CN117253229A - Deep learning-based marine mussel micronucleus cell identification and counting method and application - Google Patents
Deep learning-based marine mussel micronucleus cell identification and counting method and application Download PDFInfo
- Publication number
- CN117253229A CN117253229A CN202311534159.7A CN202311534159A CN117253229A CN 117253229 A CN117253229 A CN 117253229A CN 202311534159 A CN202311534159 A CN 202311534159A CN 117253229 A CN117253229 A CN 117253229A
- Authority
- CN
- China
- Prior art keywords
- mussel
- micronucleus
- marine
- cells
- deep learning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 241000237536 Mytilus edulis Species 0.000 title claims abstract description 83
- 235000020638 mussel Nutrition 0.000 title claims abstract description 83
- 238000013135 deep learning Methods 0.000 title claims abstract description 27
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000001514 detection method Methods 0.000 claims abstract description 34
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 238000002372 labelling Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 210000004027 cell Anatomy 0.000 claims description 87
- 239000011521 glass Substances 0.000 claims description 34
- 230000006870 function Effects 0.000 claims description 21
- OKKJLVBELUTLKV-UHFFFAOYSA-N Methanol Chemical compound OC OKKJLVBELUTLKV-UHFFFAOYSA-N 0.000 claims description 15
- 210000000601 blood cell Anatomy 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 9
- 238000004043 dyeing Methods 0.000 claims description 7
- 239000003517 fume Substances 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 claims description 6
- 239000006228 supernatant Substances 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 5
- 238000001035 drying Methods 0.000 claims description 5
- 238000005457 optimization Methods 0.000 claims description 5
- 238000007789 sealing Methods 0.000 claims description 5
- LOKCTEFSRHRXRJ-UHFFFAOYSA-I dipotassium trisodium dihydrogen phosphate hydrogen phosphate dichloride Chemical compound P(=O)(O)(O)[O-].[K+].P(=O)(O)([O-])[O-].[Na+].[Na+].[Cl-].[K+].[Cl-].[Na+] LOKCTEFSRHRXRJ-UHFFFAOYSA-I 0.000 claims description 4
- 239000002953 phosphate buffered saline Substances 0.000 claims description 4
- 239000006285 cell suspension Substances 0.000 claims description 3
- 230000006378 damage Effects 0.000 claims description 3
- 238000013136 deep learning model Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 239000003292 glue Substances 0.000 claims description 3
- 238000010438 heat treatment Methods 0.000 claims description 3
- 210000000087 hemolymph Anatomy 0.000 claims description 3
- 210000003205 muscle Anatomy 0.000 claims description 3
- 239000008363 phosphate buffer Substances 0.000 claims description 3
- 239000013535 sea water Substances 0.000 claims description 3
- 239000000725 suspension Substances 0.000 claims description 3
- 238000013526 transfer learning Methods 0.000 claims description 3
- 210000001015 abdomen Anatomy 0.000 claims description 2
- 238000007605 air drying Methods 0.000 claims description 2
- 239000012266 salt solution Substances 0.000 claims description 2
- 230000004907 flux Effects 0.000 abstract description 5
- 238000004806 packaging method and process Methods 0.000 abstract description 2
- 230000007547 defect Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000000684 flow cytometry Methods 0.000 description 4
- 238000002360 preparation method Methods 0.000 description 4
- 238000010186 staining Methods 0.000 description 4
- 238000003912 environmental pollution Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- MHAJPDPJQMAIIY-UHFFFAOYSA-N Hydrogen peroxide Chemical compound OO MHAJPDPJQMAIIY-UHFFFAOYSA-N 0.000 description 2
- 239000000090 biomarker Substances 0.000 description 2
- 230000005779 cell damage Effects 0.000 description 2
- 208000037887 cell injury Diseases 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000000338 in vitro Methods 0.000 description 2
- 238000011065 in-situ storage Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 239000008188 pellet Substances 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 108010047956 Nucleosomes Proteins 0.000 description 1
- FAPWRFPIFSIZLT-UHFFFAOYSA-M Sodium chloride Chemical compound [Na+].[Cl-] FAPWRFPIFSIZLT-UHFFFAOYSA-M 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000003855 cell nucleus Anatomy 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 210000000805 cytoplasm Anatomy 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 210000001623 nucleosome Anatomy 0.000 description 1
- 210000004940 nucleus Anatomy 0.000 description 1
- 238000000643 oven drying Methods 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000011780 sodium chloride Substances 0.000 description 1
- 238000007447 staining method Methods 0.000 description 1
- 238000003911 water pollution Methods 0.000 description 1
Classifications
-
- 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
- G06V20/693—Acquisition
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06M—COUNTING MECHANISMS; COUNTING OF OBJECTS NOT OTHERWISE PROVIDED FOR
- G06M11/00—Counting of objects distributed at random, e.g. on a surface
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/096—Transfer learning
-
- 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
- G06V10/30—Noise filtering
-
- 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
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
- G01N2015/1006—Investigating individual particles for cytology
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/80—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in fisheries management
- Y02A40/81—Aquaculture, e.g. of fish
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Chemical & Material Sciences (AREA)
- Dispersion Chemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention belongs to the technical field of micronucleus cell detection, and particularly relates to a deep learning-based marine mussel micronucleus cell identification and counting method and application, wherein the method comprises the steps of preparing a marine mussel micronucleus packaging sheet and collecting micronucleus pictures; constructing a deep learning data set; the method comprises the steps of obtaining image data of marine mussel cells, and dividing a training set and a testing set; labeling the image data of the training set and the testing set; preprocessing the image to enhance the edge and detail of the target cells; deep learning-based marine mussel micronucleus cell identification and counting. The invention has the advantages of high detection flux, high efficiency, no subjectivity and the like, and can rapidly and accurately complete the marine environment detection task when applied to marine environment detection.
Description
Technical Field
The invention belongs to the technical field of micronucleus cell detection, and particularly relates to a marine mussel micronucleus cell identification and counting method based on deep learning and application thereof.
Background
Frequent marine environmental pollution threatens the health and biological safety of marine ecology, leading to sharp reduction of biodiversity and reduction of the stability and safety of a marine ecological system. Marine mussels are used as model organisms, are widely used for camping, fixing and living, are widely distributed and have strong adaptability, and are applied to various marine environment monitoring projects. Marine mussels have a variety of sensitive biomarkers that can be used to indicate water pollution in a particular sea area over time. The detection of the micronuclei of marine mussels in vitro is an important biomarker detection method. Marine environmental pollution can cause damage to the genetic material of marine mussels, and can cause distortion, mutation and further cause deformity and tumor. Therefore, by carrying out in-vitro micronucleus detection on the marine mussels, early warning can be provided, and important roles are played for the protection and repair work of marine environments.
The marine mussel micronucleus detection method mainly comprises two steps: and (5) manual microscopic reading and flow cytometry detection.
The manual microscopic reading refers to preparation of marine mussel cell packaging sheets by a staining method, microscopic examination under a microscope, and counting of micronucleus cells according to micronucleus characteristics (nucleosomes independent of cell nuclei in cytoplasm, and the volume is about 1/16-1/3 of that of main nuclei). The method has the defects of low detection efficiency, small detection flux and relatively large error although the operation is simple and the process is easy to realize. The method has certain requirements on the professionals of the detection personnel, and whether the detection personnel can accurately identify the micronucleus cells or not and whether the micronucleus cells are counted subjectively can influence the experimental result to a certain extent.
The flow cytometry detection mainly comprises marine mussel cell sample preparation, cell fixation, cell staining, cell sample injection, flow cytometry measurement, micronucleus detection and result interpretation, so that cells with micronuclei are selected according to the staining characteristics and fluorescence intensity of the cells. The method has the advantages of large detection flux, relatively unbiased output result and the like, but has the defects of high requirement on sample quality, incapability of being applied to field in-situ monitoring and the like. And the flow cytometry is relatively expensive, and the operation flows of setting the gating conditions, analyzing the dyeing characteristics and the like are required to be searched for a long time.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a deep learning-based marine mussel micronucleus cell identification and counting method which has high detection flux, high efficiency, no defects of subjectivity and the like, and can rapidly and accurately complete marine environment detection tasks when applied to marine environment detection.
The invention provides a deep learning-based marine mussel micronucleus cell identification and counting method, which comprises the following steps:
s1, preparing a micronucleus tablet of marine mussel, and collecting micronucleus photos;
s2, constructing a deep learning data set; the method comprises the steps of obtaining image data of marine mussel cells, and dividing a training set and a testing set; labeling the image data of the training set and the testing set; preprocessing the image to enhance the edge and detail of the target cells;
s3, identifying and counting marine mussel micronucleus cells based on deep learning;
s31, performing target identification by using a deep learning algorithm YOLOv8, adopting transfer learning, initializing a YOLOv8S model by using the weight of a pre-training model on a COCO data set, and accurately positioning and identifying marine mussel micronucleus cells by training an identification model YOLOv 8S;
s32, model training: training a deep learning model by using the image data of the training set and the corresponding label information;
s33, model training optimization;
s34, model evaluation: using the image data of the test set to evaluate the identification accuracy and performance of the trained model on the new data;
s35, the trained model automatically identifies mussel cells in the picture and endows the mussel cells with identification frames and categories so as to accurately calibrate the position and category information of each cell;
s36, adding a counting function in an algorithm, so that mussel cells are identified and automatically counted;
s37, generating a statistical result: and generating a corresponding statistical report according to the counting result, wherein the statistical report comprises information such as mussel cell number, density and the like.
Further, step S1 includes:
s11, extracting blood cells of marine mussels: picking the shell from the mussel abdomen side with scissors, evacuating the seawater, extracting 400 μL of haemolymph from the posterior adductor muscle of the mussel, centrifuging at 4deg.C and 5000 rpm for 2 min to obtain cell mass with supernatant removed;
s12, cell treatment: adding 200 mu L of phosphate buffer salt solution into the centrifuged cell mass, and gently sucking the cell mass by a pipette to resuspend the cell mass; centrifuging the cell suspension in a centrifuge at 4deg.C and 5000 rpm for 2 min, and removing supernatant; resuspending the washed cell mass in 100 μl phosphate buffered saline;
s13, attaching and fixing: marking the frosted surface of a glass slide, adding the prepared cell heavy suspension into the glass slide in two drops of 50 mu L each, and placing the glass slide in a 50 ℃ oven for uniform heating and drying so that mussel blood cells can be attached to the glass slide; placing the glass slide with the attached blood cells into 100% methanol for fixation for 10 min, and then placing the glass slide into a fume hood for air drying for 10 min;
s14, dyeing and sealing: placing the air-dried glass slide in the prepared Jim Sa dye for dyeing for 5 min, and immersing the glass slide for 2 times for cleaning for 5 min each time; drying the washed glass slide in a fume hood, and sealing the glass slide with Eukitt glue;
s15, microkernel picture acquisition: and after the glass slides are dried, observing the dyed glass slides under a microscope and photographing, so as to obtain a picture to be measured.
Further, in step S2, the image preprocessing uses an OpenCV computer vision library to perform preprocessing on the acquired image set, including enhancement of the image using a normal () function and an equalzehist () function, adjustment of brightness and contrast using a convertScaleAbs () function, color balancing using a cvtColor () function, application of denoising processing, median filtering using a median blur () function to remove noise interference in the image, and adjustment of contrast according to the characteristics of the image.
Further, in step S33, the model is optimized by forward propagation, computational loss, backward propagation, parameter updating, and multiple iterations using a cross entropy loss function and a random gradient descent (SGD) optimization algorithm.
The invention relates to application of a deep learning-based marine mussel micronucleus cell identification and counting method in marine environment detection, wherein the cell damage degree is represented by calculating micronucleus occurrence frequency (the number of micronucleus cells per 1000 cells), and if the observed marine mussel micronucleus rate in a sea area reaches 30-50 per mill, the sea area is proved to have potential pollution sources.
The invention has the beneficial effects that: the invention introduces a deep learning technology, combines image processing and a machine learning algorithm, and fills the technical blank of the domestic marine environment detection field in the aspect of mussel micronucleus detection.
According to the invention, by constructing the deep learning-based marine mussel micronucleus identification model and applying an advanced computer vision algorithm, the marine mussel micronucleus detection technology is simplified, and the marine mussel micronucleus detection efficiency and precision are improved. And the field in-situ environmental pollution monitoring is conveniently carried out. The marine environment detection system can rapidly and accurately complete marine environment detection tasks.
The invention realizes the automation of identifying and counting and detecting the marine environment, realizes the automatic detection of the marine mussel micronucleus cells through an intelligent image processing system, reduces the requirement of manual operation and reduces the risk of subjective errors.
Drawings
FIG. 1 is a diagram showing the identification of sea mussel micronucleus cells according to the present invention;
FIG. 2 is a graph showing the typing and counting of marine mussel cell samples according to the present invention.
Detailed Description
For a better understanding of the present invention, embodiments of the present invention are explained in detail below with reference to fig. 1-2.
The invention provides an image recognition technology based on deep learning for recognizing and counting micronucleus cells of marine mussels, which realizes quick and accurate recognition and counting of the marine mussel cells by combining an image processing and machine learning algorithm by utilizing a target recognition technology, constructs a high-precision recognition algorithm for the marine organism micronucleus cells and solves the defects of low detection flux, low efficiency, subjectivity and the like in the conventional marine mussel micronucleus detection. In China, intelligent image processing is applied to marine mussel micronucleus detection so as to achieve the aim of marine environment detection, and the method is still in a blank stage at present.
In order to solve the technical problems, the invention is realized by the following technical scheme:
s1, preparation of micronucleus tablet of marine mussel and collection of micronucleus photo
Material preparation: marine mussels (control sea area, contaminated sea area), giemsa dye liquor, wash liquor, methanol, staining jar, eukitt's gum, hydrogen peroxide (1 mM), phosphate buffered saline solution, glass slides, coverslips, centrifuge tubes (1.5 mL), sterile syringes (1 mL), pipettes (200 μl), tips, papanicolaou tubes, centrifuges, ovens, fume hoods, microscopes, and the like. The method comprises the following steps: (1) marine mussel blood cell collection operation; (2) marine mussel blood cell cleaning and resuspension operation; (3) dripping; (4) attaching operation: oven drying to make the extracted sea mussel blood cells adhere to the glass slide; (5) methanol fixation operation; (6) Giemsa solution staining and rinsing; and (7) collecting pictures of the micronucleus cells of the marine mussels. The specific steps are as follows:
s11, extracting blood cells of marine mussels: the shells were pried off from the ventral mussel with scissors, the internal sea water was emptied, 400. Mu.L of haemolymph was extracted from the posterior adductor muscle of mussel, and centrifuged at 5000 rpm for 2 min at 4℃to obtain a cell mass from which the supernatant had been removed.
S12, cell treatment: 200. Mu.L of phosphate buffer saline was added to the centrifuged pellet, and the pellet was gently aspirated with a pipette, and resuspended. The cell suspension was centrifuged in a centrifuge for 2 min (4 ℃,5000 rpm) and the supernatant removed. The washed cell mass was resuspended in 100. Mu.L phosphate buffered saline.
S13, attaching and fixing: marking the frosted surface of the glass slide, adding the prepared cell heavy suspension into the glass slide in two drops (50 mu L of each drop), and placing the glass slide in a 50 ℃ oven for uniform heating and drying so that mussel blood cells can be attached to the glass slide. The slide after blood cell attachment was fixed in 100% methanol for 10 min, and then left to air dry in a fume hood for 10 min.
S14, dyeing and sealing: and (3) placing the air-dried glass slide in the prepared Jimsa dye for dyeing for 5 min, and immersing the glass slide in the cleaning liquid for 2 times for cleaning for 5 min each time. The washed slides were dried in a fume hood and then blocked with Eukitt glue.
S15, microkernel picture acquisition: and after the glass slides are dried, observing the dyed glass slides under a microscope and photographing, so as to obtain a picture to be measured.
S2, deep learning dataset construction
S21, image acquisition: image data of marine mussel cells are acquired by high resolution image acquisition equipment according to 8:2, dividing the acquired image data into a training set and a testing set for training and evaluating the identification model.
S22, data marking: labeling the image data of the training set and the testing set, namely adding label information for why the marine mussel cells in each image. Labeling includes accurately marking the location and boundaries of each mussel cell by drawing bounding boxes or pixel-level labeling. The labeling process may be performed manually by a professional or assisted by a semi-automatic labeling tool.
S23, image preprocessing: the acquired image set is preprocessed by using an OpenCV computer vision library, wherein the preprocessing comprises the steps of enhancing the image by using a normal () function and an equalHist () function, adjusting brightness and contrast by using a convertScaleAbs () function, performing color balance by using a cvtColor () function, applying denoising processing, removing noise interference in the image by using median filtering of a mediaBlur () function, adjusting contrast according to the characteristics of the image, and enhancing the edge and detail of target cells so as to improve the subsequent target recognition effect. The processing criteria will vary depending on the particular image. In general, the goal of enhancement operations is to improve the contrast, detail, and visual effect of the image. The criteria for adjusting brightness and contrast is to make the image appear to come from without loss of detail. The criteria for color balance is to keep the colors of the image balanced and coordinated between the different channels. Specific processing parameters and standards need to be adjusted and optimized according to actual conditions. In actual use, the best parameters and criteria are determined by multiple trials and visual evaluations.
S3, identifying and counting marine mussel micronucleus cells based on deep learning (in figure 1, MNi represents the identified micronucleus cells; C represents the identified normal cells).
S31, performing target identification by using a deep learning algorithm YOLOv8, adopting transfer learning, initializing a YOLOv8S model by using the weight of a pre-training model on a COCO data set, and accurately positioning and identifying the marine mussel micronucleus cells by training an identification model YOLOv 8S.
S32, training a model: using the image data of the training set and the corresponding label information, training a deep learning model, using the YOLOv8 algorithm.
S33, optimizing model training: the model is optimized using a cross entropy loss function and a random gradient descent (SGD) optimization algorithm to forward propagate, calculate loss, counter propagate, update parameters, and iterate multiple times to improve the accuracy and generalization ability of the model.
S34, model evaluation: and evaluating the recognition accuracy and performance of the trained model on the new data by using the image data of the test set. According to the prediction result and the label information, the evaluation index of the model on the new data is calculated, wherein the evaluation index comprises an Accuracy rate (Accuracy), a Precision rate (Precision), a Recall rate (Recall), an F1 score (F1 score) and the like.
And S35, the trained model can automatically identify mussel cells in the picture and endow the mussel cells with identification frames and categories so as to accurately mark the position and category information of each cell.
S36, automatic counting: the counting function is added in the algorithm, so that the mussel cells are automatically counted while being identified (FIG. 2, MNi represents the identified micronucleus cells; C represents the identified normal cells, and the number represents the number of the cells).
S37, generating a statistical result: and generating a corresponding statistical report according to the counting result, wherein the statistical report comprises information such as mussel cell number, density and the like. At least 1000 cells were read per treatment group, from which the amount of micronuclei were found for counting.
The application of the deep learning-based marine mussel micronucleus cell identification and counting method in marine environment detection is characterized in that the cell damage degree is represented by calculating the micronucleus occurrence frequency (the number of micronucleus cells/1000 cells), the micronucleus occurrence rate of the mussels in a control group is about 10 per mill under the general condition, and if the micronucleus rate of the marine mussels in the sea area is observed to be 30-50 per mill, the potential pollution source exists in the sea area, and the mussels are suffered from health damage to a certain extent.
The principles and embodiments of the present invention have been described with reference to specific examples, which are provided to facilitate understanding of the method and core ideas of the present invention. The foregoing is merely illustrative of the preferred embodiments of this invention, and it is noted that there is objectively no limit to the specific structure disclosed herein, since numerous modifications, adaptations and variations can be made by those skilled in the art without departing from the principles of the invention, and the above-described features can be combined in any suitable manner; such modifications, variations and combinations, or the direct application of the inventive concepts and aspects to other applications without modification, are contemplated as falling within the scope of the present invention.
Claims (5)
1. The marine mussel micronucleus cell identification and counting method based on deep learning is characterized by comprising the following steps of:
s1, preparing a micronucleus tablet of marine mussel, and collecting micronucleus photos;
s2, constructing a deep learning data set; the method comprises the steps of obtaining image data of marine mussel cells, and dividing a training set and a testing set; labeling the image data of the training set and the testing set; preprocessing the image to enhance the edge and detail of the target cells;
s3, identifying and counting marine mussel micronucleus cells based on deep learning;
s31, performing target identification by using a deep learning algorithm YOLOv8, adopting transfer learning, initializing a YOLOv8S model by using the weight of a pre-training model on a COCO data set, and accurately positioning and identifying marine mussel micronucleus cells by training an identification model YOLOv 8S;
s32, model training: training a deep learning model by using the image data of the training set and the corresponding label information;
s33, model training optimization;
s34, model evaluation: using the image data of the test set to evaluate the identification accuracy and performance of the trained model on the new data;
s35, the trained model automatically identifies mussel cells in the picture and endows the mussel cells with identification frames and categories so as to accurately calibrate the position and category information of each cell;
s36, adding a counting function in an algorithm, so that mussel cells are identified and automatically counted;
s37, generating a statistical result: and generating a corresponding statistical report according to the counting result, wherein the statistical report comprises information such as mussel cell number, density and the like.
2. The deep learning-based marine mussel micronucleus cell identification and enumeration method of claim 1, wherein step S1 comprises:
s11, extracting blood cells of marine mussels: picking the shell from the mussel abdomen side with scissors, evacuating the seawater, extracting 400 μL of haemolymph from the posterior adductor muscle of the mussel, centrifuging at 4deg.C and 5000 rpm for 2 min to obtain cell mass with supernatant removed;
s12, cell treatment: adding 200 mu L of phosphate buffer salt solution into the centrifuged cell mass, and gently sucking the cell mass by a pipette to resuspend the cell mass; centrifuging the cell suspension in a centrifuge at 4deg.C and 5000 rpm for 2 min, and removing supernatant; resuspending the washed cell mass in 100 μl phosphate buffered saline;
s13, attaching and fixing: marking the frosted surface of the glass slide, dripping the prepared cell heavy suspension on the glass slide in two drops, and placing the glass slide in a 50 ℃ oven for uniform heating and drying so that mussel blood cells can be attached on the glass slide; placing the glass slide with the attached blood cells into 100% methanol for fixation for 10 min, and then placing the glass slide into a fume hood for air drying for 10 min;
s14, dyeing and sealing: placing the air-dried glass slide in the prepared Jim Sa dye for dyeing for 5 min, and immersing the glass slide for 2 times for cleaning for 5 min each time; drying the washed glass slide in a fume hood, and sealing the glass slide with Eukitt glue;
s15, microkernel picture acquisition: and after the glass slides are dried, observing the dyed glass slides under a microscope and photographing, so as to obtain a picture to be measured.
3. The deep learning-based marine mussel micronucleus cell identification and counting method according to claim 1, wherein in step S2, the image preprocessing uses OpenCV computer vision library to preprocess the acquired image set, including enhancement of the image using normal () function and equalhist () function, adjustment of brightness and contrast using convertScaleAbs () function, color balancing using cvtdcolor () function, application of denoising processing, median filtering using mediaBlur () function to remove noise interference in the image, and adjustment of contrast according to the characteristics of the image.
4. The deep learning based marine mussel micronucleus cell identification and enumeration method of claim 1, wherein in step S33, the model is optimized by forward propagation, computational loss, back propagation, parameter updating, and multiple iterations using a cross entropy loss function and a random gradient descent optimization algorithm.
5. The application of the deep learning-based marine mussel micronucleus cell identification and counting method in marine environment detection is characterized in that the damage degree of cells is represented by calculating the occurrence frequency of micronuclei, and if the observed marine mussel micronucleus rate reaches 30-50 per mill, the sea area is proved to have potential pollution sources.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311534159.7A CN117253229A (en) | 2023-11-17 | 2023-11-17 | Deep learning-based marine mussel micronucleus cell identification and counting method and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311534159.7A CN117253229A (en) | 2023-11-17 | 2023-11-17 | Deep learning-based marine mussel micronucleus cell identification and counting method and application |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117253229A true CN117253229A (en) | 2023-12-19 |
Family
ID=89129878
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311534159.7A Pending CN117253229A (en) | 2023-11-17 | 2023-11-17 | Deep learning-based marine mussel micronucleus cell identification and counting method and application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117253229A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117611588A (en) * | 2024-01-23 | 2024-02-27 | 浙江大学海南研究院 | Marine environment pollution early detection and early warning method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106568824A (en) * | 2016-11-08 | 2017-04-19 | 浙江大学 | Method for detecting DNA damage of mussels under stress of compound heavy metals |
| CN111458269A (en) * | 2020-05-07 | 2020-07-28 | 厦门汉舒捷医疗科技有限公司 | Artificial intelligent identification method for peripheral blood lymph micronucleus cell image |
| CN113658174A (en) * | 2021-09-02 | 2021-11-16 | 北京航空航天大学 | Microkaryotic image detection method based on deep learning and image processing algorithm |
| CN114720355A (en) * | 2022-03-03 | 2022-07-08 | 浙江大学 | Particle counting method, device and system suitable for imaging flow cytometer |
| CN116994244A (en) * | 2023-08-16 | 2023-11-03 | 临海市特产技术推广总站(临海市柑桔产业技术协同创新中心) | Method for evaluating fruit yield of citrus tree based on Yolov8 |
| CN117036238A (en) * | 2023-05-31 | 2023-11-10 | 华南农业大学 | Video target tracking-based field fruit counting method and system |
-
2023
- 2023-11-17 CN CN202311534159.7A patent/CN117253229A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106568824A (en) * | 2016-11-08 | 2017-04-19 | 浙江大学 | Method for detecting DNA damage of mussels under stress of compound heavy metals |
| CN111458269A (en) * | 2020-05-07 | 2020-07-28 | 厦门汉舒捷医疗科技有限公司 | Artificial intelligent identification method for peripheral blood lymph micronucleus cell image |
| CN113658174A (en) * | 2021-09-02 | 2021-11-16 | 北京航空航天大学 | Microkaryotic image detection method based on deep learning and image processing algorithm |
| CN114720355A (en) * | 2022-03-03 | 2022-07-08 | 浙江大学 | Particle counting method, device and system suitable for imaging flow cytometer |
| CN117036238A (en) * | 2023-05-31 | 2023-11-10 | 华南农业大学 | Video target tracking-based field fruit counting method and system |
| CN116994244A (en) * | 2023-08-16 | 2023-11-03 | 临海市特产技术推广总站(临海市柑桔产业技术协同创新中心) | Method for evaluating fruit yield of citrus tree based on Yolov8 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117611588A (en) * | 2024-01-23 | 2024-02-27 | 浙江大学海南研究院 | Marine environment pollution early detection and early warning method |
| CN117611588B (en) * | 2024-01-23 | 2024-05-24 | 浙江大学海南研究院 | Marine environment pollution early detection and early warning method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10365203B2 (en) | Method and apparatus for automated whole blood sample analyses from microscopy images | |
| CN111260638B (en) | Automatic detection and counting method and system for cerebrospinal fluid cells and pathogens based on convolutional neural network | |
| CN117253229A (en) | Deep learning-based marine mussel micronucleus cell identification and counting method and application | |
| WO2018140014A1 (en) | Classifying biological samples using automated image analysis | |
| CN111504885B (en) | Analysis method of full-automatic blood smear morphological analysis device based on machine vision | |
| CN101750272A (en) | Blood cell image recognition counting method | |
| WO2020133635A1 (en) | Fully automatic sperm cell detector | |
| WO2011030481A1 (en) | Device for bacteria classification and pretreatment device for bacteria test | |
| CN101809589A (en) | Methods and systems for processing biological specimens utilizing multiple wavelengths | |
| KR20200135460A (en) | Digitalization flow method of bone marrow plate | |
| CN113450317A (en) | Immunofluorescence image detection method for gynecological clinical microbial infection | |
| Evangeline et al. | Computer aided system for human blood cell identification, classification and counting | |
| CN107367456A (en) | One kind is disposable to wash image class streaming fluorescence detection method and system | |
| CN112613505A (en) | Cell micronucleus identification, positioning and counting method based on deep learning | |
| CN112540039A (en) | Method for directly calculating number of adherent living cells | |
| Kamentsky et al. | Instrumentation for automated examinations of cellular specimens | |
| CN116757998A (en) | Screening method and device for CTC cells and CTC-like cells based on AI | |
| JP5343131B2 (en) | High-throughput system for CFU assays by using high-resolution digital imaging, dyeing and automated laboratory systems | |
| Paray et al. | Gram staining: A Brief review | |
| CN114152557B (en) | Image analysis-based blood cell counting method and system | |
| CN109633143B (en) | System and method for detecting bone marrow microcirculation environment of patient after hematopoietic stem cell transplantation | |
| CN207717473U (en) | Vaginal fluid visible component automatic staining detection device | |
| Fetter et al. | StomataCounter: a deep learning method applied to automatic stomatal identification and counting | |
| Wang et al. | A cell counting method for BEVS based on nonlinear Transformed Sliding Band Filter | |
| CN116183470B (en) | Woody plant cell programmed death detection method based on flow sorting |
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 |