KR101185728B1 - A segmentatin method of medical image and apparatus thereof - Google Patents
A segmentatin method of medical image and apparatus thereof Download PDFInfo
- Publication number
- KR101185728B1 KR101185728B1 KR1020110095149A KR20110095149A KR101185728B1 KR 101185728 B1 KR101185728 B1 KR 101185728B1 KR 1020110095149 A KR1020110095149 A KR 1020110095149A KR 20110095149 A KR20110095149 A KR 20110095149A KR 101185728 B1 KR101185728 B1 KR 101185728B1
- Authority
- KR
- South Korea
- Prior art keywords
- segmentation
- region
- medical image
- determined
- pointer
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000011218 segmentation Effects 0.000 claims abstract description 299
- 230000003902 lesion Effects 0.000 claims abstract description 122
- 238000003745 diagnosis Methods 0.000 claims abstract description 12
- 210000000988 bone and bone Anatomy 0.000 claims description 9
- 230000002093 peripheral effect Effects 0.000 claims description 8
- 208000003174 Brain Neoplasms Diseases 0.000 claims description 6
- 206010028980 Neoplasm Diseases 0.000 claims description 6
- 208000020816 lung neoplasm Diseases 0.000 claims description 5
- 208000037841 lung tumor Diseases 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 210000004072 lung Anatomy 0.000 claims 8
- 238000003709 image segmentation Methods 0.000 claims 1
- 210000004556 brain Anatomy 0.000 description 20
- 239000000284 extract Substances 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 241000270295 Serpentes Species 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000002059 diagnostic imaging Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 210000001519 tissue Anatomy 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 210000005013 brain tissue Anatomy 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 208000011581 secondary neoplasm Diseases 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
- A61B6/03—Computed tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/03—Arrangements for converting the position or the displacement of a member into a coded form
- G06F3/041—Digitisers, e.g. for touch screens or touch pads, characterised by the transducing means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/14—Digital output to display device ; Cooperation and interconnection of the display device with other functional units
-
- 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
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20101—Interactive definition of point of interest, landmark or seed
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Medical Informatics (AREA)
- Heart & Thoracic Surgery (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Radiology & Medical Imaging (AREA)
- Optics & Photonics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- High Energy & Nuclear Physics (AREA)
- Biophysics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
Description
The present invention relates to segmentation in medical images, and more particularly, to identify a lesion type in a sliced medical image, and to determine the type of the lesion, for example, a hypertumor or a lung tumor. By selecting the optimal segmentation region for the lesion type by the user's interaction using a segmentation algorithm according to a brain tumor, etc., an optimal seed for generating a three-dimensional segmentation volume of the lesion ( The present invention relates to a segmentation method and apparatus for medical imaging in which a user can select a seed and obtain an optimized three-dimensional segmentation volume.
The present invention is derived from research carried out by the Ministry of Knowledge Economy and the Korea Industrial Technology Evaluation and Management Center as part of the Knowledge Economy Technology Innovation Project (Industrial Source Technology Development Project) [Project Number: 10038419, Project Name: Intelligent Image Diagnosis and Treatment Support system].
Cancer is important for diagnosing and carefully monitoring the disease as early as possible, and doctors are interested in not only the primary tumor but also secondary tumors that may have metastasized through the rest of the body.
Such a tumor-like lesion may be diagnosed and monitored through a three-dimensional segmentation volume, which may be formed from segmentation of each of the plurality of two-dimensional medical images.
In the 3D segmentation volume according to the related art, when a doctor, that is, a user selects a specific position to be monitored in the 2D medical image, performs the 2D segmentation using the information on the 2D segmentation result and performs the 2D segmentation result. As a basis, a three-dimensional segmentation volume is generated.
Since the segmentation method according to the related art cannot know the segmentation result of the specific position selected by the user in the 2D medical image, the result generated through the 3D segmentation volume generated after the 3D segmentation process is completed can be known. . In other words, if the generated 3D segmentation volume is not satisfactory, the user may reselect the location of the lesion to be confirmed in the 2D medical image and check the result through the 3D segmentation volume. This can put a load on the system or device that creates the segmentation volume and can be inconvenient for the user.
In addition, in the conventional segmentation method, since the lesion type is determined from the slice medical image and the optimized segmentation according to the determined lesion type cannot be performed, the optimized segmentation method cannot perform the optimized two-dimensional segmentation according to the lesion. There is a problem that it is difficult to obtain a three-dimensional segmentation volume.
Thus, there is a need for a method for obtaining an optimized two-dimensional segmentation seed for a lesion type selected by a user in a sliced medical image.
The present invention is derived to solve the above problems of the prior art, by determining the lesion type in the slice medical image, and performing segmentation corresponding to the determined lesion type, to obtain an optimal segmentation seed according to the lesion type It is an object of the present invention to provide a segmentation method and apparatus thereof for medical imaging.
Specifically, the present invention uses a brightness value for the position information of the pointer and an angular profile at the corresponding location, for example, a lesion type for the corresponding location, for example, a hyper tumer, a run tumer, a brain tumer, a general tumer, or the like. To provide a segmentation method in a medical image capable of acquiring an optimal segmentation seed for generating a three-dimensional segmentation volume according to the lesion by performing a predetermined segmentation algorithm according to the determined lesion type. For the purpose of
In addition, the present invention can obtain an optimal three-dimensional segmentation volume by acquiring the optimal two-dimensional segmentation seed according to the type of lesion, thereby reducing the load for obtaining the three-dimensional segmentation volume in the medical image It is an object of the present invention to provide a segmentation method and an apparatus thereof.
In addition, the present invention by displaying the optimal two-dimensional segmentation seed determined according to the type of lesion on the slice medical image in advance, and determined by the user's selection, the segmentation in the medical image that allows the user to select the optimal segmentation seed It is an object to provide a method and an apparatus thereof.
In order to achieve the above object, a segmentation method in a medical image according to an embodiment of the present invention comprises the steps of extracting the position information of the pointer according to the user input from the slice (medical) image displayed on the screen; Determining a lesion type based on at least one of information on a slice medical image related to the extracted position information of the pointer and an angle profile based on the position information of the pointer; Determining a segmentation area including a position of the pointer using a segmentation algorithm preset according to the determined lesion type; And selecting the segmentation region as the lesion diagnosis region for the slice medical image.
The method may further include extracting a rung region and a bone region from the slice medical image, wherein the determining may include: a first preset profile of the rung region among the angle profiles; If larger than a reference value, the lesion type is determined as a rung tumor, and if the profile meeting the extracted region of the angular profile is greater than a second preset reference value, the lesion type is brain tumor. Can be determined by
The determining may include calculating a range of brightness values based on information of the slice medical image associated with the extracted position information of the pointer; Determining a first segmentation area including a position of the pointer using a segmentation algorithm preset according to the calculated range of the brightness value and the determined lesion type; Applying a preset fitting model to the determined first segmentation region; And determining an optimal segmentation region from the first segmentation region by using the fitting model.
The determining of the first segmentation area may include determining a second segmentation area including the location of the pointer using the calculated range of the brightness value; Selecting the second segmentation area into a first profile that meets a pre-extracted rung area of the angle profile and a second profile that does not meet the rung area when the determined lesion type is a rung tumer; Interpolating the second profile to the first profile to form a fence; And determining the first segmentation region from the second segmentation region based on the first profile and the formed fence.
The determining of the first segmentation area may include determining a second segmentation area including the location of the pointer using the calculated range of the brightness value; Selecting an anti-seed outside the second segmentation region when the determined lesion type is a brain tumer; Forming an anti seed region based on the brightness value of the selected anti seed; And determining the first segmentation region from the second segmentation region by using the formed anti seed region.
The determining may include determining a lesion type as a hyper tutor if the brightness value of the position information of the pointer is greater than a preset threshold value, and determining the lesion type as a hyper tumer. In this case, an area greater than or equal to the threshold value including the location of the pointer may be determined as the segmentation area.
The method may further include displaying the determined segmentation region on the slice medical image in advance, and selecting the segmentation region on the slice medical image when the segmentation region is displayed by the user. Can be selected as a diagnostic area for lesions.
The extracting may include detecting optimal position information in a predetermined peripheral area including a brightness value of the extracted position information of the pointer and position information of the pointer, and determining the optimal position information. The type of lesion may be determined based on at least one of information on the slice medical image related to the slice and an angle profile based on the optimal position information of the pointer.
The method may further include determining the selected lesion diagnosis region as a seed for segmentation of a 3D volume image; Determining a segmentation region of each of the plurality of slice medical images associated with the slice medical image based on the determined seed; And generating a 3D segmentation volume by using a segmentation region of each of the determined seeds and the plurality of slice medical images.
A segmentation apparatus in a medical image according to an embodiment of the present invention includes a position extraction unit for extracting position information of a pointer according to a user input from a slice medical image displayed on a screen; A discrimination unit for determining a lesion type based on at least one of sliced medical image information related to the extracted position information of the pointer and an angle profile based on the position information of the pointer; A determination unit to determine a segmentation area including a position of the pointer using a segmentation algorithm preset according to the determined lesion type; And a selection unit for selecting the segmentation region as the lesion diagnosis region for the slice medical image.
According to the present invention, the segmentation area is determined in the slice medical image, segmentation corresponding to the determined lesion type is displayed in advance, and the segmentation area determined in advance is selected by the user. By performing segmentation of another sliced medical image to generate a 3D segmentation volume using the selected segmentation region as a seed for the lesion, an optimal segmentation seed according to the type of lesion is obtained to generate an optimal 3D segmentation volume. have.
Furthermore, since the present invention obtains an optimal two-dimensional segmentation seed for generating a three-dimensional segmentation volume by using a segmentation algorithm according to the type of lesion, the system load for generating an optimal three-dimensional segmentation volume can be reduced.
Specifically, the present invention displays the segmentation result performed according to the determined lesion type in advance on the screen so that the user selects an optimal segmentation seed by the user, and the segmentation seed is selected by the user's selection. This reduces the load on creating three-dimensional segmentation volumes.
In other words, the present invention can determine whether the user adopts the pre-segmentation result for the lesion in the two-dimensional slice image currently displayed to the user, so that the validity of the pre-segmentation result can be quickly verified. In addition, since the pre-segmentation results adopted by the user are the first-validated results, when performing segmentation in the 3D image by using the verified pre-segmentation results as the seed region, since the seed region includes excellent information, relatively fewer resources are used. Excellent 3D segmentation results can be obtained.
In addition, since the present invention determines an optimal segmentation seed by determining the type of lesion, an optimal three-dimensional segmentation result for various types of lesions can be obtained, thereby increasing reliability of the segmentation result.
1 is a flowchart illustrating an operation of a segmentation method in a medical image according to an exemplary embodiment.
FIG. 2 shows an operation flowchart of an embodiment of step S140 shown in FIG. 1.
3 is a flowchart illustrating an embodiment of step S150 illustrated in FIG. 1.
4 is a flowchart illustrating an embodiment of step S160 shown in FIG. 1.
FIG. 5 is a diagram illustrating an example of a medical image for describing an operation flowchart illustrated in FIG. 4.
FIG. 6 illustrates an operation flowchart of another embodiment of step S160 shown in FIG. 1.
FIG. 7 is a diagram illustrating an example of a medical image for describing an operation flowchart illustrated in FIG. 6.
8 is a flowchart illustrating still another embodiment of operation S160 shown in FIG. 1.
FIG. 9 is a diagram illustrating an example of a medical image for describing an operation flowchart illustrated in FIG. 8.
10 illustrates a configuration of a segmentation apparatus in a medical image according to an embodiment of the present invention.
FIG. 11 is a diagram illustrating an example configuration of a determination unit illustrated in FIG. 10.
Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the accompanying drawings.
Preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.
However, the present invention is not limited to or limited by the embodiments. Like reference symbols in the drawings denote like elements.
Hereinafter, a segmentation method and a device in a medical image according to an embodiment of the present invention will be described in detail with reference to FIGS. 1 to 11.
1 is a flowchart illustrating an operation of a segmentation method in a medical image according to an exemplary embodiment.
Referring to FIG. 1, the segmentation method controls a pointer displayed on a slice medical image according to a user input, for example, a mouse movement, in the slice medical image selected by the user (S110).
When the pointer position is changed or the pointer is stopped by the pointer control according to the user input, the position information of the pointer is extracted (S120).
Here, the location information of the pointer may be coordinate information in the slice medical image.
When the location information of the pointer is extracted, the granular noise is removed from the slice medical image, and the optimal location information of the pointer is extracted based on the slice medical image information related to the extracted location information of the pointer (S130 and S140).
Of course, the step S130 of removing the granular noise may be performed before the slice medical image is displayed on the screen when the slice medical image is selected by the user.
The optimal position information of the pointer in step S140 may be a seed point for determining the segmentation area. The step S140 of extracting the optimal position information will be described in detail with reference to FIG. 2 as follows.
FIG. 2 shows an operation flowchart of an embodiment of step S140 shown in FIG. 1.
Referring to FIG. 2, the extracting of the optimal location information (S140) may include a circular or rectangular shape having a predetermined size around a location of a pointer, for example, a pointer location according to a user's motion or a user's input. An average value of brightness values of the area is calculated (S220).
Here, the brightness values of the circular or rectangular region mean slice medical image information corresponding to the circular or rectangular region in the slice medical image.
In operation S220, when the average value of the brightness values of the predetermined area including the pointer position is calculated, the calculated average value is compared with the brightness value of the corresponding pointer position, and the comparison result indicates that the brightness value of the pointer position is within an error range of the average value. It is determined whether there is (S230, S240).
For example, it is determined whether the brightness value of the pointer position is between two values of "average value-a" and "average value + a". Here, the value a may be dynamically determined or predetermined depending on the situation.
As a result of the determination in step S240, when the brightness value of the pointer position is out of an error range with respect to the average value, the optimum position information is extracted from the brightness values included in the predetermined region based on the average value.
Here, the optimal position information in the predetermined region may be position information having a brightness value corresponding to an average value among the brightness values included in the predetermined region. When there are a plurality of brightness values corresponding to the average value, the position and the position of the pointer The adjacent position can be extracted as the optimal position information. Of course, the position closest to the position of the pointer may be extracted as the optimal position information. However, the present invention is not limited thereto, and an arbitrary position having an average value may be arbitrarily extracted as the optimal position information.
On the other hand, if the brightness value of the pointer position exists within the error range with respect to the average value as a result of the determination in step S240, the position information of the current pointer is extracted as the optimal position information (S260).
Referring back to FIG. 1, when the optimal position information of the pointer is extracted in step S140, the slice is sliced based on at least one of slice medical image information related to the extracted optimal position information and an angle profile based on the optimal position information. The type of lesion in which the pointer is positioned in the medical image is determined (S150).
Here, the angle profile means a profile from 0 degrees to 360 degrees, and may be a profile for the same angle, for example, a 10 degree interval from 0 degrees to 360 degrees, or a profile for different angles.
If the lesion type in the medical image where the pointer is located or the lesion type present in the slice medical image is determined by step S150, the segmentation algorithm is set in advance according to the determined lesion type, and thus the pointer position information or the optimal position information. A segmentation area including a is determined (S160).
The segmentation region determined in step S160 is an optimal segmentation region determined by the segmentation algorithm for the lesion.
For example, when the determined lesion type is a rung tumer, the rung tumer segmentation algorithm is performed, and when the determined lesion type is a brain tumer, the brain tumer segmentation algorithm is performed to perform the analysis on the corresponding lesion. Determine the optimal segmentation area.
When the optimal segmentation region including the pointer position information or the optimal position information is determined, the determined segmentation region, that is, the optimal segmentation region, is previously displayed on the slice medical image (S170).
When the optimal segmentation region previously displayed on the screen is selected by the user, the segmentation region is selected as a lesion diagnosis region in the slice medical image, and the selected lesion diagnosis region is determined as a seed for performing 3D volume segmentation. (S180, S190).
As a method of determining the segmentation area displayed on the screen by the user in operation S180, various methods, such as a click or double click of a pointer and a shortcut key input, may be applied.
When the seed for the corresponding lesion is determined, a segmentation region of each of the plurality of other slice medical images related to the slice medical image is determined based on the determined seed (S200).
When the segmentation region of each of the plurality of other slice medical images is determined, a 3D segmentation volume is generated using the segmentation region of the corresponding slice medical image, that is, the segmentation region of each of the plurality of slice medical images different from the seed (S210).
On the other hand, if the segmentation region previously displayed on the screen is not selected by the user in step S180, that is, if the user is not satisfied with the segmentation region, step S110 is performed again.
As described above, the present invention may determine an optimal segmentation region for a corresponding lesion by determining a lesion type of a sliced medical image and performing a segmentation algorithm corresponding to the determined lesion, and interacting with the user with the determined segmentation region. Through the selection, the optimal segmentation seed desired by the user can be selected, and thus an optimal three-dimensional segmentation volume can be obtained.
In addition, the present invention has an advantage that the user input is simplified and the user interface is simplified because the pre-segmentation process is performed according to the location information of the pointer and the type of the lesion.
In addition, the present invention checks the segmentation region previously displayed on the slice medical image, and since the segmentation seed is determined by user selection, the 3D segmentation performance can be improved and the user's satisfaction can be improved.
FIG. 3 is a flowchart illustrating an embodiment of operation S150 of FIG. 1, which illustrates a process of determining a hyper tutor, a rung tumer, a brain tutor, and a general tumer.
Referring to FIG. 3, in determining the type of lesion (S150), it is determined whether the brightness value of the optimal position information of the pointer extracted in step S140 is greater than a preset threshold value, for example, 500 [HU] (S310). .
Here, HU is a unit of Hounsfield.
In operation S310, if the brightness value of the optimal position information is greater than the threshold value, the lesion type is determined by a hyper tutor. If the brightness value of the optimal position information is less than the threshold value, the lesion type is determined as a hypotumer. The process of determining the rung tumer, the brain tumer and the general tumer is performed (S320).
Here, the hyper-tumer means a higher brightness value than the surroundings, the hypo-tumer means a lower brightness value than the surroundings, and the hyper-tutors and hypo-tumers can be understood by those skilled in the art. Omit.
Of course, it should be appreciated that the threshold for discriminating with the hyper tumer may vary depending on the situation.
If it is determined that the lesion is a hypotumer in step S310, the rung region and the bone region are extracted from the slice medical image, and the angle profiles based on the extracted rung region, the bone region and the optimal position information are analyzed (S330 and S340). ).
Here, the angular profile to be analyzed is to analyze the angle profile that meets the rung area and the main area of the overall angle profile, and to analyze what percentage of the angle profile meets the rung area or the bone area. .
Of course, both the rung area and the main area may be extracted from the slice medical image, but only the main area may be extracted. When both areas are extracted, the angular profile is analyzed for both areas, but only one area is extracted. In the case of extraction, only the angular profile of one region needs to be analyzed.
In the method of extracting the rung region in operation S330, an area having a brightness value smaller than a first set value, for example, −200 [HU] in the slice medical image may be extracted as the rung region. For example, in the slice medical image of FIG. 5, an inner region of a body having a brightness value smaller than −200 [HU] is extracted as the
In addition, the method of extracting the region viewed in operation S330 may extract, as the region where the brightness value is greater than the second set value, for example, 5000 [HU], from the slice medical image. For example, in the slice medical image illustrated in FIG. 7, a region having a brightness value greater than 500 [HU] is extracted to the
If the angle profile that meets the rung region analyzed by step S340 is greater than a first reference value, for example, 50 [%], the lesion on which the pointer is located is determined as a rung tumer (S350 and S360).
On the other hand, if the angle profile that meets the rung area is smaller than the first reference value as a result of the determination in step S350, it is determined whether the angle profile that meets the present area analyzed by step S340 is greater than the preset second reference value, for example, 80 [%]. (S370).
As a result of the determination in step S370, when the angle profile meeting the region is greater than the second reference value, the lesion on which the pointer is located is determined by the brain tumer (S380).
On the contrary, as a result of the determination of S370, when the angle profile meeting the region is smaller than the second reference value, the lesion on which the pointer is located is determined as a general tumer except the rung and brain tumers (S490).
Of course, the first reference value and the second reference value for determining the rung and brain tumers may vary according to circumstances.
When the lesion type is determined by the process of FIG. 3, the segmentation area of the corresponding lesion is determined using a segmentation algorithm for the determined lesion type. When the determined lesion type is a rung tumer, a brain tutor, or a general tumer, the segmentation area is determined. The assumption to determine will be described with reference to FIGS. 4 to 9.
4 is an operation flowchart of an embodiment of step S160 shown in FIG. 1, and is an operation flowchart of a case in which the lesion determined by step S150 is a rung tumer.
Referring to FIG. 4, the determining of the segmentation region of the rung tumer (S160) may include a range of brightness values for determining the segmentation region based on medical image information, eg, a brightness value, of the optimal position information extracted by step S140. Calculate (S410).
Here, the brightness value range may be calculated by applying a constant standard deviation based on the brightness value for optimal position information, or may be calculated by designating a predetermined range of predetermined values.
When the brightness value range is calculated, the first segmentation area corresponding to the brightness value range is determined using the calculated brightness value range (S420).
In this case, the first segmentation region may include location information of the pointer or extracted optimal location information according to a user input.
When the first segmentation area is determined, the first segmentation area is selected as a first profile that meets the rung area and a second profile that does not meet the rung area using the angle profile (S430).
Here, since the second profile that does not meet the rung area among the first segmentation areas is a part out of the rung area, a process of limiting the area for the second profile should be performed.
That is, a fence is formed by interpolating an area corresponding to the second profile to an area corresponding to the first profile (S440).
For example, as illustrated in FIG. 5, the
When the fence is formed, the second segmentation area is determined based on the formed fence and the first segmentation area corresponding to the first profile (S450).
When the second segmentation area is determined, an
Here, the fitting model may include a deformable model, a snake model, and the like, and the fitting models such as the deformable model and the snake model are within the scope obvious to those skilled in the art for applying to the steps S460 and S470. Can be modified in
FIG. 6 is a flowchart illustrating an operation of another embodiment of step S160 illustrated in FIG. 1, and illustrates an operation flowchart of a case where the lesion determined by step S150 is a brain tumer.
Referring to FIG. 6, in the determining of the segmentation region of the brain tumer (S160), the brightness value range for determining the segmentation region is calculated based on the brightness value of the optimal position information extracted in step S140 (S610). .
When the brightness value range is calculated, the first segmentation area corresponding to the brightness value range is determined using the calculated brightness value range (S620).
In this case, the first segmentation region may include location information of the pointer or extracted optimal location information according to a user input.
When the first segmentation area is determined, an anti-seed for correcting the first segmentation area is selected (S630).
Here, the anti seed refers to a normal tissue (tissue) that is not a lesion in the brain, and the anti seed is selected at least one from the area between the first segmentation area and the main area.
For example, in the medical image illustrated in FIG. 7, a plurality of positions corresponding to normal tissues of a brain near to the
When the anti seed is selected, a range of brightness values for forming the anti seed region is calculated based on the brightness value of the selected anti seed (S640).
Here, the range of the brightness value for forming the anti seed region may be set in advance, or may be set using the brightness value of the peripheral region around the selected anti seed.
Based on the brightness value calculated in step S640, an
Here, the anti-seed region is formed to determine a more accurate segmentation region by correcting the anti-seed region because the determined first segmentation region is uncertain.
When the anti seed region is formed, a second segmentation region is determined from the first segmentation region based on the formed anti seed region and the first segmentation region (S660).
In this case, the determined second segmentation region is a segmentation region including a pointer position formed using the anti-seed region as a fence, and may be a region smaller than the first segmentation region but larger than the first segmentation region.
When the second segmentation area is determined, an
Here, the fitting model may include a deformable model, a snake model, and the like.
FIG. 8 is a flowchart illustrating operations of another embodiment of step S160 illustrated in FIG. 1, and is a flowchart illustrating operations when the lesion determined by step S150 is a general tumer.
Referring to FIG. 8, in the determining of the segmentation region of the general tutor (S160), the brightness value range for determining the segmentation region is calculated based on the brightness value of the optimal position information extracted in step S140 (S810). .
When the brightness value range is calculated, the first segmentation region corresponding to the brightness value range is determined using the calculated brightness value range (S820).
When the first segmentation region is determined, an
Here, the fitting model may include a deformable model, a snake model, and the like.
4 to 9, the optimal segmentation region for the rung tumer, the brain tumer, and the general tumer can be determined, and in the case of the hyper tumer, it corresponds to a preset brightness value range in the peripheral region including the optimal position information. The region to be determined determines the optimal segmentation region.
As described above, in the segmentation method according to the lesion type according to the present invention, by determining the type of the lesion in which the pointer is located in the slice medical image, and performing a segmentation algorithm for the lesion, an optimal segmentation region for the lesion may be determined. Since the optimal segmentation region determined in this way is used, a highly reliable three-dimensional segmentation volume can be generated for the lesion.
10 illustrates a configuration of a segmentation apparatus in a medical image according to an embodiment of the present invention.
Referring to FIG. 10, the segmentation apparatus includes a
The
Here, the
Furthermore, the
The
In this case, the
The
In this case, the
In addition, the
The
That is, the
Of course, the
The detailed configuration of the
The
The
In this case, the segmentation area selected by the
Also, although not shown in FIG. 10, a segment for determining segmentation regions of each of the plurality of slice medical images associated with the slice medical image using the segmentation seed selected by the selection unit, and each of the determined seeds and the plurality of slice medical images, respectively. It may also include a configuration for generating a three-dimensional segmentation volume for the lesion using the segmentation area of the.
FIG. 11 is a diagram illustrating an example configuration of a determination unit illustrated in FIG. 10.
Referring to FIG. 11, the
If the lesion type determined by the
When the lesion type is described as a rung tumer and a brain tumer, the following description will be given.
1) When the lesion type is rung tumer
The
The
The
The forming
The
The
2) Type of lesion is brain tumer
The
The
The
In this case, the
The forming
Herein, the brightness value range for forming the anti seed region may be preset or determined using the brightness value of the peripheral region of the selected anti seed.
The
At this time, the anti seed region serves as a fence for determining the second segmentation region.
The
The segmentation method in a medical image according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.
In the present invention as described above has been described by the specific embodiments, such as specific components and limited embodiments and drawings, but this is provided to help a more general understanding of the present invention, the present invention is not limited to the above embodiments. For those skilled in the art, various modifications and variations are possible from these descriptions.
Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, belong to the scope of the present invention .
Claims (14)
Extracting location information of a pointer according to a user input from a slice medical image displayed on a screen by a location extracting unit of the segmentation device;
The slice medical image is represented based on at least one of information of a slice medical image associated with the extracted position information of the pointer and an angle profile based on the position information of the pointer, by the determination unit of the segmentation apparatus. Determining the type of lesion;
Determining a segmentation area including a position of the pointer by using a segmentation algorithm preset according to the determined lesion type in the determination unit of the segmentation device; And
Selecting the segmentation region as the lesion diagnosis region for the slice medical image by the determination unit of the segmentation apparatus;
Segmentation method in a medical image comprising a.
Extracting a lung region and a bone region from the slice medical image
Further comprising:
The determining step
If the profile meeting the extracted lung region of the angular profile is greater than a first predetermined reference value, the lesion type is determined as a lung tumor, and the profile meeting the extracted bone region of the angular profile is The method of segmentation in a medical image, characterized in that the type of lesion is determined as a brain tumor if greater than a second reference value.
The step of determining
Calculating a range of brightness values based on information of the slice medical image associated with the extracted position information of the pointer;
Determining a first segmentation area including a position of the pointer using a segmentation algorithm preset according to the calculated range of the brightness value and the determined lesion type;
Applying a preset fitting model to the determined first segmentation region; And
Determining an optimal segmentation region from the first segmentation region using the fitting model
Segmentation method in a medical image comprising a.
The determining of the first segmentation area may include
Determining a second segmentation area including a position of the pointer using the calculated range of brightness values;
Selecting the second segmentation area into a first profile that meets a previously extracted lung area and a second profile that does not meet the lung area when the determined lesion type is a lung tumor;
Interpolating the second profile to the first profile to form a fence; And
Determining the first segmentation region from the second segmentation region based on the first profile and the formed fence.
Segmentation method in a medical image comprising a.
The determining of the first segmentation area may include
Determining a second segmentation area including a position of the pointer using the calculated range of brightness values;
Selecting an anti-seed outside the second segmentation region when the determined lesion type is a brain tumor;
Forming an anti seed region based on the brightness value of the selected anti seed; And
Determining the first segmentation region from the second segmentation region by using the formed anti seed region.
Segmentation method in a medical image comprising a.
The determining step
When the brightness value of the location information of the pointer is greater than a predetermined threshold value, the lesion type is determined as a hyper tumor,
The step of determining
And the segmentation area of the medical image is determined as the segmentation area when the determined lesion type is a hyper tumor.
Displaying the determined segmentation area on the slice medical image in advance;
Further comprising:
The selecting step
And selecting the segmentation area as the lesion diagnosis area for the slice medical image when the segmentation area displayed in advance is selected by a user.
The extracting step
Detecting optimum position information in a predetermined peripheral area including brightness values of the extracted position information of the pointer and position information of the pointer,
The determining step
In the medical image, characterized in that the lesion type is determined based on at least one of the information of the slice medical image associated with the detected optimal position information and an angle profile based on the optimal position information of the pointer. Segmentation method.
Determining the selected lesion diagnosis region as a seed for segmentation of a 3D volume image;
Determining a segmentation region of each of the plurality of slice medical images associated with the slice medical image based on the determined seed; And
Generating a 3D segmentation volume by using a segmentation region of each of the determined seeds and the plurality of slice medical images;
Segmentation method in a medical image, characterized in that it further comprises.
A discrimination unit for determining a lesion type based on at least one of sliced medical image information related to the extracted position information of the pointer and an angle profile based on the position information of the pointer;
A determination unit to determine a segmentation area including a position of the pointer using a segmentation algorithm preset according to the determined lesion type; And
A selection unit for selecting the segmentation region as the lesion diagnosis region for the slice medical image
Segmentation device in a medical image comprising a.
Mask extractor for extracting lung and bone regions from the slice medical image
Further comprising:
The determining unit
If the profile meeting the extracted lung region of the angular profile is greater than a first predetermined reference value, the lesion type is determined as a lung tumor, and the profile meeting the extracted bone region of the angular profile is The segmentation apparatus in the medical image, characterized in that the type of lesion is determined as a brain tumor if greater than a second preset reference value.
The determining unit
Calculating a range of brightness values based on information of the slice medical image associated with the extracted position information of the pointer;
A first area determiner configured to determine a first segmentation area including a location of the pointer using the calculated range of brightness values;
A selection unit for selecting the first segmentation region into a first profile that meets a previously extracted lung region and a second profile that does not meet the lung region when the determined lesion type is a lung tumor;
A forming part interpolating the second profile to the first profile to form a fence;
A second region determiner which determines a second segmentation region from the first segmentation region based on the first profile and the formed fence;
An application unit applying a preset fitting model to the determined second segmentation area; And
A third region determiner configured to determine an optimal segmentation region from the second segmentation region using the fitting model
Segmentation device in a medical image, characterized in that it comprises a.
The determining unit
Calculating a range of brightness values based on information of the slice medical image associated with the extracted position information of the pointer;
A first area determiner configured to determine a first segmentation area including a location of the pointer using the calculated range of brightness values;
An anti-seed selection unit for selecting an anti-seed outside the first segmentation region when the determined lesion type is a brain tumor;
A forming unit forming an anti seed region based on the selected brightness value of the anti seed;
A second region determiner configured to determine a second segmentation region from the first segmentation region by using the formed anti seed region;
An application unit applying a preset fitting model to the determined second segmentation area; And
A third region determiner configured to determine an optimal segmentation region from the second segmentation region using the fitting model
Segmentation device in a medical image, characterized in that it comprises a.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020110095149A KR101185728B1 (en) | 2011-09-21 | 2011-09-21 | A segmentatin method of medical image and apparatus thereof |
PCT/KR2012/007178 WO2013042889A1 (en) | 2011-09-21 | 2012-09-06 | Method and device for performing segmentation in medical images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020110095149A KR101185728B1 (en) | 2011-09-21 | 2011-09-21 | A segmentatin method of medical image and apparatus thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
KR101185728B1 true KR101185728B1 (en) | 2012-09-25 |
Family
ID=47114090
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
KR1020110095149A KR101185728B1 (en) | 2011-09-21 | 2011-09-21 | A segmentatin method of medical image and apparatus thereof |
Country Status (2)
Country | Link |
---|---|
KR (1) | KR101185728B1 (en) |
WO (1) | WO2013042889A1 (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102099350B1 (en) * | 2019-06-07 | 2020-04-09 | 주식회사 뷰노 | Method for aiding quantification of lesions in medical imagery and apparatus using the same |
CN113712594A (en) * | 2020-05-25 | 2021-11-30 | 株式会社日立制作所 | Medical image processing apparatus and medical imaging apparatus |
WO2022139068A1 (en) * | 2020-12-22 | 2022-06-30 | 주식회사 딥노이드 | Deep learning-based lung disease diagnosis assistance system and deep learning-based lung disease diagnosis assistance method |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111784711A (en) * | 2020-07-08 | 2020-10-16 | 麦克奥迪(厦门)医疗诊断系统有限公司 | Lung pathology image classification and segmentation method based on deep learning |
WO2024046142A1 (en) * | 2022-08-30 | 2024-03-07 | Subtle Medical, Inc. | Systems and methods for image segmentation of pet/ct using cascaded and ensembled convolutional neural networks |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10192256A (en) | 1996-11-01 | 1998-07-28 | General Electric Co <Ge> | Image segmentation system and method to segment plural slice images into anatomical structure |
US20080269598A1 (en) | 2005-02-11 | 2008-10-30 | Koninklijke Philips Electronics N.V. | Identifying Abnormal Tissue in Images of Computed Tomography |
JP2008272480A (en) | 2007-05-07 | 2008-11-13 | General Electric Co <Ge> | Method and device for improving and/or validating 3d-segmentation |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101014563B1 (en) * | 2009-08-07 | 2011-02-16 | 주식회사 메디슨 | Ultrasound system and method for performing segmentation of vessel |
-
2011
- 2011-09-21 KR KR1020110095149A patent/KR101185728B1/en active IP Right Grant
-
2012
- 2012-09-06 WO PCT/KR2012/007178 patent/WO2013042889A1/en active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10192256A (en) | 1996-11-01 | 1998-07-28 | General Electric Co <Ge> | Image segmentation system and method to segment plural slice images into anatomical structure |
US20080269598A1 (en) | 2005-02-11 | 2008-10-30 | Koninklijke Philips Electronics N.V. | Identifying Abnormal Tissue in Images of Computed Tomography |
JP2008272480A (en) | 2007-05-07 | 2008-11-13 | General Electric Co <Ge> | Method and device for improving and/or validating 3d-segmentation |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102099350B1 (en) * | 2019-06-07 | 2020-04-09 | 주식회사 뷰노 | Method for aiding quantification of lesions in medical imagery and apparatus using the same |
CN113712594A (en) * | 2020-05-25 | 2021-11-30 | 株式会社日立制作所 | Medical image processing apparatus and medical imaging apparatus |
CN113712594B (en) * | 2020-05-25 | 2023-12-26 | 富士胶片医疗健康株式会社 | Medical image processing apparatus and medical imaging apparatus |
WO2022139068A1 (en) * | 2020-12-22 | 2022-06-30 | 주식회사 딥노이드 | Deep learning-based lung disease diagnosis assistance system and deep learning-based lung disease diagnosis assistance method |
Also Published As
Publication number | Publication date |
---|---|
WO2013042889A1 (en) | 2013-03-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10147223B2 (en) | Apparatus and method for computer-aided diagnosis | |
US8611623B2 (en) | Network construction apparatus, method and program | |
KR102154733B1 (en) | Apparatus and method for estimating whether malignant tumor is in object by using medical image | |
JP6059261B2 (en) | Intelligent landmark selection to improve registration accuracy in multimodal image integration | |
JP6530371B2 (en) | Interactive follow-up visualization | |
CA2776203C (en) | Determining contours of a vessel using an active contouring model | |
US9536316B2 (en) | Apparatus and method for lesion segmentation and detection in medical images | |
US9959622B2 (en) | Method and apparatus for supporting diagnosis of region of interest by providing comparison image | |
US8483462B2 (en) | Object centric data reformation with application to rib visualization | |
US10991102B2 (en) | Image processing apparatus and image processing method | |
KR101185728B1 (en) | A segmentatin method of medical image and apparatus thereof | |
EP1947606A1 (en) | Medical image processing apparatus and medical image processing method | |
JP2019530490A (en) | Computer-aided detection using multiple images from different views of the region of interest to improve detection accuracy | |
US8165376B2 (en) | System and method for automatic detection of rib metastasis in computed tomography volume | |
JP2015066311A (en) | Image processor, image processing method, program for controlling image processor, and recording medium | |
JP5388614B2 (en) | Medical image processing apparatus, image diagnostic apparatus, and medical image processing program | |
KR101185727B1 (en) | A segmentatin method of medical image and apparatus thereof | |
KR20210098381A (en) | Device and method for visualizing image of lesion | |
WO2012026090A1 (en) | Image display device, method and program | |
US10390799B2 (en) | Apparatus and method for interpolating lesion detection | |
EP4252657A1 (en) | Medical image abnormality detection system and abnormality detection method | |
KR101654824B1 (en) | Method and apparatus for controlling key point matching of medical images | |
US20080137928A1 (en) | Method and System for Registering CT Data Sets | |
JP2007512890A (en) | How to determine the structure of a moving object | |
JP6746639B2 (en) | Image processing apparatus and method, and diagnostic support system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
A201 | Request for examination | ||
A302 | Request for accelerated examination | ||
E902 | Notification of reason for refusal | ||
E701 | Decision to grant or registration of patent right | ||
GRNT | Written decision to grant | ||
FPAY | Annual fee payment |
Payment date: 20150915 Year of fee payment: 4 |
|
FPAY | Annual fee payment |
Payment date: 20160906 Year of fee payment: 5 |
|
FPAY | Annual fee payment |
Payment date: 20180830 Year of fee payment: 7 |
|
FPAY | Annual fee payment |
Payment date: 20190910 Year of fee payment: 8 |