CN117078696A - Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium - Google Patents
Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN117078696A CN117078696A CN202311047266.7A CN202311047266A CN117078696A CN 117078696 A CN117078696 A CN 117078696A CN 202311047266 A CN202311047266 A CN 202311047266A CN 117078696 A CN117078696 A CN 117078696A
- Authority
- CN
- China
- Prior art keywords
- dimensional medical
- image
- medical image
- target
- images
- 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
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000003709 image segmentation Methods 0.000 claims abstract description 75
- 238000012545 processing Methods 0.000 claims abstract description 24
- 230000011218 segmentation Effects 0.000 claims abstract description 21
- 230000001186 cumulative effect Effects 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 9
- 241001270131 Agaricus moelleri Species 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 2
- 239000013598 vector Substances 0.000 description 9
- 238000010586 diagram Methods 0.000 description 6
- 238000003062 neural network model Methods 0.000 description 5
- 230000004927 fusion Effects 0.000 description 4
- 238000002059 diagnostic imaging Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 238000011946 reduction process Methods 0.000 description 2
- 210000004872 soft tissue Anatomy 0.000 description 2
- 108010038083 amyloid fibril protein AS-SAM Proteins 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000002583 angiography Methods 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000002591 computed tomography Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000013421 nuclear magnetic resonance imaging Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The application provides a method and a device for segmenting a three-dimensional medical image, electronic equipment and a storage medium. The method comprises the steps of acquiring a three-dimensional medical image; performing dimension reduction processing on the three-dimensional medical images to generate a plurality of two-dimensional medical images; and performing image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by using an image segmentation model. According to the method, the dimension reduction processing is carried out on the three-dimensional medical image to generate a plurality of two-dimensional medical images, and the complexity of the two-dimensional medical image relative to the three-dimensional medical image is greatly reduced, so that the accuracy of a segmentation result can be improved when the three-dimensional medical image is segmented through the segmentation of the two-dimensional medical image.
Description
Technical Field
The present application relates to the field of medical technology, and in particular, to a method and apparatus for segmenting three-dimensional medical images, an electronic device, and a storage medium.
Background
In the medical technical field, the segmentation of a three-dimensional medical image is a complex and key step in the medical image processing and analyzing process, and the purpose of the segmentation is to segment parts with certain special meanings in the three-dimensional medical image, such as the segmentation of the three-dimensional medical image of an organ, so as to obtain an image of a lesion part in the organ, thereby providing a reliable basis for clinical diagnosis and treatment.
At present, a deep neural network model is generally used for dividing a three-dimensional medical image, but due to the fact that the complexity of the three-dimensional medical image is high, the accuracy of a division result is poor in the manner of dividing the three-dimensional medical image by the deep neural network model.
Disclosure of Invention
The embodiment of the application aims to provide a method, a device, electronic equipment and a storage medium for segmenting a three-dimensional medical image, which are used for solving the problem of poor accuracy of a segmentation result of the three-dimensional medical image in the prior art.
An embodiment of the present application provides a method for segmenting a three-dimensional medical image, including:
acquiring a three-dimensional medical image;
performing dimension reduction processing on the three-dimensional medical images to generate a plurality of two-dimensional medical images;
and performing image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by using an image segmentation model.
In one embodiment, the image segmentation model is used to segment at least one two-dimensional medical image of the plurality of two-dimensional medical images, and specifically includes:
screening at least one two-dimensional medical image from the plurality of two-dimensional medical images to serve as a target two-dimensional medical image; the definition of the target two-dimensional medical image is larger than a preset threshold;
and performing image segmentation on the target two-dimensional medical image by using the image segmentation model.
In an embodiment, before image segmentation of the target two-dimensional medical image using the image segmentation model, the method further comprises:
performing image enhancement on the target two-dimensional medical image; the method comprises the steps of,
the image segmentation model is utilized to segment the target two-dimensional medical image, and the method specifically comprises the following steps: and performing image segmentation on the target two-dimensional medical image after image enhancement by using the image segmentation model.
In one embodiment, the image enhancement of the target two-dimensional medical image specifically includes: and carrying out image enhancement on the target two-dimensional medical image in a histogram equalization mode.
In an embodiment, the image enhancement of the target two-dimensional medical image by means of histogram equalization specifically includes:
dividing a plurality of gray scales;
counting the number of pixels in each gray level for each pixel in the target two-dimensional medical image;
according to the number of pixels in each gray level, calculating the duty ratio of the pixels in each gray level to obtain probability distribution of the pixels in each gray level;
according to the probability distribution of each gray level, calculating the cumulative probability distribution of each gray level;
according to the cumulative probability distribution of each gray level, calculating the gray value of each gray level after equalization;
and mapping the gray values obtained by equalizing the gray levels to pixels in the corresponding gray levels so as to realize image enhancement of the target two-dimensional medical image.
In one embodiment, the image segmentation model specifically includes a SAM model; the method comprises the steps of,
the image segmentation model is utilized to segment the target two-dimensional medical image, and the method specifically comprises the following steps: and performing image segmentation on the target two-dimensional medical image by using the SAM model.
In one embodiment, the dimension reduction processing is performed on the three-dimensional medical image to generate a plurality of two-dimensional medical images, which specifically includes:
reading the three-dimensional medical image by using Python;
setting a window for the three-dimensional medical image, and setting a window width and a window level of the window, wherein the window width refers to a pixel range of an image of a target area in the three-dimensional medical image observed by the window, and the window level refers to an average number of an upper limit and a lower limit of pixels in the target area;
and respectively intercepting images of each target area in the three-dimensional medical image observed by the window by moving the window for a plurality of times so as to generate corresponding two-dimensional medical images through the respectively intercepted images.
A second aspect of an embodiment of the present application provides a three-dimensional medical image segmentation apparatus, including:
the acquisition unit is used for acquiring the three-dimensional medical image;
the dimension reduction processing unit is used for generating a plurality of two-dimensional medical images by dimension reduction processing on the three-dimensional medical images;
and the segmentation unit is used for carrying out image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by utilizing an image segmentation model.
A third aspect of an embodiment of the present application provides an electronic device, including:
a memory for storing a computer program;
a processor configured to perform the method according to any one of the embodiments of the method of the present application.
A fourth aspect of an embodiment of the present application provides a storage medium, including: a program which, when run on an electronic device, causes the electronic device to perform the method according to any one of the method embodiments of the application.
The method for segmenting the three-dimensional medical image comprises the steps of firstly obtaining the three-dimensional medical image, then performing dimension reduction processing on the three-dimensional medical image to generate a plurality of two-dimensional medical images, and then performing image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by using an image segmentation model. According to the method, the dimension reduction processing is carried out on the three-dimensional medical image to generate a plurality of two-dimensional medical images, and the complexity of the two-dimensional medical image relative to the three-dimensional medical image is greatly reduced, so that the accuracy of a segmentation result can be improved when the three-dimensional medical image is segmented through the segmentation of the two-dimensional medical image.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application;
FIG. 2 is a schematic diagram illustrating interaction between an electronic device and a client according to an embodiment of the present application;
FIG. 3 is a flowchart illustrating a method for segmenting a three-dimensional medical image according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a SAM model according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of a three-dimensional medical image segmentation apparatus according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application. In the description of the present application, terms such as "first," "second," "third," and the like are used merely to distinguish between descriptions and are not to be construed as indicating or implying relative importance or order.
As described above, in the medical technical field, the segmentation of the three-dimensional medical image is a complex and critical step in the medical image processing and analysis process, and the three-dimensional medical image is usually segmented by using the deep neural network model at present, but the segmentation method using the deep neural network model is poor in accuracy due to the high complexity of the three-dimensional medical image.
In view of the above, the embodiments of the present application provide a method, an apparatus, an electronic device, and a storage medium for segmenting a three-dimensional medical image, which can be used to solve the technical problem. As shown in fig. 1, which is a schematic diagram of a specific structure of an electronic device provided in this embodiment, the electronic device 1 includes: at least one processor 11 and a memory 12, one processor being exemplified in fig. 1. The processor 11 and the memory 12 may be connected by a bus 10, the memory 12 storing instructions executable by the processor 11, the instructions being executable by the processor 11 to cause the electronic device 1 to perform all or part of the flow of the method in the embodiments described below. The electronic device 1 may be a mobile phone, a notebook computer, a desktop computer, or a large server or a server cluster formed by the mobile phone, the notebook computer, the desktop computer, or the large server or the server cluster.
As shown in fig. 2, the electronic device 1 may also be docked to one or more clients 2 as a server, in which case the electronic device 1 may be in data communication with each client 2 separately. The client 2 may be a user's mobile phone, a notebook computer, a desktop computer, etc.
In an embodiment, a user may send a three-dimensional medical image to the electronic device 1 as a server through the client 2, so that the electronic device 1 may receive the three-dimensional medical image, and the method provided by the embodiment of the present application is executed to realize the segmentation of the three-dimensional medical image.
Fig. 3 is a schematic flowchart of a method for segmenting a three-dimensional medical image according to an embodiment of the present application, and part or all of the steps of the method may be performed by the electronic device 1 shown in fig. 1 as a server. The method may be described specifically herein, and includes the steps of:
step S31: and acquiring a three-dimensional medical image.
The three-dimensional medical image may be an internal tissue image obtained by a medical imaging system in a non-invasive manner with respect to a human body or a part of a human body. Depending on the medical imaging system, the three-dimensional medical image may include angiography, computed tomography, nuclear magnetic resonance imaging, and the like.
For the specific implementation manner of acquiring the three-dimensional medical image in the step S31, for example, the three-dimensional medical image may be acquired, or the three-dimensional medical image may be acquired from a three-dimensional medical image library. For example, in the process of diagnosing a patient, the three-dimensional medical image may be directly acquired by the medical imaging system, or the acquired three-dimensional medical image may be stored in a three-dimensional medical image library, and then the three-dimensional medical image may be acquired from the three-dimensional medical image library.
In an embodiment, the specific implementation manner of the step S31 of acquiring the three-dimensional medical image during the patient 'S medical institution visit may be to acquire the three-dimensional medical image submitted during the patient' S visit.
It should be noted that, in another embodiment, the specific implementation manner of acquiring the three-dimensional medical image in the step S31 may also be that the three-dimensional medical image sent by the client is acquired, for example, when the patient performs online medical treatment, the three-dimensional medical image may be sent to the server through the client. At this time, the server may receive the three-dimensional medical image transmitted by the client. In addition, considering that the server may be connected with a plurality of clients, so that a plurality of patients can send three-dimensional medical images to the server through different clients respectively, at this time, the situation that the plurality of patients send three-dimensional medical images to the server through different clients respectively may occur in a short time, and further the operation pressure of the server is caused to be larger, for this situation, a message queue may be generally constructed on the server in advance, so after the different clients send three-dimensional medical images to the server respectively, the three-dimensional medical images may be added to the message queue according to the sequence of the three-dimensional medical images received by the server, and then according to the service processing situation of the server, the corresponding three-dimensional medical images are acquired from the message queue according to the sequence of the three-dimensional medical images in the message queue, thereby reducing the operation pressure of the server.
Step S32: the three-dimensional medical image is subjected to dimension reduction processing to generate a plurality of two-dimensional medical images.
In practical application, the three-dimensional medical image is different from a common image, for example, the three-dimensional medical image lacks simple linear characteristics, artifacts are easy to exist, the gray level between different soft tissues is approximate, and the like, so that the complexity of the three-dimensional medical image is higher than that of the common image, and the accuracy is easy to be poor when the three-dimensional medical image is directly segmented.
In the method provided by the embodiment of the application, the step S32 is utilized to perform dimension reduction processing on the three-dimensional medical image, so as to generate a plurality of two-dimensional medical images, and the complexity of the two-dimensional medical images is greatly reduced relative to the three-dimensional medical images, so that the two-dimensional medical images are convenient to divide.
In practical application, the specific implementation manner of performing the dimension reduction processing on the three-dimensional medical image may be that performing the dimension reduction processing on the three-dimensional medical image by using Python, for example, the three-dimensional medical image may be read by using Python first, then a Window is set for the three-dimensional medical image, and a Window Width (Window Width) and a Window Level (Window Level) of the Window are set, where the Window Width refers to a pixel range of an image of a target area in the three-dimensional medical image observed by the Window; window level refers to the average of the upper and lower limits of pixels in the target area; then, the window is moved for a plurality of times, and the images of all target areas in the three-dimensional medical image observed by the window are respectively intercepted, so that corresponding two-dimensional medical images are generated through the respectively intercepted images.
For example, an image of a target region observed by a window in the three-dimensional medical image may be first captured, and a two-dimensional medical image generated from the captured image; then the window can be moved in the three-dimensional medical image, then the image of the next target area is intercepted, and another two-dimensional medical image is generated through the intercepted image; by analogy, a plurality of two-dimensional medical images may be generated.
Step S33: image segmentation is performed on at least one two-dimensional medical image of the plurality of two-dimensional medical images using an image segmentation model.
After the three-dimensional medical image is subjected to the dimension reduction process to generate a plurality of two-dimensional medical images in step S32, at least one of the plurality of two-dimensional medical images may be subjected to image segmentation by using the image segmentation model, so that the at least one two-dimensional medical image is segmented by using the image segmentation model to realize the segmentation of the three-dimensional medical image. The image segmentation model can be a deep neural network model or other types of image segmentation models because the image segmentation model needs to be segmented into two-dimensional medical images.
It should be noted that, in view of the medical field, the segmentation of the two-dimensional medical image needs to be high to accuracy, so the image segmentation model may be a segmented arbitrary model (Segment Anything Model, hereinafter referred to as SAM model). As shown in fig. 4, the SAM model includes three modules, namely an image encoding module (image encoder), a prompt encoding module (prompt encoder), and a template decoding module (mask decoder).
The image encoding module is configured to map a two-dimensional medical image (i.e., an image in fig. 4) input to the SAM model to an image feature space, so as to obtain an image feature vector (image mapping) of the two-dimensional medical image.
The prompt encoding module is configured to map a prompt (prompt) input by a user to a prompt feature space, so as to obtain a prompt feature vector (prompt scrolling) of the prompt, where the prompt may include, for example, points, box, text. In addition, the prompt may be generally divided into a sparse prompt (sparse prompt) and a dense prompt (dense prompt), where the prompt encoding module can map the two types of prompts to the prompt feature space, so as to obtain corresponding prompt feature vectors.
The template decoding module can acquire the image feature vector and the prompt feature vector on the one hand, and fuse the image feature vector and the prompt feature vector to generate a fused feature vector, wherein the fusion mode can be weighted summation, inner product fusion or other fusion modes; on the other hand, the fusion feature vector may also be decoded, so as to segment the two-dimensional medical image, such as generating a mask in a region to be segmented in the two-dimensional medical image.
For example, for the specific implementation of the step S33, each two-dimensional medical image in the at least one two-dimensional medical image may be input to the SAM model, so that the input two-dimensional medical image is segmented by the SAM model.
The method for segmenting the three-dimensional medical image comprises the steps of firstly obtaining the three-dimensional medical image, then performing dimension reduction processing on the three-dimensional medical image to generate a plurality of two-dimensional medical images, and then performing image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by using an image segmentation model. According to the method, the dimension reduction processing is carried out on the three-dimensional medical image to generate a plurality of two-dimensional medical images, and the complexity of the two-dimensional medical image relative to the three-dimensional medical image is greatly reduced, so that the accuracy of a segmentation result can be improved when the three-dimensional medical image is segmented through the segmentation of the two-dimensional medical image.
In practical applications, considering that the three-dimensional medical images lack simple linear characteristics, artifacts are easy to exist, and gray scale approximations among different soft tissues are similar, after the above-mentioned step S32 is performed to generate a plurality of two-dimensional medical images, the two-dimensional medical images may also have artifacts or have poor definition due to gray scale approximations, so that it is generally required to screen out a part of the two-dimensional medical images with higher definition from the plurality of two-dimensional medical images, and then divide the screened part of the two-dimensional medical images.
Therefore, for the specific implementation manner of step S33, at least one two-dimensional medical image may be first selected from the plurality of two-dimensional medical images as a target two-dimensional medical image, where the sharpness of the target two-dimensional medical image is greater than a preset threshold; then, the image segmentation model is utilized to carry out image segmentation on the target two-dimensional medical image. The magnitude of the preset threshold value can be set according to actual needs.
In addition, in order to further improve the accuracy of the segmentation result, before the image segmentation is performed on the target two-dimensional medical image by using the image segmentation model, the method may further include performing image enhancement on the target two-dimensional medical image, so that the image segmentation is performed on the target two-dimensional medical image by using the image segmentation model, specifically, performing image segmentation on the target two-dimensional medical image after image enhancement by using the image segmentation model, so that the accuracy of the segmentation result can be further improved by performing image segmentation on the target two-dimensional medical image after image enhancement relative to performing image segmentation directly on the target two-dimensional medical image.
The image enhancement method for the target two-dimensional medical image can be various, for example, the image enhancement method for the target two-dimensional medical image can be realized by a histogram equalization method, specifically, a plurality of gray scales can be firstly divided, and then the number of pixels in each gray scale can be counted for each pixel in the target two-dimensional medical image; then, according to the number of pixels in each gray level, calculating the duty ratio of the pixels in each gray level, so as to obtain the probability distribution of the pixels in each gray level; then, according to the probability distribution of each gray level, calculating the cumulative probability distribution of each gray level, and further according to the cumulative probability distribution of each gray level, calculating the gray value of each gray level after equalization, wherein the gray value of a certain gray level after equalization=the cumulative probability distribution of the gray level (maximum gray level-1); after the gray value of each gray level after the equalization is obtained, the gray value is mapped to each pixel in the corresponding gray level, so that the image enhancement of the target two-dimensional medical image is realized, and the contrast of the image can be increased through the image enhancement, thereby facilitating the subsequent image segmentation.
Based on the same inventive concept as the method for segmenting a three-dimensional medical image provided by the embodiment of the present application, the embodiment of the present application also provides a device for segmenting a three-dimensional medical image, and for the device embodiment, reference may be made to the corresponding content of the method embodiment if it is unclear. As shown in fig. 5, which is a schematic structural diagram of the apparatus 40, the apparatus 40 includes: an acquisition unit 401, a dimension reduction processing unit 402, and a division unit 403, wherein:
an acquisition unit 401 for acquiring a three-dimensional medical image;
a dimension reduction processing unit 402, configured to generate a plurality of two-dimensional medical images by performing dimension reduction processing on the three-dimensional medical images;
a segmentation unit 403 for performing image segmentation on at least one of the plurality of two-dimensional medical images by using an image segmentation model
With the device 40 provided by the embodiment of the present application, since the device 40 adopts the same inventive concept as the method for segmenting a three-dimensional medical image provided by the embodiment of the present application, the device 40 can solve the technical problem on the premise that the method can solve the technical problem, and the description thereof is omitted herein.
In addition, in practical application, the technical effects obtained by combining the device 40 with specific hardware devices, cloud technology and the like are also within the protection scope of the present application, for example, different units in the device 40 are distributed in different nodes in the distributed cluster by adopting a distributed cluster manner, so as to improve efficiency and the like.
In practical application, the image segmentation of at least one two-dimensional medical image in the plurality of two-dimensional medical images by using the image segmentation model may specifically include: screening at least one two-dimensional medical image from the plurality of two-dimensional medical images to serve as a target two-dimensional medical image; the definition of the target two-dimensional medical image is larger than a preset threshold; and performing image segmentation on the target two-dimensional medical image by using the image segmentation model.
Before the image segmentation model is utilized to carry out image segmentation on the target two-dimensional medical image, the method can further comprise the step of carrying out image enhancement on the target two-dimensional medical image; and performing image segmentation on the target two-dimensional medical image by using the image segmentation model, wherein the method specifically comprises the following steps of: and performing image segmentation on the target two-dimensional medical image after image enhancement by using the image segmentation model.
The image enhancement of the target two-dimensional medical image may specifically include: and carrying out image enhancement on the target two-dimensional medical image in a histogram equalization mode.
The image enhancement of the target two-dimensional medical image by means of histogram equalization may specifically include:
dividing a plurality of gray scales;
counting the number of pixels in each gray level for each pixel in the target two-dimensional medical image;
according to the number of pixels in each gray level, calculating the duty ratio of the pixels in each gray level to obtain probability distribution of the pixels in each gray level;
according to the probability distribution of each gray level, calculating the cumulative probability distribution of each gray level;
according to the cumulative probability distribution of each gray level, calculating the gray value of each gray level after equalization;
and mapping the gray values obtained by equalizing the gray levels to pixels in the corresponding gray levels so as to realize image enhancement of the target two-dimensional medical image.
Wherein, the image segmentation model specifically comprises a SAM model; the method comprises the steps of,
the image segmentation of the target two-dimensional medical image by using the image segmentation model may specifically include: and performing image segmentation on the target two-dimensional medical image by using the SAM model.
The method according to claim 1, wherein the step of performing a dimension reduction process on the three-dimensional medical image to generate a plurality of two-dimensional medical images may specifically include: reading the three-dimensional medical image by using Python; setting a window for the three-dimensional medical image, and setting a window width and a window level of the window, wherein the window width refers to a pixel range of an image of a target area in the three-dimensional medical image observed by the window, and the window level refers to an average number of an upper limit and a lower limit of pixels in the target area; and respectively intercepting images of each target area in the three-dimensional medical image observed by the window by moving the window for a plurality of times so as to generate corresponding two-dimensional medical images through the respectively intercepted images.
The embodiment of the application also provides a storage medium, which comprises: a program which, when run on an electronic device, causes the electronic device to perform all or part of the flow of the method in the above-described embodiments. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a Flash Memory (Flash Memory), a Hard Disk (HDD), or a Solid State Drive (SSD), etc. The storage medium may also comprise a combination of memories of the kind described above.
Although embodiments of the present application have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the application, and such modifications and variations are within the scope of the application as defined by the appended claims.
Claims (10)
1. A method for segmenting a three-dimensional medical image, comprising:
acquiring a three-dimensional medical image;
performing dimension reduction processing on the three-dimensional medical images to generate a plurality of two-dimensional medical images;
and performing image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by using an image segmentation model.
2. The method of claim 1, wherein image segmentation of at least one of the plurality of two-dimensional medical images using an image segmentation model, comprises:
screening at least one two-dimensional medical image from the plurality of two-dimensional medical images to serve as a target two-dimensional medical image; the definition of the target two-dimensional medical image is larger than a preset threshold;
and performing image segmentation on the target two-dimensional medical image by using the image segmentation model.
3. The method of claim 2, wherein prior to image segmentation of the target two-dimensional medical image using the image segmentation model, the method further comprises:
performing image enhancement on the target two-dimensional medical image; the method comprises the steps of,
the image segmentation model is utilized to segment the target two-dimensional medical image, and the method specifically comprises the following steps: and performing image segmentation on the target two-dimensional medical image after image enhancement by using the image segmentation model.
4. A method according to claim 3, wherein image enhancement of the target two-dimensional medical image specifically comprises: and carrying out image enhancement on the target two-dimensional medical image in a histogram equalization mode.
5. The method according to claim 4, wherein the image enhancement of the target two-dimensional medical image by means of histogram equalization, in particular comprises:
dividing a plurality of gray scales;
counting the number of pixels in each gray level for each pixel in the target two-dimensional medical image;
according to the number of pixels in each gray level, calculating the duty ratio of the pixels in each gray level to obtain probability distribution of the pixels in each gray level;
according to the probability distribution of each gray level, calculating the cumulative probability distribution of each gray level;
according to the cumulative probability distribution of each gray level, calculating the gray value of each gray level after equalization;
and mapping the gray values obtained by equalizing the gray levels to pixels in the corresponding gray levels so as to realize image enhancement of the target two-dimensional medical image.
6. The method of claim 2, wherein the image segmentation model specifically comprises a SAM model; the method comprises the steps of,
the image segmentation model is utilized to segment the target two-dimensional medical image, and the method specifically comprises the following steps: and performing image segmentation on the target two-dimensional medical image by using the SAM model.
7. The method according to claim 1, wherein the three-dimensional medical image is subjected to dimension reduction processing to generate a plurality of two-dimensional medical images, and specifically comprises:
reading the three-dimensional medical image by using Python;
setting a window for the three-dimensional medical image, and setting a window width and a window level of the window, wherein the window width refers to a pixel range of an image of a target area in the three-dimensional medical image observed by the window, and the window level refers to an average number of an upper limit and a lower limit of pixels in the target area;
and respectively intercepting images of each target area in the three-dimensional medical image observed by the window by moving the window for a plurality of times so as to generate corresponding two-dimensional medical images through the respectively intercepted images.
8. A segmentation apparatus for three-dimensional medical images, comprising:
the acquisition unit is used for acquiring the three-dimensional medical image;
the dimension reduction processing unit is used for generating a plurality of two-dimensional medical images by dimension reduction processing on the three-dimensional medical images;
and the segmentation unit is used for carrying out image segmentation on at least one two-dimensional medical image in the plurality of two-dimensional medical images by utilizing an image segmentation model.
9. An electronic device, comprising:
a memory for storing a computer program;
a processor configured to perform the method of any one of claims 1 to 7.
10. A storage medium, comprising: program which, when run on an electronic device, causes the electronic device to perform the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311047266.7A CN117078696A (en) | 2023-08-18 | 2023-08-18 | Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311047266.7A CN117078696A (en) | 2023-08-18 | 2023-08-18 | Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117078696A true CN117078696A (en) | 2023-11-17 |
Family
ID=88709251
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311047266.7A Pending CN117078696A (en) | 2023-08-18 | 2023-08-18 | Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117078696A (en) |
-
2023
- 2023-08-18 CN CN202311047266.7A patent/CN117078696A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110321920B (en) | Image classification method and device, computer readable storage medium and computer equipment | |
| CN109409503B (en) | Neural network training method, image conversion method, device, equipment and medium | |
| CN111080584B (en) | Quality control method for medical image, computer device and readable storage medium | |
| CN110599421B (en) | Model training method, video fuzzy frame conversion method, device and storage medium | |
| US10867375B2 (en) | Forecasting images for image processing | |
| US11151703B2 (en) | Artifact removal in medical imaging | |
| CN111598989B (en) | Image rendering parameter setting method and device, electronic equipment and storage medium | |
| CN110782502A (en) | PET scattering estimation system based on deep learning and method using perception neural network model | |
| CN116664713B (en) | Training method of ultrasound contrast image generation model and image generation method | |
| CN112767505A (en) | Image processing method, training method, device, electronic terminal and storage medium | |
| Gheorghiță et al. | Improving robustness of automatic cardiac function quantification from cine magnetic resonance imaging using synthetic image data | |
| CN116433695B (en) | Mammary gland region extraction method and system of mammary gland molybdenum target image | |
| CN111260636B (en) | Model training method and device, image processing method and device, and medium | |
| WO2024051018A1 (en) | Pet parameter image enhancement method and apparatus, device, and storage medium | |
| CN111489318A (en) | Medical image enhancement method and computer-readable storage medium | |
| CN113888566B (en) | Target contour curve determination method and device, electronic equipment and storage medium | |
| CN114419375B (en) | Image classification method, training device, electronic equipment and storage medium | |
| CN117078696A (en) | Method and device for segmenting three-dimensional medical image, electronic equipment and storage medium | |
| CN113096238B (en) | X-ray diagram simulation method and device, electronic equipment and storage medium | |
| CN112530554B (en) | Scanning positioning method and device, storage medium and electronic equipment | |
| CN114723723A (en) | Medical image processing method, computer device and storage medium | |
| CN112233126A (en) | Windowing method and device for medical image | |
| CN115984229B (en) | Model training method, breast measurement device, electronic equipment and medium | |
| CN114037775B (en) | Bone structure growth method and device, electronic equipment and storage medium | |
| CN113239978B (en) | Method and device for correlation of medical image preprocessing model and analysis model |
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 |