CN113989254A - Pathological section processing method, system and storage medium - Google Patents
Pathological section processing method, system and storage medium Download PDFInfo
- Publication number
- CN113989254A CN113989254A CN202111301387.0A CN202111301387A CN113989254A CN 113989254 A CN113989254 A CN 113989254A CN 202111301387 A CN202111301387 A CN 202111301387A CN 113989254 A CN113989254 A CN 113989254A
- Authority
- CN
- China
- Prior art keywords
- image
- pathological section
- pathological
- cell clusters
- processing
- 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
Images
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
- 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
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a pathological section processing method, a pathological section processing system and a storage medium, which relate to the technical field of computer images and comprise the following specific steps: acquiring a panoramic image of the pathological section; performing sharpening processing on the panoramic image to obtain a first image; and performing quality scoring on the first image, selecting the first image with the highest score, performing three-dimensional modeling, and generating an animation image. According to the invention, the two-dimensional pathological section is visualized into the three-dimensional animation image, so that the pathology can be visually and accurately analyzed and treated, and the risks of missed diagnosis and misdiagnosis of diseases are reduced to a certain extent; the method has the advantages that the method carries out the sharpening processing on the fuzzy picture based on the deep neural network, has multi-scene adaptability, can carry out corresponding training aiming at different fuzzy types and obtain corresponding better processing results, and makes up the defects that the plasticity in the prior art is insufficient and only specific fuzzy can be removed.
Description
Technical Field
The invention relates to the technical field of computer images, in particular to a pathological section processing method, a pathological section processing system and a storage medium.
Background
The pathological medicine is the core of the diagnosis of tumor and cancer, and the pathological section is the key of the pathological diagnosis and is the most important link in the pathology. The pathological diagnosis made by the pathologist in clinic and scientific research usually depends on pathological section, and the histomorphology of the pathological section is observed to judge and attribute the disease. In the traditional mode, a doctor needs to manually observe the slices point by point through a high power microscope; with the improvement of science and technology in the medical industry, the traditional manual operation is gradually replaced by the automatic medical equipment. In the prior art, a pathological section scanner can generate a corresponding high-definition image from a human medical tissue section in a high-power scanning mode, and can digitally store and manage the section scanning image. Therefore, the pathological section scanner can not only improve the observation speed of the tissue section to a great extent, but also ensure the accuracy and high efficiency of doctor diagnosis, and make up for the blank of the traditional mode for storing the slice scanning images.
In the prior art, a doctor generally marks suspicious regions on each pair of slice images respectively by software and medical theory, and finally looks up the suspicious regions one by one to perform overall judgment, which is very direct, but needs to invest in thinking of stereovision, when the regions are small, small suspicious structures are easy to miss, and digital pathological sections are only pictures of a single focus layer, and the risks of missed diagnosis and misdiagnosis are increased when the diseases are confirmed by observing the pictures of the single focus layer compared with the diseases confirmed by directly observing the diseases through a microscope because the lost information is too much, so that for technicians in the field, how to process the medical record slices to reduce misdiagnosis rate and improve the diagnosis efficiency of the doctor is an urgent problem to be solved.
Disclosure of Invention
In view of this, the invention provides a pathological section processing method, a pathological section processing system and a storage medium, which are used for displaying a pathological section in a three-dimensional animation manner, so that doctors can observe the pathological section more three-dimensionally and comprehensively, and the misdiagnosis rate can be reduced to a certain extent.
In order to achieve the purpose, the invention adopts the following technical scheme: on one hand, a pathological section processing method is provided, and the specific steps comprise the following steps:
acquiring a panoramic image of the pathological section;
performing sharpening processing on the panoramic image to obtain a first image;
and performing quality scoring on the first image, selecting the first image with the highest score, performing three-dimensional modeling, and generating an animation image.
Optionally, the method further comprises performing preliminary discrimination and labeling on the animation image through a digital pathology image database.
Optionally, the panoramic image is obtained by:
dividing the pathological section into a plurality of sub-regions to obtain a regional pathological image set corresponding to each sub-region;
and selecting a region focusing image from each region pathological image set for splicing to obtain a panoramic image of the pathological section.
Optionally, the regional pathology image set includes regional focusing images of the corresponding sub-regions at different focusing levels.
Optionally, the method further includes performing sharpening processing on the panoramic image, and the specific steps are as follows:
acquiring a data text, wherein the data text is an original pathological section image;
processing the data text through a fuzzy algorithm to obtain a second image;
training a BiCycleGAN network by taking the data text and the second image as training data to obtain the trained BiCycleGAN network;
and inputting the panoramic image into the BiCycleGAN network which is trained to carry out sharpening processing.
Optionally, the step of performing preliminary discrimination and labeling on the animation image is as follows:
detecting and positioning the cell nucleus of the animation image;
and determining cell clusters according to the cell nucleus, and determining whether the cell clusters are abnormal samples according to the number of the cell clusters.
Optionally, the specific step of determining whether the sample is an abnormal sample according to the number of the cell clusters comprises:
obtaining the number of cells in each of the cell clusters;
acquiring the number of standard cell clusters and the number of standard cells in the standard cell clusters;
and comparing the number of the cell clusters with the number of the standard cell clusters, and comparing the number of the cells in the cell clusters with the number of the standard cells in the standard cell clusters to determine whether the cell clusters are abnormal samples.
On the other hand, the pathological section processing system comprises an image acquisition module, an image sharpening processing module, an image quality grading module and a three-dimensional processing module; wherein the content of the first and second substances,
the image acquisition module is used for acquiring a panoramic image of the pathological section;
the image sharpening processing module is used for sharpening the panoramic image to obtain a first image;
the image quality scoring module is used for performing quality scoring on the first image;
and the three-dimensional processing module is used for selecting the first image with the highest score to perform three-dimensional modeling and generate an animation image.
Optionally, the system further comprises a preliminary screening module, which is used for performing preliminary discrimination and labeling on the animation image through a digital pathological image database.
Finally, a computer storage medium is provided, on which a computer program is stored which, when being executed by a processor, carries out the steps of a method of pathological section processing as described.
Compared with the prior art, the pathological section processing method, the pathological section processing system and the storage medium have the following beneficial technical effects:
(1) the two-dimensional pathological section can be visualized into a three-dimensional animation image, so that the pathology can be visually and accurately analyzed and treated, and the risks of missed diagnosis and misdiagnosis of diseases are reduced to a certain extent.
(2) The method has the advantages that the method carries out the sharpening processing on the fuzzy picture based on the deep neural network, has multi-scene adaptability, can carry out corresponding training aiming at different fuzzy types and obtain corresponding better processing results, and makes up the defects that the plasticity in the prior art is insufficient and only specific fuzzy can be removed.
(3) The preliminary marking and distinguishing of the image can reduce the workload of a pathologist and assist diagnosis.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of a method of the present invention;
fig. 2 is a system configuration diagram of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment 1 of the invention discloses a pathological section processing method, as shown in fig. 1, the method comprises the following specific steps:
s1, acquiring a panoramic image of the pathological section;
s2, performing sharpening processing on the panoramic image to obtain a first image;
and S3, performing quality scoring on the first image, selecting the first image with the highest score, performing three-dimensional modeling, and generating an animation image.
Further, the panoramic image is obtained by the following steps:
s11, dividing the pathological section into a plurality of sub-regions to obtain a regional pathological image set corresponding to each sub-region;
and S12, selecting a region focusing image from each region pathological image set for splicing to obtain a pathological section panoramic image.
Furthermore, the regional pathology image set comprises regional focusing images of the corresponding sub-regions at different focusing levels. The focusing layer is a layer of the pathological section aligned with the microscope camera, and the focal length of the microscope camera is changed by adjusting the screw in the actual scanning and shooting process, so that the microscope camera scans the layer of the pathological section sub-region. The digital image data is three-channel floating-point type digital image data.
Further, the method also comprises the step of carrying out sharpening processing on the panoramic image, and the specific steps are as follows:
s21, acquiring a data text which is an original pathological section image;
s22, processing the data text through a fuzzy algorithm to obtain a second image;
s23, training the BiCycleGAN network by taking the data text and the second image as training data to obtain the trained BiCycleGAN network;
and S24, inputting the panoramic image into the trained BiCycleGAN network for sharpening.
Further, the method further comprises the step of carrying out preliminary discrimination and labeling on the animation image through a digital pathological image database.
Further, the step of performing preliminary discrimination and labeling on the animation image comprises the following steps:
s41, detecting and positioning the cell nucleus of the animation image;
and S42, determining cell clusters according to the cell nucleus, and determining whether the cell clusters are abnormal samples according to the number of the cell clusters.
Specifically, the specific steps of determining whether the sample is an abnormal sample according to the number of the cell clusters are as follows:
s421, obtaining the number of cells in each cell cluster;
s422, acquiring the number of standard cell clusters and the number of standard cells in the standard cell clusters;
and S423, comparing the number of the cell clusters with the number of the standard cell clusters, and comparing the number of the cells in the cell clusters with the number of the standard cells in the standard cell clusters to determine whether the abnormal sample exists.
Further, acquiring panoramic image data of the pathological section, generating a corresponding three-dimensional array according to the digital image data, and converting the three-dimensional array into a binary three-dimensional array; and carrying out three-dimensional modeling according to the binary three-dimensional array to generate a pathological section animation image. The method comprises the following specific steps:
s31, converting the RGB image format of the digital image data into HSV format;
because of the specificity of pathological markers, the original RGB image is easily interfered by white regions when performing color segmentation, requires a very accurate threshold to obtain a clear boundary, and is unstable. The HSV format has ideal anti-interference characteristics, and the divided boundary has only a few noise points.
S32, performing threshold segmentation in the HSV space by adopting a threshold segmentation method to obtain a boundary graph;
the threshold segmentation method is an image segmentation technology based on regions, and the principle is to divide image pixels into a plurality of classes. The image thresholding segmentation is the most common traditional image segmentation method, and becomes the most basic and widely applied segmentation technology in image segmentation due to simple implementation, small calculation amount and stable performance. It is particularly suitable for images where the object and background occupy different gray scale ranges. It not only can compress a great amount of data, but also greatly simplifies the analysis and processing steps, and thus is a necessary image preprocessing process before image analysis, feature extraction and pattern recognition in many cases. The purpose of image thresholding is to divide the set of pixels by gray level, each resulting subset forming a region corresponding to the real scene, each region having consistent properties within it, while adjacent regions do not have such consistent properties. Such a division can be achieved by choosing one or more threshold values from the grey scale.
And S33, performing morphological hole filling according to the boundary graph to obtain a segmentation area.
The morphological closing operation is introduced in the embodiment, and the hole of the closed boundary can be effectively and rapidly filled without generating the outward erosion of the self boundary.
S34, performing augmentation and expansion of the non-data boundary in each direction according to the divided areas;
s35, setting a proper isosurface for the increased and expanded region result extraction surface by adopting a Marchang-cube algorithm to obtain a triangular grid PolyData data structure;
and S36, eliminating an isolated structure in the triangular mesh PolyData data structure, and obtaining a three-dimensional model file.
The embodiment 2 of the invention discloses a pathological section processing system, which comprises an image acquisition module, an image sharpening processing module, an image quality grading module and a three-dimensional processing module, wherein the image acquisition module, the image sharpening processing module, the image quality grading module and the three-dimensional processing module are arranged in sequence; wherein the content of the first and second substances,
the image acquisition module is used for acquiring a panoramic image of the pathological section;
the image sharpening processing module is used for sharpening the panoramic image to obtain a first image;
the image quality scoring module is used for scoring the quality of the first image;
and the three-dimensional processing module is used for selecting the first image with the highest score to perform three-dimensional modeling and generate an animation image.
The system also comprises a prescreening module which is used for preliminarily distinguishing and labeling the animation images through a digital pathological image database, so that the workload of a pathologist can be reduced, and diagnosis can be assisted.
Finally, a computer storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, realizes the steps of a method of pathological section processing.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A pathological section processing method is characterized by comprising the following specific steps:
acquiring a panoramic image of the pathological section;
performing sharpening processing on the panoramic image to obtain a first image;
and performing quality scoring on the first image, selecting the first image with the highest score, performing three-dimensional modeling, and generating an animation image.
2. The pathological section processing method according to claim 1, further comprising performing preliminary discrimination and labeling on the animated image through a digital pathological image database.
3. The pathological section processing method according to claim 1, wherein the panoramic image is obtained by:
dividing the pathological section into a plurality of sub-regions to obtain a regional pathological image set corresponding to each sub-region;
and selecting a region focusing image from each region pathological image set for splicing to obtain a panoramic image of the pathological section.
4. The pathological section processing method according to claim 3, wherein the set of regional pathological images includes regional in-focus images of corresponding sub-regions at different in-focus planes.
5. The pathological section processing method according to claim 1, further comprising a step of performing sharpening processing on the panoramic image, the method comprising the specific steps of:
acquiring a data text, wherein the data text is an original pathological section image;
processing the data text through a fuzzy algorithm to obtain a second image;
training a BiCycleGAN network by taking the data text and the second image as training data to obtain the trained BiCycleGAN network;
and inputting the panoramic image into the BiCycleGAN network which is trained to carry out sharpening processing.
6. The pathological section processing method according to claim 2, wherein the step of preliminarily discriminating and labeling the animated image comprises:
detecting and positioning the cell nucleus of the animation image;
and determining cell clusters according to the cell nucleus, and determining whether the cell clusters are abnormal samples according to the number of the cell clusters.
7. The pathological section treatment method according to claim 6, wherein the step of determining whether the sample is abnormal or not according to the number of the cell clusters comprises:
obtaining the number of cells in each of the cell clusters;
acquiring the number of standard cell clusters and the number of standard cells in the standard cell clusters;
and comparing the number of the cell clusters with the number of the standard cell clusters, and comparing the number of the cells in the cell clusters with the number of the standard cells in the standard cell clusters to determine whether the cell clusters are abnormal samples.
8. A pathological section processing system is characterized by comprising an image acquisition module, an image sharpening processing module, an image quality grading module and a three-dimensional processing module; wherein the content of the first and second substances,
the image acquisition module is used for acquiring a panoramic image of the pathological section;
the image sharpening processing module is used for sharpening the panoramic image to obtain a first image;
the image quality scoring module is used for performing quality scoring on the first image;
and the three-dimensional processing module is used for selecting the first image with the highest score to perform three-dimensional modeling and generate an animation image.
9. The pathological section processing system according to claim 8, further comprising a prescreening module for performing preliminary discrimination and labeling on the animated image through a digital pathological image database.
10. A computer storage medium, characterized in that the computer storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of a method of pathological section treatment according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111301387.0A CN113989254A (en) | 2021-11-04 | 2021-11-04 | Pathological section processing method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111301387.0A CN113989254A (en) | 2021-11-04 | 2021-11-04 | Pathological section processing method, system and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113989254A true CN113989254A (en) | 2022-01-28 |
Family
ID=79746483
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111301387.0A Pending CN113989254A (en) | 2021-11-04 | 2021-11-04 | Pathological section processing method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113989254A (en) |
-
2021
- 2021-11-04 CN CN202111301387.0A patent/CN113989254A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111951221B (en) | Glomerular cell image recognition method based on deep neural network | |
| CN111524137B (en) | Cell identification counting method and device based on image identification and computer equipment | |
| JP7026826B2 (en) | Image processing methods, electronic devices and storage media | |
| US5796861A (en) | Mosaic construction, processing, and review of very large electronic micrograph composites | |
| CN106611413A (en) | Image segmentation method and system | |
| CN111415352B (en) | Cancer metastasis panoramic pathological section analysis method based on deep cascade network | |
| CN110838100A (en) | Colonoscope pathological section screening and segmenting system based on sliding window | |
| CN111161272B (en) | Embryo tissue segmentation method based on generation of confrontation network | |
| CN110288613B (en) | Tissue pathology image segmentation method for ultrahigh pixels | |
| CN112990214A (en) | Medical image feature recognition prediction model | |
| CN114092450A (en) | Real-time image segmentation method, system and device based on gastroscopy video | |
| Kromp et al. | Deep Learning architectures for generalized immunofluorescence based nuclear image segmentation | |
| Kanwal et al. | Quantifying the effect of color processing on blood and damaged tissue detection in whole slide images | |
| CN115100474A (en) | Thyroid gland puncture image classification method based on topological feature analysis | |
| CN111339899B (en) | Catheter feature acquisition method, device, equipment, medium and intelligent microscope | |
| CN112508860B (en) | Artificial intelligence interpretation method and system for positive check of immunohistochemical image | |
| CN116779093B (en) | Method and device for generating medical image structured report and computer equipment | |
| CN113989254A (en) | Pathological section processing method, system and storage medium | |
| CN115880245A (en) | Self-supervision-based breast cancer disease classification method | |
| CN109948706B (en) | Micro-calcification cluster detection method combining deep learning and feature multi-scale fusion | |
| CN117058292B (en) | Tone scale map rendering system based on digital pathological image | |
| JP6329651B1 (en) | Image processing apparatus and image processing method | |
| Quinones et al. | Impact of Patch Extraction Variables on Histopathological Imagery Classification Using Convolution Neural Networks | |
| JP2018125019A (en) | Image processing apparatus and image processing method | |
| CN117542527B (en) | Lung nodule tracking and change trend prediction method, device, equipment and storage medium |
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 |