CN117132570A - Automatic tissue positioning calculation method and system based on CT image - Google Patents

Automatic tissue positioning calculation method and system based on CT image Download PDF

Info

Publication number
CN117132570A
CN117132570A CN202311109282.4A CN202311109282A CN117132570A CN 117132570 A CN117132570 A CN 117132570A CN 202311109282 A CN202311109282 A CN 202311109282A CN 117132570 A CN117132570 A CN 117132570A
Authority
CN
China
Prior art keywords
fat
muscle
spine
image
region
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
Application number
CN202311109282.4A
Other languages
Chinese (zh)
Inventor
李迎燕
叶宏伟
李晓光
王瑶法
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Minfound Medical Systems Co Ltd
Original Assignee
Minfound Medical Systems Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Minfound Medical Systems Co Ltd filed Critical Minfound Medical Systems Co Ltd
Priority to CN202311109282.4A priority Critical patent/CN117132570A/en
Publication of CN117132570A publication Critical patent/CN117132570A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration by the use of local operators
    • G06T5/30—Erosion or dilatation, e.g. thinning
    • 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
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/24—Aligning, centring, orientation detection or correction of the image
    • G06V10/245—Aligning, centring, orientation detection or correction of the image by locating a pattern; Special marks for positioning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10081—Computed x-ray tomography [CT]
    • 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
    • G06T2207/30008—Bone
    • G06T2207/30012—Spine; Backbone

Abstract

The invention provides a tissue automatic positioning calculation method and system based on a CT image, wherein a spine bone in the CT image is segmented through a deep learning network, and a spine mask is obtained; performing coronal plane projection on the spinal mask, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash; triangular searching is carried out on two sides of the center point of the spinal bone ash so as to locate and obtain muscles and fat; wherein the search direction of the muscle and fat is opposite. The invention directly uses local tomographic images, and utilizes the special stable relation of the cancellous bone, the muscle and the fat in the human body to automatically identify and position the cancellous bone, the muscle and the fat; the method from cylindrical deduction to free form is used in the positioning process, so that the measured value is more accurate.

Description

Automatic tissue positioning calculation method and system based on CT image
Technical Field
The invention relates to the technical field of hospital image processing, in particular to an automatic tissue positioning calculation method and system based on CT images.
Background
The more dependent on the use of computed tomography CT scan data for prediction and assisted diagnosis of clinical disease in the clinic today. Currently, most commonly, bone density is measured, which is one of the main grounds for diagnosing osteoporosis. These require measurements of muscle, fat, and bone mass regions of interest based on measured CT values. In the measurement process, the professional who can identify the muscle, fat and bone areas in the scanned imaging image is required to operate, and the medical staff can obtain the result after the professional performs complicated operation on the muscle, fat and bone areas.
Current methods for diagnosing osteoporosis are DXA and QCT, DXA (using dual-x-rayabsolution) principles to directly detect x-ray absorption, and the requirements for equipment are not high. QCT requires relying on the phastom data for auxiliary measurement, and special phastom and software also bring inconvenience to detection. At present, a method for measuring cortical bone, muscle and fat is divided directly in fault code scanning data. However, in the method, the steps of manually dividing muscles and fat exist, so that the manual error is large, the work is complex, and the efficiency is low.
Disclosure of Invention
In order to overcome the technical defects, the invention aims to provide a tissue automatic positioning calculation method and system based on CT images, which can automatically identify and position bone ash, muscle and fat.
The invention discloses a tissue automatic positioning calculation method based on CT images, which comprises the following steps: dividing backbone bones in the CT image through a deep learning network to obtain a backbone mask; performing coronal plane projection on the spinal mask, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash; triangular searching is carried out on two sides of the center point of the spinal bone ash so as to locate and obtain muscles and fat; wherein the search direction of the muscle and fat is opposite.
Preferably, the performing coronal plane projection on the spinal mask, and performing smoothing processing on the projected image by using a filter kernel includes: the filter kernel size is
Preferably, the spine segmentation of the smoothed image by locating the center point of the spine to locate the bone ash of the spine includes: the spine divided from the smoothed image is a near-square body, and the coordinates (x 1 ,y 1 ) And (x) 2 ,y 2 ) The method comprises the steps of carrying out a first treatment on the surface of the Acquiring a center point (x 1 ,x 2 )/2，(y 1 ,y 2 ) 2; and determining the bone ash of the spine according to the central point.
Preferably, the performing a triangular search on both sides of the center point of the spinal bone ash to locate the acquisition of muscle and fat includes: dividing a central axis for distinguishing a muscle area from a fat area, wherein the two sides of the central axis are respectively provided with the muscle area and the fat area; searching was performed by defining a search range of 120 degrees in the muscle region and the fat region on both sides of the central axis, respectively.
Preferably, the performing a triangular search on both sides of the center point of the spinal bone ash to locate the acquisition of muscle and fat further comprises: a cylindrical region is used in the search region to determine whether muscle tissue or fat tissue is present, thereby demarcating the muscle or fat region.
Preferably, the determining whether the muscle tissue or the fat tissue exists using the cylindrical region in the search region, thereby demarcating the muscle or the fat region includes: and calculating the Hu value of the muscle or fat region, and defining an initial muscle region and an initial fat region in the muscle or fat region according to the Hu value range of the muscle tissue and the Hu value range of the fat tissue.
Preferably, the defining the initial muscle area and the initial fat area in the muscle or fat area according to the Hu value range of the muscle tissue and the Hu value range of the fat tissue further includes: and corroding and expanding the cylindrical region, and acquiring an accurate muscle region and an accurate fat region again.
Preferably, the dividing is used for distinguishing a central axis of the muscle area and the fat area, and the two sides of the central axis are respectively provided with the muscle area and the fat area; searching in a search range of 120 degrees defined in a muscle region and a fat region on both sides of the central axis, respectively, includes: p is p 1 P, points in the muscle area 2 Is a point in the fat region, p 0 Is a point on the central axis; let p be 0 (x 0 ,y 0 ,z i )，p 1 (x 1 ,y 1 ,z i )，p 1 (x 2 ,y 2 ,z i )；p 1 Satisfy the following requirementsOr (b)
Then p is 2 The initial search positions of (a) are:
or->
And the search range is:
or->Or (b)
Or->
Where H is the height value of the i-th layer image and W is the image width.
Preferably, the segmenting the spine bone in the CT image through the deep learning network, the obtaining the spine mask includes: the deep learning network is a segmentation network, and the segmentation network comprises a Resunate network, an nnunate network, a Unet++ network and a Nestnet network.
The invention also discloses a tissue automatic positioning computing system based on the CT image, which comprises a segmentation module and a positioning module, wherein the segmentation module comprises a first segmentation unit and a second segmentation unit based on a deep learning network; dividing the spine bone in the CT image through a deep learning network in the first dividing unit to obtain a spine mask; performing coronal plane projection on the spine mask through the second segmentation unit, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash; triangular searching is conducted on two sides of the center point of the spinal bone ash through the positioning module so as to obtain muscles and fat in a positioning mode; wherein the search direction of the muscle and fat is opposite.
After the technical scheme is adopted, compared with the prior art, the method has the following beneficial effects:
1. the invention directly uses local tomographic images, and utilizes the special stable relation of the cancellous bone, the muscle and the fat in the human body to automatically identify and position the cancellous bone, the muscle and the fat; the method from cylindrical deduction to free form is used in the positioning process, so that the measured value is more accurate.
Drawings
FIG. 1 is a flow chart of the automatic tissue positioning calculation method based on CT images provided by the invention;
fig. 2 is a schematic diagram of a search area of the method for searching and positioning muscles and fat provided by the invention.
Detailed Description
Advantages of the invention are further illustrated in the following description, taken in conjunction with the accompanying drawings and detailed description.
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.
The terminology used in the present disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used in this disclosure and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present disclosure. The word "if" as used herein may be interpreted as "at … …" or "at … …" or "responsive to a determination", depending on the context.
In the description of the present invention, it should be understood that the terms "longitudinal," "transverse," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate orientations or positional relationships based on the orientation or positional relationships shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and therefore should not be construed as limiting the present invention.
In the description of the present invention, unless otherwise specified and defined, it should be noted that the terms "mounted," "connected," and "coupled" are to be construed broadly, and may be, for example, mechanical or electrical, or may be in communication with each other between two elements, directly or indirectly through intermediaries, as would be understood by those skilled in the art, in view of the specific meaning of the terms described above.
In the following description, suffixes such as "module", "component", or "unit" for representing elements are used only for facilitating the description of the present invention, and are not of specific significance per se. Thus, "module" and "component" may be used in combination.
Lumbar positioning refers to the process of determining the position of the human lumbar spine (lumber vertebrae) or identifying a specific lumbar spine. Lumbar vertebrae are a group of vertebrae located in the lower portion of the spine, generally numbered L1 through L5. Lumbar positioning is commonly used in the fields of medical imaging, spinal surgery, and diagnosis. Computed Tomography (CT) can provide more detailed images of the lumbar spine and allow localization in three dimensions. A doctor can accurately locate the lumbar spine using CT scan images, determine the location and number of a particular lumbar spine.
Referring to fig. 1, the invention discloses a tissue automatic positioning calculation method based on CT images, which comprises the following steps:
s100, segmenting backbone bones in a CT image through a deep learning network to obtain a backbone mask;
s200, performing coronal plane projection on the spinal mask, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash;
s300, performing triangle search on two sides of a central point of the spinal bone ash so as to locate and obtain muscles and fat; wherein the search direction of the muscle and fat is opposite.
Bone ash, muscle and fat can be positioned through the three steps, so that each vertebra is segmented. In particular, the method comprises the steps of,
in step S100, using several (e.g., 50) cases of data, a segmentation of the spine bone is performed, and a rough spine mask is segmented, which is advantageous for further using image recognition technology on the segmented spine for the next step to perform vertebral body segmentation. The deep learning network used is a split network, where the split network includes, but is not limited to, a ResUnet network, an nnUnet network, a Unet++ network, a Nestnet network.
Step S200 is to precisely locate the bone ash of the spine by using an image processing method. Firstly, utilizing the characteristics of the spine, carrying out coronal plane projection, designing a filter kernel to carry out smoothing treatment on bone tissues in a smooth CT image, and carrying out spine segmentation on the smoothed image and keeping the original central point of the spine unchanged.
A Filter Kernel, also known as a convolution Kernel (Convolution Kernel) or convolution matrix (Convolution Matrix), is a small matrix or small Kernel used for filtering operations in image processing and signal processing. The filter kernel defines the weight and response characteristics of the filtering operation for performing convolution operations on the input image or signal to achieve different filtering effects. The filter kernel is typically a square matrix, with each element representing the weight of the pixel or signal at a corresponding location during the filtering process, for calculating the value of the output pixel or signal. The specific values of the filter kernel determine the effect of the filtering, and different values may enable different filtering operations, such as smoothing, sharpening, edge detection, etc. What is achieved here is a smoothing purpose. The filter kernel size is
The spine segmented in the smoothed image is a near-square body, and the coordinates (x 1 ,y 1 ) And (x) 2 ,y 2 ) From the two pairs of angles, the center point (x 1 ,x 2 )/2，(y 1 ,y 2 ) 2; last rootThe bone ash of the spine is determined according to the central point.
In step S300, the specific relationship of the specific tissue of the human body, the outside of the bone wraps the muscle, the outside of the muscle generates fat, and the physiological structure relationship is still preserved in the computed tomography image (CT image), so the invention can induce the relative inherent stable relationship among the bone, the muscle and the fat. In order to prevent errors caused by local tissue injury of the calculation result, a search strategy of radiating triangles to two sides of a cancellous bone center point is adopted in positioning, and muscles and fat are positioned. And in the process, the muscles and the fat search direction are opposite. And using the cylindrical region as a rough judgment basis for finding out the corresponding tissue in the search region, namely judging whether muscle tissue or fat tissue exists in the search region by using the cylindrical region, thereby demarcating the muscle or fat region.
The searching is performed by using 120-degree left and right sides of the central axis to search towards the middle. Namely: dividing a central axis for distinguishing a muscle area from a fat area, wherein the two sides of the central axis are respectively provided with the muscle area and the fat area; searches were performed by defining 120-degree search ranges in the muscle region and the fat region on both sides of the central axis, respectively.
For a given x-ray computed tomography image, different tissues have different HU values, and for lumbar vertebra data, CT image information has information of bones (lumbar vertebra), muscles and fat, so that the HU values can be obtained through calculation. Roughly judging the muscle range and the fat range according to the Hu value range, namely: the Hu value of the muscle or fat region is calculated, and an initial muscle region and an initial fat region in the muscle or fat region are defined according to the Hu value range of the muscle tissue and the Hu value range of the fat tissue.
Finally, after the initial muscle area and the initial fat area in the muscle or fat area are defined according to the Hu value range of the muscle tissue and the Hu value range of the fat tissue, the cylindrical area is eroded and expanded, and the accurate muscle area and the accurate fat area are obtained again.
Triangle search is a space search method based on triangles, and is commonly used for three-dimensional model rendering, collision detection, shape matching and other applications in computer graphics and computer vision. It accelerates queries and computations by dividing the three-dimensional space into a series of triangles and exploiting the characteristics of the triangles during the search. In triangle searching, there are typically two main steps:
1. triangle construction: first, a three-dimensional model is divided into a set of triangles. This may be done by a triangulation algorithm (e.g. Delaunay triangulation) or manual modeling. Each triangle is defined by three vertices and corresponding attributes of normals, texture coordinates, and the like.
2. Triangle search: in performing a query or calculation, a search is performed using the characteristics of the triangle. This includes the following aspects:
1) Triangle intersection detection: for detecting the intersection relationship of the ray and the triangle. For example, ray tracing or a pick operation may be performed using intersection tests of rays with triangles.
2) Triangle Bounding Box (Bounding Box): each triangle may define a bounding box for quickly excluding triangles that do not intersect the ray or other geometry. First, a minimum bounding box of the triangle is calculated, and then an intersection test is performed with the ray or bounding box of the query to determine if an intersection is possible.
3) Space division structure: triangle-based spatial segmentation structures (e.g., mesh, BVH, etc.) may further accelerate the search process. These structures group triangles to form a hierarchy such that queries need only be performed on a particular subset of space.
Through triangle search optimization and acceleration technology, various inquiry and calculation operations can be efficiently performed in a three-dimensional scene, and the performance and efficiency of applications such as rendering and collision detection are improved.
The search area cylinder determination is a spatial search algorithm that determines whether a given point is inside a certain area cylinder. The column is typically made up of a bottom polygon and a top plane, and may be used to represent a building, an obstacle or other object having a column structure. The following is a basic step of a common search area cylindrical judgment algorithm:
1. defining a column: a cylindrical bottom polygon and a top plane are determined. The bottom polygon may be represented by a set of vertices, which may be rectangular, polygonal, or other shape. The top plane may be defined by the center point and the normal direction of the cylinder.
2. Space division structure: to speed up the search process, a spatial segmentation structure, such as a grid, octree, or BVH (Bounding Volume Hierarchy), may be constructed. These structures divide the space into hierarchical sub-regions such that searches need only be performed within specific sub-regions.
3. And (3) point position judgment: for a given point, traversing the space division structure is firstly carried out, and determining the subarea where the point is located. Then, for the pillars within this sub-region, the following judgment is made:
1) And (3) judging the bottom polygon: using point-and-polygon internal test algorithms, e.g.
The intersection test of the ray and the polygon or the point-inside-polygon test to determine if the point is inside the bottom polygon.
2) Top plane judgment: by calculating the distance of the point to the top plane and comparing it to the height of the cylinder, it is determined whether the point is above the top plane.
3) And (3) comprehensive judgment: and determining whether the point is positioned in the column by combining the bottom polygon judgment and the top plane judgment.
By searching the regional column judgment algorithm, whether a given point is positioned inside a certain regional column can be judged, which plays an important role in applications such as path planning, collision detection, virtual environment modeling and the like.
Further, referring to fig. 2, in the process of searching by defining a search range of 120 degrees in a muscle area and a fat area on both sides of the central axis, a search strategy formula is explained, and a certain diagonal is used for explanation, which is defined as follows: p is p 1 P, points in the muscle area 2 Is a point in the fat region, p 0 Is a point on the central axis; let p be 0 (x 0 ,y 0 ,z i )，p 1 (x 1 ,y 1 ,z i )，p 1 (x 2 ,y 2 ,z i )；
p 1 Satisfy the following requirementsOr->
Then p is 2 The initial search positions of (a) are:
or->
And the search range is:
or->Or (b)
Or->
Where H is the height value of the i-th layer image and W is the image width.
So far, after each vertebra can be automatically segmented, the scanned image contains the coccyx by default, but not necessarily contains all lumbar vertebrae nodules, and on the basis of the image processing, the final coccyx is used as a positioning, and the image recognition targets are ordered in reverse order, so that the formula L is given i =5-i, obtaining lumbar name. L (L) i Includes L 1 、L 2 、L 3 、L 4 、L 5 。
The invention also discloses a tissue automatic positioning computing system based on the CT image, which comprises a segmentation module and a positioning module, wherein the segmentation module comprises a first segmentation unit and a second segmentation unit based on a deep learning network; dividing the spine bone in the CT image through a deep learning network in a first dividing unit to obtain a spine mask; performing coronal plane projection on the spine mask through a second segmentation unit, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash; triangular searching is conducted on two sides of a center point of the spinal bone ash through a positioning module so as to obtain muscles and fat in a positioning mode; wherein the search direction of the muscle and fat is opposite.
The invention directly uses local tomographic images, and utilizes the special stable relation of the cancellous bone, the muscle and the fat in the human body to automatically identify and position the cancellous bone, the muscle and the fat; the method from cylindrical deduction to free form is used in the positioning process, so that the measured value is more accurate.
It should be noted that the embodiments of the present invention are preferred and not limited in any way, and any person skilled in the art may make use of the above-disclosed technical content to change or modify the same into equivalent effective embodiments without departing from the technical scope of the present invention, and any modification or equivalent change and modification of the above-described embodiments according to the technical substance of the present invention still falls within the scope of the technical scope of the present invention.

Claims (10)

1. The automatic tissue positioning calculation method based on the CT image is characterized by comprising the following steps of:
dividing backbone bones in the CT image through a deep learning network to obtain a backbone mask;
performing coronal plane projection on the spinal mask, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash;
triangular searching is carried out on two sides of the center point of the spinal bone ash so as to locate and obtain muscles and fat; wherein the search direction of the muscle and fat is opposite.
2. The method of claim 1, wherein performing coronal plane projection on the spinal mask, and performing smoothing on the projected image using a filter kernel comprises:
the filter kernel size is
3. The method according to claim 1, wherein the performing the spine segmentation on the smoothed image by locating the center point of the spine to locate the bone ash of the spine comprises:
the spine divided from the smoothed image is a near-square body, and the coordinates (x 1 ,y 1 ) And (x) 2 ,y 2 )；
Acquiring a center point (x 1 ,x 2 )/2，(y 1 ,y 2 )/2；
And determining the bone ash of the spine according to the central point.
4. The method of claim 1, wherein performing a triangular search on both sides of the center point of the spinal bone ash to locate the acquisition of muscle and fat comprises:
dividing a central axis for distinguishing a muscle area from a fat area, wherein the two sides of the central axis are respectively provided with the muscle area and the fat area;
searching was performed by defining a search range of 120 degrees in the muscle region and the fat region on both sides of the central axis, respectively.
5. The method of claim 4, wherein performing a triangular search on both sides of the center point of the spinal bone ash to locate the acquisition of muscle and fat further comprises:
a cylindrical region is used in the search region to determine whether muscle tissue or fat tissue is present, thereby demarcating the muscle or fat region.
6. The method of claim 5, wherein determining whether muscle tissue or fat tissue is present in the search area using a cylindrical area, thereby defining the muscle or fat area, comprises:
and calculating the Hu value of the muscle or fat region, and defining an initial muscle region and an initial fat region in the muscle or fat region according to the Hu value range of the muscle tissue and the Hu value range of the fat tissue.
7. The method according to claim 6, wherein the defining the initial muscle region and the initial fat region from the muscle tissue Hu value range and the fat tissue Hu value range further comprises:
and corroding and expanding the cylindrical region, and acquiring an accurate muscle region and an accurate fat region again.
8. The method according to claim 7, wherein the dividing is used for distinguishing a central axis of the muscle area and the fat area, and the two sides of the central axis are respectively the muscle area and the fat area; searching in a search range of 120 degrees defined in a muscle region and a fat region on both sides of the central axis, respectively, includes:
p 1 p, points in the muscle area 2 Is a point in the fat region, p 0 Is a point on the central axis;
let p be 0 (x 0 ,y 0 ,z i )，p 1 (x 1 ,y 1 ,z i )，p 1 (x 2 ,y 2 ,z i )；p 1 Satisfy the following requirementsOr->
Then p is 2 The initial search positions of (a) are:
or->
And the search range is:
or->Or (b)
Or->
Where H is the height value of the i-th layer image and W is the image width.
9. The method of claim 1, wherein the segmenting the spine bone in the CT image via the deep learning network, obtaining the spine mask comprises:
the deep learning network is a segmentation network, and the segmentation network comprises a Resunate network, an nnunate network, a Unet++ network and a Nestnet network.
10. The automatic tissue positioning computing system based on the CT image is characterized by comprising a segmentation module and a positioning module, wherein the segmentation module comprises a first segmentation unit and a second segmentation unit based on a deep learning network;
dividing the spine bone in the CT image through a deep learning network in the first dividing unit to obtain a spine mask;
performing coronal plane projection on the spine mask through the second segmentation unit, and performing smoothing treatment on the projected image by adopting a filter kernel; performing spine segmentation on the smoothed image by positioning a spine central point so as to obtain spine bone ash;
triangular searching is conducted on two sides of the center point of the spinal bone ash through the positioning module so as to obtain muscles and fat in a positioning mode; wherein the search direction of the muscle and fat is opposite.
CN202311109282.4A 2023-08-31 2023-08-31 Automatic tissue positioning calculation method and system based on CT image Pending CN117132570A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311109282.4A CN117132570A (en) 2023-08-31 2023-08-31 Automatic tissue positioning calculation method and system based on CT image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311109282.4A CN117132570A (en) 2023-08-31 2023-08-31 Automatic tissue positioning calculation method and system based on CT image

Publications (1)

Publication Number Publication Date
CN117132570A true CN117132570A (en) 2023-11-28

Family

ID=88862598

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311109282.4A Pending CN117132570A (en) 2023-08-31 2023-08-31 Automatic tissue positioning calculation method and system based on CT image

Country Status (1)

Country Link
CN (1) CN117132570A (en)

Similar Documents

Publication Publication Date Title
US9153021B2 (en) Estimating risk of future bone fracture utilizing three-dimensional bone density model
US7505617B2 (en) Fiducial-less tracking with non-rigid image registration
Lu et al. Automatic re-contouring in 4D radiotherapy
US7259762B2 (en) Method and system for automatically transforming CT studies to a common reference frame
US7639866B2 (en) Method of radiographic imaging for three-dimensional reconstruction, and a computer program and apparatus for implementing the method
US6028907A (en) System and method for three-dimensional geometric modeling by extracting and merging two-dimensional contours from CT slice data and CT scout data
US7616818B2 (en) Method of determining the orientation of an image
CN100528078C (en) Method and system for airway measurement
US7492968B2 (en) System and method for segmenting a structure of interest using an interpolation of a separating surface in an area of attachment to a structure having similar properties
US20060002631A1 (en) ROI selection in image registration
US20080117210A1 (en) Virtual endoscopy
US20110274330A1 (en) Automated characterization of time-dependent tissue change
US20030099391A1 (en) Automated lung nodule segmentation using dynamic progamming and EM based classifcation
CN103156632B (en) Lumbar vertebra three-dimensional measurement method
JP2015530155A (en) Analysis Morphomics: High-speed medical image automatic analysis method
CN108830852A (en) Three-D ultrasonic tumour auxiliary measurement system and method
CN111166332B (en) Radiotherapy target region delineation method based on magnetic resonance spectrum and magnetic resonance image
JP2007534416A5 (en)
US20030154201A1 (en) Data storage format for topography data
CN1399763A (en) 3-D shape measurements using statistical curvature analysis
CN117132570A (en) Automatic tissue positioning calculation method and system based on CT image
JP4822669B2 (en) How to determine image orientation
Rusinek et al. Quantitative and qualitative comparison of volumetric and surface rendering techniques
US20040171931A1 (en) Method to produce a volume data set
WO2022128956A1 (en) Method for determining the screw trajectory of a pedicle bone screw

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