otherwise, the image is a platelet count image.
In a preferred embodiment, the salient regions of the classified images are structurally extracted.
Specifically, the classified image is converted into a binary image by an automatic segmentation method based on a threshold value, a salient region (namely a foreground part) and a background part are segmented, wherein the part with lower gray level value is divided into the background, and the rest is the foreground.
The grey values of the foreground and the background in the image are replaced, the boundary of the salient region is smoothed under the condition that the areas of the foreground and the background are not obviously changed, burrs or narrow connection possibly existing in the salient region are eliminated, and the edges of all the salient regions in the image are extracted from the binarized image.
Calculating morphological characteristics of the salient region, including the area, roundness and centroid distance of each complete recess structure;
area S is defined as the total number of pixels within the edge of each salient region;
roundness is defined as: c=4pi×s/L 2 ;
Wherein the perimeter L is the total number of pixels on the edge of each salient region;
the centroid distance d is defined as the centroid distance of two adjacent salient regions, wherein the centroid coordinate of each salient region is the average of all pixel point coordinates within its edge.
According to the application, through obtaining a cerebrospinal fluid cell image, preprocessing the cerebrospinal fluid cell image through filtering; establishing an identification window, and performing cell morphology segmentation on the filtered cerebrospinal fluid cell image; calculating a density function value of the image after cell morphology segmentation by adopting a Gaussian kernel function; the method realizes the mapping from the density function value of the input cell image to the distribution value based on the feature image convolution network, obtains the cerebrospinal fluid cell classification, improves the segmentation accuracy, solves the technical problems of unbalanced brightness and low contrast, does not consume a large amount of calculation resources, and is suitable for large-scale rapid analysis of the cell image.
In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, produces a flow or function in accordance with embodiments of the present application, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in or transmitted across a computer-readable storage medium. The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.
While the application has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the application. Therefore, the protection scope of the application is subject to the protection scope of the claims.
Claims (8)
1. An artificial intelligence cerebrospinal fluid cell automatic analysis method, characterized by comprising the following steps:
s1, acquiring a cerebrospinal fluid cell image, and preprocessing the cerebrospinal fluid cell image through filtering;
s2, establishing an identification window, and performing cell morphology segmentation on the filtered cerebrospinal fluid cell image;
s3, calculating a density function value of the image after cell morphology segmentation by adopting a Gaussian kernel function;
and S4, mapping the density function value of the input cell image to the distribution value based on the feature image convolution network to obtain the cerebrospinal fluid cell classification.
2. The method according to claim 1, wherein in step S3: image I after segmentation of N cell morphologies 1 ,I 2 ,…,I i ,…,I N For arbitrary cell image I i With 2D point set P i Marking P i ={P 1(i) ,…P j(i) ,…P C(i) C (I) is cell image I i The number of labelled cells, j (I) is the cell image I i Number of cells at j-th marker, cell image I i Density function value F of (2) i (p) is calculated by the following formula:
;
where p represents a pixel and,representing a 2D Gaussian kernel function at p, the Gaussian kernel variance being +.>。
3. The method for automated analysis of cerebrospinal fluid cells according to claim 1, wherein, in step S4,
the mapping of the density function value of the input cell image to the distribution value is realized according to the following formula:
;
wherein D (I i ) Represents the value of the distribution,representative will Density function value F i (p) an input mapping function as a variable, c' being a mapping function parameter.
4. According to claimThe method for automatically analyzing cerebrospinal fluid cells according to 3, wherein the mapping function is trained by a gradient descent algorithm using a training datasetThe gradient, loss function J, was calculated as follows:
;
wherein the method comprises the steps ofIs the predicted value, y j’ Is a true value for ∈>Is +.>The gradient descent method is used for minimizing the loss value of the loss function J, and m is the iteration number.
5. The method for automated analysis of cerebrospinal fluid cells according to claim 4, wherein said step of,
;
wherein the method comprises the steps ofIs a loss function vs. parameter->A is the step size of the update.
6. The method according to claim 1, wherein in step S1, the cell image f is subjected to a first filtering and a second filtering to obtain a post-filtering imageImage f dg =f*(G 1 -G 2 ) -representing a convolution operation;
first filter function G 1 For preserving low frequency information in the cell image f, a second filter function G 2 For filtering noise in the cell image f;
for the filtered image f dg And performing self-adaptive thresholding to obtain a cell region, and performing hole filling and area constraint on the cell region to obtain a preprocessed image.
7. The method according to claim 1, wherein in step S2, the identification window is moved pixel by pixel, coordinates of an upper left corner and a lower right corner of the identification window are obtained every time the identification window is moved, the coordinates of the upper left corner and the lower right corner are converted into a radius and a center of a circular frame, and the morphology of the individual cells is divided according to the circular frame.
8. The method according to claim 7, wherein the coordinates of the upper left corner and the lower right corner of the recognition window are (x 1 ,y 1 ) And (x) 2 ,y 2 ) The center C and the radius R of the circular frame are calculated by the following formula,
;
。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311091847.0A CN116823823B (en) | 2023-08-29 | 2023-08-29 | Artificial intelligence cerebrospinal fluid cell automatic analysis method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311091847.0A CN116823823B (en) | 2023-08-29 | 2023-08-29 | Artificial intelligence cerebrospinal fluid cell automatic analysis method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116823823A true CN116823823A (en) | 2023-09-29 |
CN116823823B CN116823823B (en) | 2023-11-14 |
Family
ID=88126110
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311091847.0A Active CN116823823B (en) | 2023-08-29 | 2023-08-29 | Artificial intelligence cerebrospinal fluid cell automatic analysis method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116823823B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2948499A1 (en) * | 2016-11-16 | 2018-05-16 | The Governing Council Of The University Of Toronto | System and method for classifying and segmenting microscopy images with deep multiple instance learning |
CN110189292A (en) * | 2019-04-15 | 2019-08-30 | 浙江工业大学 | A kind of cancer cell detection method based on Faster R-CNN and density estimation |
CN110910388A (en) * | 2019-10-23 | 2020-03-24 | 浙江工业大学 | Cancer cell image segmentation method based on U-Net and density estimation |
CN111260638A (en) * | 2020-01-19 | 2020-06-09 | 陕西未来健康科技有限公司 | Automatic detection and counting method and system for cerebrospinal fluid cells and pathogens based on convolutional neural network |
CN112595655A (en) * | 2020-10-28 | 2021-04-02 | 宁夏医科大学总医院 | Automatic cerebrospinal fluid cell analysis system based on artificial intelligence |
CN112595654A (en) * | 2020-10-28 | 2021-04-02 | 宁夏医科大学总医院 | Cerebrospinal fluid cell image feature library and establishing method thereof |
CN113327233A (en) * | 2021-05-28 | 2021-08-31 | 北京理工大学重庆创新中心 | Cell image detection method based on transfer learning |
CN114494197A (en) * | 2022-01-26 | 2022-05-13 | 重庆大学 | Cerebrospinal fluid cell identification and classification method for small-complexity sample |
-
2023
- 2023-08-29 CN CN202311091847.0A patent/CN116823823B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2948499A1 (en) * | 2016-11-16 | 2018-05-16 | The Governing Council Of The University Of Toronto | System and method for classifying and segmenting microscopy images with deep multiple instance learning |
CN110189292A (en) * | 2019-04-15 | 2019-08-30 | 浙江工业大学 | A kind of cancer cell detection method based on Faster R-CNN and density estimation |
CN110910388A (en) * | 2019-10-23 | 2020-03-24 | 浙江工业大学 | Cancer cell image segmentation method based on U-Net and density estimation |
CN111260638A (en) * | 2020-01-19 | 2020-06-09 | 陕西未来健康科技有限公司 | Automatic detection and counting method and system for cerebrospinal fluid cells and pathogens based on convolutional neural network |
CN112595655A (en) * | 2020-10-28 | 2021-04-02 | 宁夏医科大学总医院 | Automatic cerebrospinal fluid cell analysis system based on artificial intelligence |
CN112595654A (en) * | 2020-10-28 | 2021-04-02 | 宁夏医科大学总医院 | Cerebrospinal fluid cell image feature library and establishing method thereof |
CN113327233A (en) * | 2021-05-28 | 2021-08-31 | 北京理工大学重庆创新中心 | Cell image detection method based on transfer learning |
CN114494197A (en) * | 2022-01-26 | 2022-05-13 | 重庆大学 | Cerebrospinal fluid cell identification and classification method for small-complexity sample |
Also Published As
Publication number | Publication date |
---|---|
CN116823823B (en) | 2023-11-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108426994A (en) | Digital holographic microscopy data are analyzed for hematology application | |
Tonti et al. | An automated approach to the segmentation of HEp-2 cells for the indirect immunofluorescence ANA test | |
CN110705403A (en) | Cell sorting method, cell sorting device, cell sorting medium, and electronic apparatus | |
Hortinela et al. | Identification of abnormal red blood cells and diagnosing specific types of anemia using image processing and support vector machine | |
CN112215790A (en) | KI67 index analysis method based on deep learning | |
CN110796661B (en) | Fungal microscopic image segmentation detection method and system based on convolutional neural network | |
CN110517273B (en) | Cytology image segmentation method based on dynamic gradient threshold | |
Maji et al. | An automated method for counting and characterizing red blood cells using mathematical morphology | |
CN113658174B (en) | Microkernel histology image detection method based on deep learning and image processing algorithm | |
CN110148126B (en) | Blood leukocyte segmentation method based on color component combination and contour fitting | |
AU2021349226C1 (en) | Critical component detection using deep learning and attention | |
US20230066976A1 (en) | Systematic characterization of objects in a biological sample | |
Jagadeesh et al. | Image processing based approach to cancer cell prediction in blood samples | |
Mohapatra et al. | Unsupervised leukocyte image segmentation using rough fuzzy clustering. | |
CN112088296A (en) | Method for detecting cells having at least one abnormality in a cytological sample | |
CA3195891A1 (en) | Training end-to-end weakly supervised networks at the specimen (supra-image) level | |
Ghosh et al. | Entropy based divergence for leukocyte image segmentation | |
CN116823823B (en) | Artificial intelligence cerebrospinal fluid cell automatic analysis method | |
Chitra et al. | Detection of aml in blood microscopic images using local binary pattern and supervised classifier | |
Pavithra et al. | White blood cell analysis using watershed and circular hough transform technique | |
CN114742803A (en) | Platelet aggregation detection method combining deep learning and digital image processing algorithm | |
Mahdy et al. | Automatic counting of infected white blood cells using multi-level thresholding | |
Bapure | Automated image analysis for nuclear morphometry using h&e and feulgen stains in prostate biopsies | |
ATICI et al. | Determination of blood group by image processing using digital images | |
Prabaharan et al. | Segmentation of human spermatozoa using improved Havrda-Chavrat entropy-based thresholding method |
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 |