CN112837324A - Automatic tumor image region segmentation system and method based on improved level set - Google Patents
Automatic tumor image region segmentation system and method based on improved level set Download PDFInfo
- Publication number
- CN112837324A CN112837324A CN202110084653.2A CN202110084653A CN112837324A CN 112837324 A CN112837324 A CN 112837324A CN 202110084653 A CN202110084653 A CN 202110084653A CN 112837324 A CN112837324 A CN 112837324A
- Authority
- CN
- China
- Prior art keywords
- tumor
- image
- module
- tumor image
- level set
- 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
- 206010028980 Neoplasm Diseases 0.000 title claims abstract description 228
- 230000011218 segmentation Effects 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 49
- 230000004927 fusion Effects 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000007405 data analysis Methods 0.000 claims abstract description 20
- 238000000605 extraction Methods 0.000 claims abstract description 19
- 238000003062 neural network model Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 13
- 238000003066 decision tree Methods 0.000 claims abstract description 11
- 238000007637 random forest analysis Methods 0.000 claims abstract description 4
- 238000004458 analytical method Methods 0.000 claims description 14
- 230000002708 enhancing effect Effects 0.000 claims description 12
- 238000009499 grossing Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 3
- 238000000354 decomposition reaction Methods 0.000 claims description 3
- 238000009792 diffusion process Methods 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 238000003759 clinical diagnosis Methods 0.000 abstract 1
- 238000003745 diagnosis Methods 0.000 description 6
- 230000003203 everyday effect Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 208000014674 injury Diseases 0.000 description 2
- 230000001575 pathological effect Effects 0.000 description 2
- 230000008733 trauma Effects 0.000 description 2
- 238000005481 NMR spectroscopy Methods 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000036210 malignancy Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010827 pathological analysis Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- 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
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- 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)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention belongs to the technical field of image region segmentation, and discloses an automatic tumor image region segmentation system and method based on an improved level set, wherein the automatic tumor image region segmentation system based on the improved level set comprises the following steps: the tumor image processing system comprises a tumor image acquisition module, an image preprocessing module, a central processing module, an image region segmentation module, an image feature extraction module, an image feature fusion module, an image data analysis module, a cloud storage module and a display module. The image feature extraction module is used for setting the neural network model into a multi-branch neural network structure, so that two-dimensional tumor features and three-dimensional tumor features of the tumor image can be captured simultaneously; the image data analysis module adopts a random forest algorithm to generate a plurality of decision trees, and the classification results obtained by the two-dimensional SVM algorithm are reclassified according to the decision trees, so that the classification accuracy is improved, and the method has great practical value in clinical diagnosis and treatment.
Description
Technical Field
The invention belongs to the technical field of image region segmentation, and particularly relates to an automatic tumor image region segmentation system and method based on an improved level set.
Background
At present, in the process of diagnosis and treatment of tumors, compared with some clinical analysis modes, the medical imaging technology can quantify tumor heterogeneity by extracting massive features from tumor images, so that quantitative analysis of tumor image data is achieved noninvasively. With the continuous increase of the data volume of clinical images, the development of a data-driven segmentation model is of great significance for the diagnosis and treatment of tumors. There are two main categories of traditional tumor diagnosis methods: 1. the radiologist judges the type and the malignancy degree of the tumor according to personal experience by checking the nuclear magnetic resonance image. 2. A small amount of tumor tissue is extracted through a puncture surgery, and then diagnosis is made by using various pathological analysis methods, gene detection and other means. However, there is great uncertainty about the way in which a radiologist makes a decision by viewing a medical image. Since doctors need to read a large number of images every day, the judgment result is easily interfered by external and internal factors, and a high misjudgment rate is caused to a great extent; the method of extracting a small amount of tumor tissues through the puncture operation to carry out pathological examination can obtain diagnosis with high accuracy, but the puncture operation inevitably causes certain trauma to the patient, which can increase the pain of the patient, and compared with the imaging examination, the method has long time required by the puncture examination and cannot be popularized on a large scale.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) there is great uncertainty in the way radiologists make their decisions by looking at medical images. Since doctors need to read a large number of images every day, the judgment result is easily interfered by external and internal factors, and a high misjudgment rate is caused to a great extent.
(2) The method of extracting a small amount of tumor tissues through the puncture operation to carry out pathological examination can obtain diagnosis with high accuracy, but the puncture operation inevitably causes certain trauma to the patient, which can increase the pain of the patient, and compared with the imaging examination, the method has long time required by the puncture examination and cannot be popularized on a large scale.
Disclosure of Invention
In view of the problems of the prior art, the present invention provides an automatic tumor image region segmentation system and method based on an improved level set.
The invention is realized by an automatic tumor image region segmentation method based on an improved level set, which comprises the following steps:
acquiring original tumor image data to be segmented by using an image acquisition device through a tumor image acquisition module;
step two, carrying out denoising, cutting and enhancing processing on the acquired original tumor image data to be segmented by an image preprocessing module by utilizing an image preprocessing program;
the denoising, cutting and enhancing treatment of the acquired original tumor image data to be segmented comprises the following steps:
(2.1) carrying out multi-scale decomposition on the acquired original tumor image data by utilizing an isotropic thermal diffusion equation;
(2.2) determining a denoising intensity coefficient, and denoising and controlling denoising intensity of the tumor image data in each scale by utilizing a penalty weighted least squares algorithm based on the determined denoising intensity coefficient;
(2.3) smoothing, enhancing and compensating the denoised residual edge; sequentially traversing the scales, and fusing the compensated tumor image data of all scales to obtain complete enhanced tumor image data;
(2.4) dividing a focus area and a background area of the enhanced tumor image, and cutting an edge background area;
thirdly, controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set by using a central processing module and a central processor;
fourthly, performing region division on the processed tumor image by using a region division program through an image region division module, and dividing the tumor image into corresponding target regions;
fifthly, extracting high-dimensional quantitative features of each target area of the tumor image by using a feature extraction program and a neural network model through an image feature extraction module;
step six, fusing the extracted high-dimensional quantitative characteristics of each target area by using a characteristic fusion program through an image characteristic fusion module to obtain tumor image fusion characteristics;
analyzing the tumor state by using an image data analysis module according to the extracted high-dimensional quantitative characteristics of each target region of the tumor image by using a data analysis program;
the analyzing the tumor state comprises:
(a) obtaining a first classification probability according to tumor image fusion characteristics through a tumor data analysis module, and taking the first classification probability as a first classification dimension;
(b) generating a plurality of decision trees by using a data analysis program based on a random forest algorithm, obtaining a second classification probability according to the decision trees, and taking the second classification probability as a second classification dimension;
(c) taking the first classification dimension and the second classification dimension as a first two-dimensional feature, obtaining a third classification probability based on a two-dimensional SVM algorithm, and taking the third classification probability as a first classification result;
step eight, storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result by using a cloud database server through a cloud storage module;
and step nine, displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result by using a display through a display module.
Further, in the first step, the tumor image includes a two-dimensional tumor image and a three-dimensional tumor image.
Further, in step four, the method for performing region segmentation on the processed tumor image by the image region segmentation module includes:
(1) constructing a hypergraph according to the processed tumor images, and determining the tumor images to be an initial zero level set;
(2) performing a modified level set approach on the initial zero level set to determine a tumor region;
(3) performing edge smoothing on the tumor region according to a morphological operation.
Further, the performing the improved level set method on the initial zero level set to determine a tumor region comprises:
1) adding the gradient field information of the preprocessed tumor image to be segmented into the initial zero level set to construct a new evolution equation;
2) performing downsampling on a coarse tumor region in the tumor image at a resolution that is the same as a resolution of the original tumor image to obtain an initial zero level set;
3) and carrying out multiple iterations on the initial zero level set on the evolution equation according to a finite difference method to determine a tumor region.
Further, in step five, the method for extracting high-dimensional quantitative features of each target region of the tumor image includes:
(I) dividing the target area into a set of element graphs with preset sizes by an image feature extraction module;
(II) respectively acquiring feature maps of corresponding dimensions of corresponding target regions by using a feature extraction program based on the convolution layers in the neural network model according to the element map set;
(III) acquiring a feature subset from the feature map corresponding to each dimension of the target region through a pooling layer of the neural network model, and using the feature subset as a high-dimensional quantitative feature of the tumor image.
Further, in the sixth step, the method for fusing the extracted high-dimensional quantitative features of each target region includes:
and inputting the high-dimensional quantitative characteristics of each target area into a full-connection network layer as an input value, and carrying out nonlinear combination on the high-dimensional quantitative characteristics of each target area through a multilayer full-connection network and setting an activation function of a full-connection output node as a nonlinear function to obtain tumor image fusion characteristics.
Further, each of the decision trees includes a number of the features that are randomly selected.
Another object of the present invention is to provide an improved level set-based automatic tumor image region segmentation system for implementing the improved level set-based automatic tumor image region segmentation method, which includes:
the tumor image processing system comprises a tumor image acquisition module, an image preprocessing module, a central processing module, an image region segmentation module, an image feature extraction module, an image feature fusion module, an image data analysis module, a cloud storage module and a display module.
The tumor image acquisition module is connected with the central control module and used for acquiring original tumor image data to be segmented through the image acquisition device; the tumor image comprises a two-dimensional tumor image and a three-dimensional tumor image;
the image preprocessing module is connected with the central control module and is used for carrying out denoising, cutting and enhancing processing on the acquired original tumor image data to be segmented through an image preprocessing program;
the central processing module is connected with the tumor image acquisition module, the image preprocessing module, the image region segmentation module, the image feature extraction module, the image feature fusion module, the image data analysis module, the cloud storage module and the display module and is used for controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set through a central processing unit;
the image area segmentation module is connected with the central control module and is used for carrying out area segmentation on the processed tumor image through an area segmentation program and segmenting the tumor image into corresponding target areas;
the image feature extraction module is connected with the central control module and used for extracting high-dimensional quantitative features of each target area of the tumor image by utilizing a neural network model through a feature extraction program;
the image feature fusion module is connected with the central control module and used for fusing the extracted high-dimensional quantitative features of each target area through a feature fusion program to obtain tumor image fusion features;
the data analysis module is connected with the central control module and is used for analyzing the tumor state according to the extracted high-dimensional quantitative characteristics of each target area of the tumor image through a data analysis program;
the cloud storage module is connected with the central control module and used for storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result through the cloud database server;
and the display module is connected with the central control module and used for displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result through the display.
It is another object of the present invention to provide a computer program product stored on a computer readable medium, comprising a computer readable program for providing a user input interface to implement the method for automatic segmentation of tumor image regions based on an improved level set when executed on an electronic device.
It is another object of the present invention to provide a computer-readable storage medium storing instructions which, when executed on a computer, cause the computer to perform the method for automatic segmentation of tumor image regions based on an improved level set.
By combining all the technical schemes, the invention has the advantages and positive effects that: according to the automatic tumor image region segmentation system based on the improved level set, the neural network model is set into the multi-branch neural network structure corresponding to the dimensionality of the tumor image through the image feature extraction module, the multi-branch neural network structure is used for extracting the tumor image features of the tumor image with the corresponding dimensionality, and the two-dimensional tumor features and the three-dimensional tumor features of the tumor image can be captured simultaneously; and finally, classifying the classification results obtained by the two classes of algorithms by using a two-dimensional SVM algorithm to obtain a third classification probability, so that the classification accuracy is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
Fig. 1 is a flowchart of an automatic tumor image region segmentation method based on an improved level set according to an embodiment of the present invention.
Fig. 2 is a flowchart of a method for performing region segmentation on a processed tumor image by an image region segmentation module according to an embodiment of the present invention.
Fig. 3 is a flowchart of a method for extracting high-dimensional quantitative features of each target region of a tumor image according to an embodiment of the present invention.
Fig. 4 is a flowchart of a method for analyzing a tumor status by a tumor data analysis module according to an embodiment of the present invention.
FIG. 5 is a block diagram of an embodiment of an automatic tumor image region segmentation system based on an improved level set;
in the figure: 1. a tumor image acquisition module; 2. an image preprocessing module; 3. a central processing module; 4. an image area segmentation module; 5. an image feature extraction module; 6. an image feature fusion module; 7. an image data analysis module; 8. a cloud storage module; 9. and a display module.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In view of the problems of the prior art, the present invention provides an automatic tumor image region segmentation system and method based on an improved level set, which will be described in detail below with reference to the accompanying drawings.
As shown in fig. 1, an embodiment of the present invention provides an automatic tumor image region segmentation method based on an improved level set, which includes the following steps:
s101, acquiring original tumor image data to be segmented by using an image acquisition device through a tumor image acquisition module; the tumor image includes a two-dimensional tumor image and a three-dimensional tumor image.
S102, denoising, cutting and enhancing the acquired original tumor image data to be segmented by using an image preprocessing program through an image preprocessing module.
And S103, controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set by using a central processing module and a central processor.
And S104, performing region division on the processed tumor image by using a region division program through an image region division module, and dividing the tumor image into corresponding target regions.
And S105, extracting the high-dimensional quantitative features of each target area of the tumor image by using a feature extraction program and a neural network model through an image feature extraction module.
And S106, fusing the extracted high-dimensional quantitative characteristics of each target area by using a characteristic fusion program through an image characteristic fusion module to obtain tumor image fusion characteristics.
And S107, analyzing the tumor state by using the image data analysis module and a data analysis program according to the extracted high-dimensional quantitative characteristics of each target area of the tumor image.
And S108, storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result by using the cloud database server through the cloud storage module.
And S109, displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result by using a display through a display module.
In step S102, the denoising, clipping, and enhancing processing performed on the acquired original tumor image data to be segmented provided by the embodiment of the present invention includes:
(2.1) carrying out multi-scale decomposition on the acquired original tumor image data by utilizing an isotropic thermal diffusion equation;
(2.2) determining a denoising intensity coefficient, and denoising and controlling denoising intensity of the tumor image data in each scale by utilizing a penalty weighted least squares algorithm based on the determined denoising intensity coefficient;
(2.3) smoothing, enhancing and compensating the denoised residual edge; sequentially traversing the scales, and fusing the compensated tumor image data of all scales to obtain complete enhanced tumor image data;
and (2.4) dividing the focus area and the background area of the enhanced tumor image, and cutting the edge background area.
As shown in fig. 2, in step S104, the method for performing region segmentation on the processed tumor image by the image region segmentation module according to the embodiment of the present invention includes:
s201, constructing a hypergraph according to the processed tumor images, and determining the tumor images to be an initial zero level set.
S202, performing a modified level set method on the initial zero level set to determine a tumor region.
And S203, performing edge smoothing processing on the tumor region according to morphological operation.
The embodiment of the invention provides a method for performing improved level set on an initial zero level set so as to determine a tumor region, which comprises the following steps:
1) and adding the gradient field information of the preprocessed tumor image to be segmented into the initial zero level set to construct a new evolution equation.
2) Performing downsampling on a coarse tumor region in the tumor image at a resolution that is the same as a resolution of the original tumor image to obtain an initial zero level set.
3) And carrying out multiple iterations on the initial zero level set on the evolution equation according to a finite difference method to determine a tumor region.
As shown in fig. 3, in step S105, the method for extracting high-dimensional quantitative features of each target region of a tumor image according to an embodiment of the present invention includes:
s301, the target area is divided into a set of element images with preset sizes through an image feature extraction module.
S302, respectively obtaining feature maps corresponding to the dimensions of the target area by utilizing a feature extraction program based on the convolution layer in the neural network model according to the element map set.
And S303, acquiring a feature subset from the feature map corresponding to each dimension of the target region through a pooling layer of the neural network model, and taking the feature subset as a high-dimensional quantitative feature of the tumor image.
In step S106, the method for fusing the extracted high-dimensional quantitative features of each target region provided in the embodiment of the present invention includes:
and inputting the high-dimensional quantitative characteristics of each target area into a full-connection network layer as an input value, and carrying out nonlinear combination on the high-dimensional quantitative characteristics of each target area through a multilayer full-connection network and setting an activation function of a full-connection output node as a nonlinear function to obtain tumor image fusion characteristics.
As shown in fig. 4, in step S107, the method for analyzing a tumor state by a tumor data analysis module according to an embodiment of the present invention includes:
s401, obtaining a first classification probability according to tumor image fusion characteristics through a tumor data analysis module, and taking the first classification probability as a first classification dimension.
S402, generating a plurality of decision trees based on a random forest algorithm by using a data analysis program, obtaining a second classification probability according to the decision trees, and taking the second classification probability as a second classification dimensionality.
S403, taking the first classification dimension and the second classification dimension as a first two-dimensional feature, obtaining a third classification probability based on a two-dimensional SVM algorithm, and taking the third classification probability as a first classification result.
Each decision tree provided by the embodiment of the invention comprises a plurality of randomly selected features.
As shown in fig. 5, an embodiment of the present invention provides an automatic tumor image region segmentation system based on an improved level set, which includes: the system comprises a tumor image acquisition module 1, an image preprocessing module 2, a central processing module 3, an image region segmentation module 4, an image feature extraction module 5, an image feature fusion module 6, an image data analysis module 7, a cloud storage module 8 and a display module 9.
The tumor image acquisition module 1 is connected with the central control module 3 and used for acquiring original tumor image data to be segmented through an image acquisition device; the tumor image comprises a two-dimensional tumor image and a three-dimensional tumor image;
the image preprocessing module 2 is connected with the central control module 3 and is used for carrying out denoising, cutting and enhancing processing on the acquired original tumor image data to be segmented through an image preprocessing program;
the central processing module 3 is connected with the tumor image acquisition module 1, the image preprocessing module 2, the image region segmentation module 4, the image feature extraction module 5, the image feature fusion module 6, the image data analysis module 7, the cloud storage module 8 and the display module 9, and is used for controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set through a central processing unit;
the image area segmentation module 4 is connected with the central control module 3 and is used for carrying out area division on the processed tumor image through an area segmentation program and segmenting the tumor image into corresponding target areas;
the image feature extraction module 5 is connected with the central control module 3 and used for extracting high-dimensional quantitative features of each target area of the tumor image by utilizing a neural network model through a feature extraction program;
the image feature fusion module 6 is connected with the central control module 3 and is used for fusing the extracted high-dimensional quantitative features of each target area through a feature fusion program to obtain tumor image fusion features;
the data analysis module 7 is connected with the central control module 3 and is used for analyzing the tumor state according to the extracted high-dimensional quantitative characteristics of each target area of the tumor image through a data analysis program;
the cloud storage module 8 is connected with the central control module 3 and used for storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result through a cloud database server;
and the display module 9 is connected with the central control module 3 and is used for displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result through a display.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. An improved level set based automatic tumor image region segmentation method, comprising the following steps:
acquiring original tumor image data to be segmented by using an image acquisition device through a tumor image acquisition module;
step two, carrying out denoising, cutting and enhancing processing on the acquired original tumor image data to be segmented by an image preprocessing module by utilizing an image preprocessing program;
the denoising, cutting and enhancing treatment of the acquired original tumor image data to be segmented comprises the following steps:
(2.1) carrying out multi-scale decomposition on the acquired original tumor image data by utilizing an isotropic thermal diffusion equation;
(2.2) determining a denoising intensity coefficient, and denoising and controlling denoising intensity of the tumor image data in each scale by utilizing a penalty weighted least squares algorithm based on the determined denoising intensity coefficient;
(2.3) smoothing, enhancing and compensating the denoised residual edge; sequentially traversing the scales, and fusing the compensated tumor image data of all scales to obtain complete enhanced tumor image data;
(2.4) dividing a focus area and a background area of the enhanced tumor image, and cutting an edge background area;
thirdly, controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set by using a central processing module and a central processor;
fourthly, performing region division on the processed tumor image by using a region division program through an image region division module, and dividing the tumor image into corresponding target regions;
fifthly, extracting high-dimensional quantitative features of each target area of the tumor image by using a feature extraction program and a neural network model through an image feature extraction module;
step six, fusing the extracted high-dimensional quantitative characteristics of each target area by using a characteristic fusion program through an image characteristic fusion module to obtain tumor image fusion characteristics;
analyzing the tumor state by using an image data analysis module according to the extracted high-dimensional quantitative characteristics of each target region of the tumor image by using a data analysis program;
the analyzing the tumor state comprises:
(a) obtaining a first classification probability according to tumor image fusion characteristics through a tumor data analysis module, and taking the first classification probability as a first classification dimension;
(b) generating a plurality of decision trees by using a data analysis program based on a random forest algorithm, obtaining a second classification probability according to the decision trees, and taking the second classification probability as a second classification dimension;
(c) taking the first classification dimension and the second classification dimension as a first two-dimensional feature, obtaining a third classification probability based on a two-dimensional SVM algorithm, and taking the third classification probability as a first classification result;
step eight, storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result by using a cloud database server through a cloud storage module;
and step nine, displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result by using a display through a display module.
2. The method according to claim 1, wherein in step one, the tumor image comprises a two-dimensional tumor image and a three-dimensional tumor image.
3. The method according to claim 1, wherein in step four, the method for performing region segmentation on the processed tumor image by the image region segmentation module comprises:
(1) constructing a hypergraph according to the processed tumor images, and determining the tumor images to be an initial zero level set;
(2) performing a modified level set approach on the initial zero level set to determine a tumor region;
(3) performing edge smoothing on the tumor region according to a morphological operation.
4. The method of claim 3, wherein the performing the modified level set method on the initial zero level set to determine the tumor region comprises:
1) adding the gradient field information of the preprocessed tumor image to be segmented into the initial zero level set to construct a new evolution equation;
2) performing downsampling on a coarse tumor region in the tumor image at a resolution that is the same as a resolution of the original tumor image to obtain an initial zero level set;
3) and carrying out multiple iterations on the initial zero level set on the evolution equation according to a finite difference method to determine a tumor region.
5. The method according to claim 1, wherein in step five, the method for extracting high-dimensional quantitative features of each target region of the tumor image comprises:
(I) dividing the target area into a set of element graphs with preset sizes by an image feature extraction module;
(II) respectively acquiring feature maps of corresponding dimensions of corresponding target regions by using a feature extraction program based on the convolution layers in the neural network model according to the element map set;
(III) acquiring a feature subset from the feature map corresponding to each dimension of the target region through a pooling layer of the neural network model, and using the feature subset as a high-dimensional quantitative feature of the tumor image.
6. The method for automatic segmentation of tumor image regions based on improved level set according to claim 1, wherein in step six, the method for fusing the extracted high-dimensional quantitative features of each target region comprises:
and inputting the high-dimensional quantitative characteristics of each target area into a full-connection network layer as an input value, and carrying out nonlinear combination on the high-dimensional quantitative characteristics of each target area through a multilayer full-connection network and setting an activation function of a full-connection output node as a nonlinear function to obtain tumor image fusion characteristics.
7. The method according to claim 1, wherein each of the decision trees comprises a plurality of randomly selected features.
8. An improved level set based automatic tumor image region segmentation system, comprising:
the system comprises a tumor image acquisition module, an image preprocessing module, a central processing module, an image region segmentation module, an image feature extraction module, an image feature fusion module, an image data analysis module, a cloud storage module and a display module;
the tumor image acquisition module is connected with the central control module and used for acquiring original tumor image data to be segmented through the image acquisition device; the tumor image comprises a two-dimensional tumor image and a three-dimensional tumor image;
the image preprocessing module is connected with the central control module and is used for carrying out denoising, cutting and enhancing processing on the acquired original tumor image data to be segmented through an image preprocessing program;
the central processing module is connected with the tumor image acquisition module, the image preprocessing module, the image region segmentation module, the image feature extraction module, the image feature fusion module, the image data analysis module, the cloud storage module and the display module and is used for controlling the work of each module of the automatic tumor image region segmentation system based on the improved level set through a central processing unit;
the image area segmentation module is connected with the central control module and is used for carrying out area segmentation on the processed tumor image through an area segmentation program and segmenting the tumor image into corresponding target areas;
the image feature extraction module is connected with the central control module and used for extracting high-dimensional quantitative features of each target area of the tumor image by utilizing a neural network model through a feature extraction program;
the image feature fusion module is connected with the central control module and used for fusing the extracted high-dimensional quantitative features of each target area through a feature fusion program to obtain tumor image fusion features;
the data analysis module is connected with the central control module and is used for analyzing the tumor state according to the extracted high-dimensional quantitative characteristics of each target area of the tumor image through a data analysis program;
the cloud storage module is connected with the central control module and used for storing the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the tumor state analysis result through the cloud database server;
and the display module is connected with the central control module and used for displaying the acquired original tumor image data to be segmented, the target area, the high-dimensional quantitative characteristics and the real-time data of the tumor state analysis result through the display.
9. A computer program product stored on a computer readable medium, comprising a computer readable program for providing a user input interface for implementing the method of automatic tumor image region segmentation based on an improved level set according to any one of claims 2 to 8 when executed on an electronic device.
10. A computer-readable storage medium storing instructions which, when executed on a computer, cause the computer to perform the method for automatic segmentation of tumor image regions based on improved level set according to any one of claims 2 to 8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110084653.2A CN112837324A (en) | 2021-01-21 | 2021-01-21 | Automatic tumor image region segmentation system and method based on improved level set |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110084653.2A CN112837324A (en) | 2021-01-21 | 2021-01-21 | Automatic tumor image region segmentation system and method based on improved level set |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112837324A true CN112837324A (en) | 2021-05-25 |
Family
ID=75929317
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110084653.2A Pending CN112837324A (en) | 2021-01-21 | 2021-01-21 | Automatic tumor image region segmentation system and method based on improved level set |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112837324A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113838071A (en) * | 2021-09-29 | 2021-12-24 | 杭州海心智医信息科技有限公司 | Tumor feature extraction device |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104715484A (en) * | 2015-03-20 | 2015-06-17 | 中国科学院自动化研究所 | Automatic tumor area partition method based on improved level set |
CN106157275A (en) * | 2015-04-21 | 2016-11-23 | 杭州与盟医疗技术有限公司 | Multiple dimensioned CT Image denoising algorithm based on thermic vibrating screen |
CN108538369A (en) * | 2018-03-20 | 2018-09-14 | 中南大学湘雅医院 | Method for analyzing central nervous system tumor image data |
CN109493325A (en) * | 2018-10-23 | 2019-03-19 | 清华大学 | Tumor Heterogeneity analysis system based on CT images |
CN109635846A (en) * | 2018-11-16 | 2019-04-16 | 哈尔滨工业大学(深圳) | A kind of multiclass medical image judgment method and system |
CN110427954A (en) * | 2019-07-26 | 2019-11-08 | 中国科学院自动化研究所 | The image group feature extracting method of multizone based on tumor imaging |
CN111210441A (en) * | 2020-01-02 | 2020-05-29 | 苏州瑞派宁科技有限公司 | Tumor prediction method and device, cloud platform and computer-readable storage medium |
-
2021
- 2021-01-21 CN CN202110084653.2A patent/CN112837324A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104715484A (en) * | 2015-03-20 | 2015-06-17 | 中国科学院自动化研究所 | Automatic tumor area partition method based on improved level set |
CN106157275A (en) * | 2015-04-21 | 2016-11-23 | 杭州与盟医疗技术有限公司 | Multiple dimensioned CT Image denoising algorithm based on thermic vibrating screen |
CN108538369A (en) * | 2018-03-20 | 2018-09-14 | 中南大学湘雅医院 | Method for analyzing central nervous system tumor image data |
CN109493325A (en) * | 2018-10-23 | 2019-03-19 | 清华大学 | Tumor Heterogeneity analysis system based on CT images |
CN109635846A (en) * | 2018-11-16 | 2019-04-16 | 哈尔滨工业大学(深圳) | A kind of multiclass medical image judgment method and system |
CN110427954A (en) * | 2019-07-26 | 2019-11-08 | 中国科学院自动化研究所 | The image group feature extracting method of multizone based on tumor imaging |
CN111210441A (en) * | 2020-01-02 | 2020-05-29 | 苏州瑞派宁科技有限公司 | Tumor prediction method and device, cloud platform and computer-readable storage medium |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113838071A (en) * | 2021-09-29 | 2021-12-24 | 杭州海心智医信息科技有限公司 | Tumor feature extraction device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11935644B2 (en) | Deep learning automated dermatopathology | |
US10853409B2 (en) | Systems and methods for image search | |
CN107665491A (en) | The recognition methods of pathological image and system | |
NL2003805A (en) | Systems, apparatus and processes for automated medical image segmentation using a statistical model. | |
CN116433605A (en) | Medical image analysis mobile augmented reality system and method based on cloud intelligence | |
CN112070785A (en) | Medical image analysis method based on computer vision | |
Aygüneş et al. | Graph convolutional networks for region of interest classification in breast histopathology | |
CN113781387A (en) | Model training method, image processing method, device, equipment and storage medium | |
JP2023504875A (en) | Computer vision-based vessel feature acquisition method, intelligent microscope, vessel tissue feature acquisition device, computer program, and computer equipment | |
Nair et al. | Prediction and Classification of CT images for Early Detection of Lung Cancer Using Various Segmentation Models | |
CN112837324A (en) | Automatic tumor image region segmentation system and method based on improved level set | |
CN113822846A (en) | Method, apparatus, device and medium for determining region of interest in medical image | |
CN111598144B (en) | Training method and device for image recognition model | |
Wu et al. | Semiautomatic segmentation of glioma on mobile devices | |
US11615895B2 (en) | Systems and methods for automatically managing image data | |
Lin et al. | Image segmentation by improved minimum spanning tree with fractional differential and Canny detector | |
Wajeed et al. | A Breast Cancer Image Classification Algorithm with 2c Multiclass Support Vector Machine | |
CN114241198A (en) | Method, device, equipment and storage medium for obtaining local imagery omics characteristics | |
CN114170415A (en) | TMB classification method and system based on histopathology image depth domain adaptation | |
Romo-Bucheli et al. | Nuclei graph local features for basal cell carcinoma classification in whole slide images | |
Mohideen et al. | A guided ant colony algorithm for detection of microcalcification clusters | |
Mansour | Multimodal biomedical image retrieval and indexing system using handcrafted with deep convolution neural network features | |
Qi et al. | Discrete Wavelet Transform-Based CNN for Breast Cancer Classification from Histopathology Images | |
Rahman et al. | ADBNet: An Attention-Guided Deep Broad Convolutional Neural Network for the Classification of Breast Cancer Histopathology Images | |
EP3776475A1 (en) | Methods of generating an encoded representation of an image and systems of operating thereof |
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 |