Method and storage medium recording program for extracting ventricles of brain from computed tomography image

Abstract

According to the present embodiments, a method for extracting a ventricle from a CT image is provided, comprising: a step of binarizing a CT image based on a first reference value; a step of performing noise reduction on the binarized CT image based on a second reference value; a step of identifying the largest volume region in the CT image where noise reduction has been performed as a ventricle region; a step of generating a skeleton by performing skeletonization on a horizontal plane slice of the identified ventricle region; a step of determining whether the generated skeleton has one of a predetermined structural feature; and a step of extracting a ventricle from the structure of the generated skeleton if it is determined that the generated skeleton has one of the predetermined structural feature.

Description

Technology Field

[0001] The embodiments of the present disclosure relate to a storage medium recording a method and program for extracting ventricles from a CT image. Background Technology

[0002] Computerized Tomography (CT) is an imaging examination capable of accurately identifying the anatomical structure of the human body, and it holds high diagnostic value in the early diagnosis of diseases and determining the presence or absence of abnormalities in lesions. For example, head CT images can clearly reveal neuroimaging abnormalities such as cerebral hemorrhage or tumors, making it widely used for rapidly screening and diagnosing patient abnormalities. In particular, it is widely utilized for the early diagnosis of diseases due to its advantages of being faster and less expensive compared to other imaging methods such as Positron Emission Tomography (PET) or Magnetic Resonance Imaging (MRI).

[0003] The ventricles of the brain are four interconnected cavities located within the brain parenchyma and are filled with cerebrospinal fluid. Changes in the size and shape of the ventricles are considered indicators of various brain diseases, and observing them can help identify the causes of neurological abnormalities. For example, changes in ventricular size can be caused by hydrocephalus or the rupture of a cerebral aneurysm. Additionally, increased intracranial pressure (ICP) resulting from injury can lead to death or serious complications; therefore, it may be necessary to determine the location and size of the ventricles to assess the severity of ICP. Prior art literature

[65535] WO2004-077359 A1, "METHOD AND APPARATUS FOR EXTRACTING CEREBRAL VENTRICULAR SYSTEM FROM IMAGES"(2004.09.10)US 8831328 B2, "Method and system for segmenting a brain image"(2014.09.09)KR 10-2300231 B1, "A METHOD FOR PROVIDING DISEASE INFORMATION AND DEVICE PERFORMING THE SAME" (2021.09.03) The problem to be solved

[0004] Compared to other imaging tests, CT has low spatial resolution, making it difficult to segment brain regions and observe the presence of diseases. There is a need for measures to effectively utilize CT for disease diagnosis and early response by enabling rapid and accurate ventricular measurements through CT imaging. means of solving the problem

[0005] Ventricle extraction according to the embodiments enables accurate extraction of the region or location and size of the ventricles by detecting specific structural features in brain images through morphological operations. If the initial result of structural feature detection is inaccurate, a method or algorithm is provided to repeatedly perform the process until accurate structural feature detection is achieved by adjusting the threshold for binarization and the filtering size.

[0006] A method for extracting a ventricle from a CT image according to one embodiment comprises: a step of binarizing a CT image based on a first reference value; a step of performing noise reduction on the binarized CT image based on a second reference value; a step of identifying the largest volume region in the CT image where noise reduction has been performed as a ventricle region; a step of generating a skeleton by performing skeletonization on a horizontal plane slice of the identified ventricle region; a step of determining whether the generated skeleton has one of a predetermined structural feature; and a step of extracting a ventricle from the structure of the generated skeleton if it is determined that the generated skeleton has one of the predetermined structural feature.

[0007] According to one embodiment, the first reference value is a value of HU (Hounsfield Unit), the noise reduction is performed by median filtering, and the second reference value is the size of the median filtering.

[0008] According to one embodiment, the predetermined structural feature comprises: a first structure in which the generated skeleton includes four vertices and three lines connecting any one of the four vertices to the other three vertices on the generated skeleton are common only from any one vertex to a first node; and a second structure in which the generated skeleton includes four vertices and three lines connecting any one of the four vertices to the other three vertices on the generated skeleton are common from any one vertex to a first node and two of the three lines are common from the first node to a second node located at a different position from the first node.

[0009] According to one embodiment, the predetermined structural features include X-shaped and H-shaped forms.

[0010] According to one embodiment, if it is determined that the generated skeleton does not have one of the predetermined structural features, at least one of the first reference value and the second reference value is adjusted, and then binarization of the CT image, noise reduction, identification of the ventricular region, skeletonization is performed, and a determination of whether it has one of the predetermined structural features is performed again. Effects of the invention

[0011] The present invention enables the measurement (extraction) of the location and size of the ventricles by obtaining morphologically segmented images of the ventricles through an automated image processing method for brain CT images, thereby allowing brain CT images obtained through a CT device to be utilized for diagnosis, such as brain lesion screening. Brief explanation of the drawing

[0012] FIG. 1 is a flowchart of a method for extracting ventricles from a CT image according to one embodiment. Figure 2 shows examples of images in which the HU threshold was applied to extract the ventricles from CT images. Figure 3 shows examples of images with median filtering applied to CT images with the HU threshold applied. Figure 4 shows an example of skeletonization for a binary object. Figure 5 shows an example where the skeleton does not have predetermined structural features. FIG. 6 illustrates conditions 1 and 2 having predetermined structural features according to one embodiment. Specific details for implementing the invention

[0013] Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the embodiments, if it is determined that a detailed description of related known components or functions may obscure the essence of the technical concept, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless otherwise specified.

[0014] Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms.

[0015] In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another.

[0016] In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used.

[0017] Meanwhile, where numerical values ​​or corresponding information regarding a component (e.g., reference values, etc.) are mentioned, even without separate explicit notation, the numerical values ​​or corresponding information may be interpreted as including a range of error that may occur due to various factors (e.g., process factors, internal or external shocks, noise, etc.).

[0018] The embodiments are described in detail below with reference to the drawings.

[0019] FIG. 1 is a flowchart of a method for extracting ventricles from a CT image according to one embodiment.

[0020] The flowchart of Fig. 1 briefly illustrates the operations from S10 to S70.

[0021] In S10, the CT image is binarized based on a first reference value. The binarization may be performed based on a value (threshold) of HU (Hounsfield Unit). In this case, the first reference value is the value of HU. According to one embodiment, the value of HU may correspond to a predetermined value (e.g., 28). For example, when binarization is performed, parts having a value greater than or equal to the predetermined value may appear black, and parts having a value less than the predetermined value may appear white.

[0022] Figure 2 shows examples of images in which the HU threshold was applied to extract the ventricles from CT images.

[0023] Referring to Fig. 2, images in which HU thresholds of 15, 20, and 25 are applied to the original CT image are shown in sequence.

[0024] Mathematical Equation 1 represents the mathematical expression for binarizing CT images through a specific threshold of HU.

[0025]

[0026] In mathematical formula 1, src represents the input image, dst represents the output image, and T represents a specific threshold value, respectively.

[0027] In S20, noise reduction is performed on the binarized CT image based on a second reference value. In one embodiment, the noise reduction may be performed by median filtering. The second reference value may be the size of the median filtering. Median filtering is used to reduce noise in an image while simultaneously preserving the properties of the original image. It outputs the value of the original pixel (the corresponding pixel) by replacing it with the median value among the values ​​of a pixel in the image and the neighbor values ​​adjacent to that pixel, while the values ​​of the pixel and neighbors adjacent to that pixel are sorted. The size of the median filtering represents the range of neighbors adjacent to a point and may correspond, for example, to the number of surrounding pixels used when calculating the median value from each pixel. The larger the window size of the median filtering, the smoother or blurrier the image becomes. According to one embodiment, the size of the median filtering here may correspond to a predetermined initial value (e.g., 3).

[0028] Figure 3 shows examples of images with median filtering applied to CT images with the HU threshold applied.

[0029] Referring to Fig. 3, CT images with a HU threshold (HU 20) applied and median filtering sizes of 3, 5, and 7 applied, respectively, are shown in sequence.

[0030] Mathematical Equation 2 represents the mathematical expression for binarizing CT images through a specific threshold of HU.

[0031]

[0032] In mathematical formula 2, y(n,m) represents the output point, x(n,m) represents the input point, N represents the window size, and median represents a function that calculates the median of a given set.

[0033] In S30, the largest volume area in the CT image where noise reduction has been performed is identified (specifically identified) as the ventricular region. The CT image where noise reduction has been performed may be divided into multiple spaces (regions), and among them, the largest volume area (widest area) may be identified as the ventricular space (ventricular region). Each space (region) is, for example, an independent space composed of continuous white and can be easily distinguished from other spaces. If the CT image where noise reduction has been performed is not divided into multiple regions but has only one region, the entire area of ​​the CT image may be identified as the ventricular region.

[0034] In various embodiments of the present invention, the CT image that is binarized, the CT image that undergoes noise reduction, and the identification of the ventricular region are assumed to be processed in three dimensions, but cases where they are processed in two dimensions may also be included in embodiments of the present invention.

[0035] In S40, a skeleton is generated by performing skeletonization on a horizontal plane slice (two-dimensional cross-section) of the identified ventricular region. Skeletonization refers to reducing a binary object to represent it as an extension of pixels. In various embodiments of the present invention, the skeleton may be treated as a plurality of adjacent points coming together to form one or more lines.

[0036] Figure 4 shows an example of skeletonization for a binary object.

[0037] Referring to Fig. 4, it can be seen that a stelton (right) was generated by performing skeletonization on the original image (left) of a horse and a plant.

[0038] In S50, it is determined whether the generated skeleton has one of the predetermined structural features. In various embodiments of the present invention, the predetermined structural features are determined in two dimensions, and a point that has one or two adjacent points in two dimensions corresponds to a point that constitutes a line, and in particular, a point that has one adjacent point in two dimensions corresponds to a vertex of a line.

[0039] In one embodiment, the predetermined structural feature may include: a first structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common only from any one vertex to a first node; and a second structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common from any one vertex to a first node, and two of the three edges are common from the first node to a second node located at a different position from the first node. In this case, it is determined whether the generated skeleton has one of the structural features of the first structure and the second structure.

[0040] In another embodiment, the predetermined structural feature may be a structure in which the generated skeleton includes four vertices, an edge connecting the first vertex and the second vertex among the four vertices and an edge connecting the third vertex and the fourth vertex among the four vertices intersect at a single node, and a first shortest edge connecting the first vertex and the second vertex on the generated skeleton is common to a second shortest edge connecting the third vertex and the fourth vertex on the generated skeleton at a single node or a single edge. Alternatively, the predetermined structural feature may include an X-shape and an H-shape. In this case, it is determined whether the generated skeleton has one of the structural features of an X-shape and an H-shape.

[0041] Figure 5 shows an example where the skeleton does not have predetermined structural features.

[0042] In the example of Fig. 5, the skeleton has an 11 shape, and it is determined that this does not correspond to a predetermined structural feature.

[0043] The above-determined structural features are not limited to the embodiments described above, and may be implemented by determining in S50 whether the skeleton has other structural features other than the structural features described above.

[0044] In S60, if it is determined that the generated skeleton has one of the predetermined structural features, a ventricle is extracted from the structure of the generated skeleton. In one embodiment, the extraction of the ventricle may include specifying the location and size of the ventricle.

[0045] In S70, if it is determined that the generated skeleton does not have one of the predetermined structural features (if it is determined that it does not correspond to the predetermined structural features), at least one of the first reference value (e.g., HU) and the second reference value (e.g., the size of the median filtering) can be adjusted, and binarization of the CT image in S10, noise reduction in S20, identification of the ventricular region in S30, skeletonization in S40, and determination of whether it has one of the predetermined structural features in S50 can be performed again by the adjusted first reference value and / or the adjusted second reference value.

[0046] The HU value required to distinguish the ventricles from other tissues varies from person to person. For example, for person A, a value of 25 HU or less may be considered a ventricle, while for person B, it may be 18 HU or less. Therefore, to avoid overestimating or underestimating the ventricles based on HU, it is necessary to apply different HU thresholds (reference values) to each individual for judgment. Similar to the HU reference value, the size of the median filtering (smoothing size) can also be adaptively adjusted to accurately estimate the ventricles.

[0047] According to various embodiments, adjusting at least one of the first reference value (HU threshold) and the second reference value (size of median filtering) may correspond to raising one of the first reference value and the second reference value, each having a predetermined range, by a predetermined unit or lowering it to an initial value. For example, an example is described below in which the first reference value has a range of 28 to 18 as the HU threshold and is reduced by 2 units according to the adjustment, and the second reference value has a range of 3 to 7 as the size of median filtering and is increased by 1 unit according to the adjustment.

[0048] The following is a description of an example of an adjustment of at least one of a first reference value and a second reference value in a method for extracting ventricles from a CT image. In this example, the first reference value is reduced by a predetermined unit (2) every time the adjustment is performed 5 times from a maximum value (28) to a minimum value (10), and the previous value is maintained in the remaining adjustments. In this example, the second reference value is increased by a predetermined unit (1) every time the adjustment is performed from a minimum value (3) to a maximum value (7) (if the adjustment was made to a maximum value in the previous adjustment, it is adjusted from a maximum value to a minimum value), and this process is repeated. In this embodiment, the initial value of the first reference value of S10 is set to HU threshold = 28 and the initial value of the second reference value of S20 is set to correspond to the size of the median filtering = 3. Then, in S70, it is determined that one of the predetermined structural features is not present, and S10 through S50 are performed again after adjustment of at least one of the first reference value and the second reference value. In this case, the second first reference value of S10 is maintained at HU threshold = 28 and the second second reference value of S20 is set to the size of the median filtering = 4 (increased by 1). Until the adjustment of at least one of the first reference value and the second reference value of S70 is performed 4 times, the first reference value of S10 is maintained at HU threshold = 28 and the second reference value of S20 is adjusted to the size of the median filtering = 7. Since the second reference value has been adjusted to the maximum value of a predetermined range, when adjusting at least one of the first reference value and the second reference value again in S70, the first reference value is reduced by a predetermined unit (2) to a HU threshold of 26, and the second reference value is adjusted to the size of the median filtering, which is the minimum value of the predetermined value, to 3. After that, whenever adjusting at least one of the first reference value and the second reference value, the second reference value is increased by a predetermined unit (1), and the first reference value is maintained at a HU threshold of 26 until it is adjusted to the size of the median filtering, which is the maximum value of the predetermined value, to 7.

[0049] If the process of performing adjustments to at least one of the first reference value and the second reference value of S70 is repeated without determining that the generated skeleton has one of the aforementioned predetermined structural features, the first reference value is adjusted to a minimum value (HU threshold = 18) and the second reference value is adjusted to a maximum value (size of median filtering = 7), and the algorithm terminates without extracting the ventricles from the CT image.

[0050] That is, the algorithm can be implemented such that, even though adjustment of at least one of the first reference value and the second reference value is repeated within a given range, if there is no case where it is determined that the generated skeleton has one of the predetermined structural features until the first reference value and the second reference value each reach a final value, the algorithm terminates without extracting the ventricles from the CT image.

[0051] In one embodiment, if the algorithm terminates because it is not determined that the generated skeleton has one of the predetermined structural features, the algorithm may be performed again by changing the horizontal plane slice of the identified ventricular region (changing the position of the cross-section).

[0052] Whether the generated skeleton possesses predetermined structural characteristics can be determined by utilizing graph theory.

[0053] FIG. 6 illustrates conditions 1 and 2 having predetermined structural features according to one embodiment.

[0054] In the example of Figure 6, condition 1 having a predetermined structural feature is that vertex v5 is adjacent to v1, v2, v3, and v4, respectively, and condition 2 having a predetermined structural feature is that vertex v5 is adjacent to v6, v5 is adjacent to v1 and v2, and v6 is adjacent to v2 and v4. Here, two vertices being adjacent can be defined as having only one line connecting the two vertices on the skeleton.

[0055] The embodiments described above may be implemented through various means. For example, the embodiments may be implemented by hardware, firmware, software, or a combination thereof.

[0056] In the case of implementation by hardware, the method for extracting ventricles from CT images according to the embodiments may be implemented by one or more ASICs (Application Specific Integrated Circuits), DSPs (Digital Signal Processors), DSPDs (Digital Signal Processing Devices), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), processors, controllers, microcontrollers, or microprocessors.

[0057] In the case of implementation by firmware or software, the method for extracting ventricles from CT images according to the embodiments may be implemented in the form of a device, procedure, or function that performs the functions or operations described above. The software code may be stored in a memory unit and executed by a processor. The memory unit may be located inside or outside the processor and may exchange data with the processor by various means already known.

[0058] Additionally, terms such as "system," "processor," "controller," "component," "module," "interface," "model," or "unit" described above may generally refer to computer-related entities, hardware, combinations of hardware and software, software, or running software. For example, the aforementioned components may be, but are not limited to, processes driven by a processor, processors, controllers, control processors, objects, execution threads, programs, and / or computers. For example, both the application running on the controller or processor and the controller or processor may be components. One or more components may reside within a process and / or execution thread, and the components may be located on a single device (e.g., a system, a computing device, etc.) or distributed across two or more devices.

[0059] Meanwhile, another embodiment provides a computer program stored on a computer recording medium for performing a method of extracting ventricles from the aforementioned CT image. Additionally, another embodiment provides a computer-readable recording medium that records a program for realizing the method of extracting ventricles from the aforementioned CT image. The program recorded on the recording medium can execute the aforementioned steps by being read, installed, and executed on a computer.

[0060] In this way, for a computer to read a program recorded on a recording medium and execute functions implemented in the program, the aforementioned program may include code encoded in computer languages ​​such as C, C++, JAVA, and machine language, which can be read by the computer's processor (CPU) through the computer's device interface. The method for extracting ventricles from the aforementioned CT image can be classified as an image processing algorithm and can be implemented through functions of computer vision processing packages such as Python's Numpy, OpenCV, Scikit-image, and SimpleITK.

[0061] Such code may include functional code related to functions that define the aforementioned functions, and may also include control code related to execution procedures necessary for a computer processor to execute the aforementioned functions according to a predetermined procedure.

[0062] In addition, this code may further include memory reference-related code regarding where (address) in the computer's internal or external memory additional information or media required for the computer's processor to execute the aforementioned functions should be referenced.

[0063] In addition, if the computer processor needs to communicate with any other computer or server located remotely in order to execute the aforementioned functions, the code may further include communication-related code regarding how the computer processor should communicate with any other computer or server located remotely using the computer's communication module, and what information or media should be transmitted or received during communication.

[0064] A computer-readable recording medium that has recorded a program as described above includes, for example, ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical media storage device, etc., and may also include one implemented in the form of a carrier wave (for example, transmission via the Internet).

[0065] In addition, computer-readable recording media are distributed across networked computer systems, allowing computer-readable code to be stored and executed in a distributed manner.

[0066] Furthermore, the functional program for implementing the present invention, and the related code and code segments, etc., may be easily inferred or modified by programmers skilled in the art to which the present invention belongs, taking into account the system environment of a computer that reads a recording medium and executes the program.

[0067] The method for extracting the ventricles from the aforementioned CT image may also be implemented in the form of a recording medium containing computer-executable instructions, such as applications or program modules executed by a computer. A computer-readable medium may be any available medium accessible by a computer and includes both volatile and non-volatile media, as well as removable and inseparable media. Additionally, a computer-readable medium may include all computer storage media. Computer storage media include both volatile and non-volatile, removable and inseparable media implemented by any method or technique for storing information, such as computer-readable instructions, data structures, program modules, or other data.

[0068] The method for extracting ventricles from the aforementioned CT image may be executed by an application installed by default on the terminal (which may include a program included in a platform or operating system installed by default on the terminal), or by an application (i.e., a program) directly installed by the user on the master terminal through an application provider server, such as an application store server, an application, or a web server related to the service. In this sense, the method for extracting ventricles from the aforementioned CT image may be implemented by an application (i.e., a program) that is installed by default on the terminal or directly installed by the user, and may be recorded on a computer-readable recording medium such as the terminal.

[0069] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will understand that other specific forms can be easily modified without altering the technical spirit or essential features of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.

[0070] The scope of the present invention is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts thereof should be interpreted as being included within the scope of the present invention.

[0071] The foregoing description is merely an illustrative explanation of the technical concept of the present disclosure, and those skilled in the art to which the present disclosure pertains may make various modifications and variations within the scope of the essential characteristics of the technical concept. Furthermore, since these embodiments are intended to explain, not limit, the scope of the technical concept is not limited by these embodiments. The scope of protection of the present disclosure shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present disclosure.

Claims

Claim 1 A method for extracting ventricles from a CT image executed by a computer processor, comprising: a step of binarizing the CT image based on a first reference value; a step of performing noise reduction on the binarized CT image based on a second reference value; a step of identifying the largest volume region in the noise-reduced CT image as a ventricular region; a step of generating a skeleton by performing skeletonization on a horizontal plane slice of the identified ventricular region; a step of determining whether the generated skeleton has one of a predetermined structural feature; and, if it is determined that the generated skeleton has one of the predetermined structural feature, extracting the ventricles from the structure of the generated skeleton, and if it is determined that the generated skeleton does not have one of the predetermined structural feature, adjusting at least one of the first reference value and the second reference value, and then performing the binarization, noise reduction, ventricular region identification, skeleton generation, and determination steps again. Claim 2 A method for extracting ventricles from a CT image, wherein, in claim 1, the first reference value is a value of HU (Hounsfield Unit), the noise reduction is performed by median filtering, and the second reference value is the size of the median filtering. Claim 3 A method for extracting ventricles from a CT image, wherein the predetermined structural features include: a first structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common only from any one vertex to a first node; and a second structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common from any one vertex to a first node and two of the three edges are common from the first node to a second node located at a different position from the first node. Claim 4 A method for extracting ventricles from a CT image, wherein the predetermined structural features of claim 1 include X-shaped and H-shaped shapes. Claim 5 delete Claim 6 A computer-readable storage medium storing a program for extracting ventricles from a CT image, wherein when the program is executed by a computer processor, the computer processor performs the following operations: binarizing the CT image based on a first reference value; performing noise reduction on the binarized CT image based on a second reference value; identifying the largest volume region in the noise-reduced CT image as the ventricular region; generating a skeleton by performing skeletonization on a horizontal plane slice of the identified ventricular region; and determining whether the generated skeleton has one of a predetermined structural features. A computer-readable storage medium having a program that performs the operation of extracting a ventricle from the structure of the generated skeleton if it is determined that the generated skeleton has one of the predetermined structural features, and if it is determined that the generated skeleton does not have one of the predetermined structural features, adjusting at least one of the first reference value and the second reference value, and then performing the binarization, noise reduction, ventricle region identification, skeleton generation, and determination operation again. Claim 7 A computer-readable storage medium having a program written thereon, wherein, in claim 6, the first reference value is a value of HU (Hounsfield Unit), the noise reduction is performed by median filtering, and the second reference value is the size of the median filtering. Claim 8 A computer-readable storage medium storing a program, wherein the predetermined structural features include: a first structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common only from any one vertex to a first node; and a second structure in which the generated skeleton includes four vertices and three edges connecting any one of the four vertices to the other three vertices on the generated skeleton are common from any one vertex to a first node and two of the three edges are common from the first node to a second node located at a different position from the first node. Claim 9 In paragraph 6, the above-mentioned predetermined structural features include X-shaped and H-shaped forms, a computer-readable storage medium that records a program. Claim 10 delete

Images