KR102298249B1 - Method and apparatus for brain image correction using brain structure - Google Patents

Abstract

A brain image correction method using a brain structure performed by a computer is provided. The method includes obtaining a head image including the brain of a subject, dividing the head image into a plurality of regions based on a brain structure, and performing correction using a layer arrangement condition of the brain for the plurality of regions comprising the steps of performing

Description

Brain image correction method and apparatus using brain structure {METHOD AND APPARATUS FOR BRAIN IMAGE CORRECTION USING BRAIN STRUCTURE}

The present invention relates to a method and apparatus for correcting a brain image using a brain structure.

Electrical brain stimulation refers to a method of finally applying an electric current to the brain by attaching electrodes inside or outside the head and allowing current to flow. Electrical brain stimulation is a non-invasive treatment method that can be performed simply, and is widely used to treat various brain diseases depending on the location and type of stimulation.

In addition, an electroencephalogram (EEG) capable of measuring an electrical activity according to a brain activity of a subject is also widely used in neurology and neuropsychiatric treatment.

Electrical brain stimulation and EEG electroencephalography are both non-invasive tests and treatment methods, and have the advantage of being simple to operate. In addition, in order to perform such a procedure, a brain image of each subject to be treated is acquired, and treatment is performed through the acquired brain image. At this time, it is important to obtain a brain image suitable for the purpose of the procedure, and in particular, it is important to obtain a brain image obtained by classifying each brain region to match the actual structure of the brain. Therefore, it is required to develop a method for classifying each brain region to match the actual structure of the brain through brain imaging.

An object of the present invention is to provide a method and apparatus for correcting a brain image using a brain structure.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method and apparatus for acquiring a brain image divided into a plurality of brain regions conforming to the layer arrangement structure of the brain.

An object of the present invention is to provide a method and apparatus for detecting and correcting a brain region that does not satisfy the layer arrangement structure of the brain in a brain image.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

A method of correcting a brain image using a brain structure performed by a computer according to an embodiment of the present invention includes: acquiring a head image including the brain of a subject; dividing the head image into a plurality of regions based on the brain structure and performing correction using a layer arrangement condition of the brain for the plurality of regions.

In an embodiment of the present invention, the dividing into the plurality of regions may include dividing the brain in the head image into the plurality of regions by using a learning model that labels the brain based on the brain structure.

In an embodiment of the present invention, each of the plurality of regions may correspond to each brain region labeled based on a brain structure through the learning model.

In an embodiment of the present invention, the performing of the correction includes determining whether the plurality of regions meet the layer arrangement condition of the brain, and the layer arrangement condition of the brain among the plurality of areas and extracting and correcting a region that does not match, and the layer arrangement condition of the brain may be set based on the layer structure of the brain arranged in the order of skin, skull, cerebrospinal fluid, and brain internal regions.

In one embodiment of the present invention, the layer arrangement condition of the brain is a first condition that a layer disposed outside of a layer corresponding to the skin should not exist, a layer corresponding to the cerebrospinal fluid is in contact with a layer corresponding to the skin The second condition not to be, the layer corresponding to the inner brain region does not come into contact with the layer corresponding to the skull or skin, the third condition, the layer corresponding to the white matter in the brain inner region is a layer corresponding to the gray matter in the inner brain region It may include at least one of a fourth condition that should not exist outside and a fifth condition that an arrangement distribution of all layers in the layer structure of the brain must exist within a certain range.

In one embodiment of the present invention, the step of extracting and correcting a region that does not meet the layer arrangement condition of the brain includes the layer arrangement condition of the brain so as to match the order of the skin, skull, cerebrospinal fluid, and brain internal region. Areas that do not match can be relocated.

In one embodiment of the present invention, the method may further include performing 3D brain modeling of the object based on the head image corrected using the layer arrangement condition of the brain.

In one embodiment of the present invention, the method may further include simulating electrical stimulation of the brain of the object based on the head image corrected using the layer arrangement condition of the brain.

An apparatus according to an embodiment of the present invention includes a memory storing one or more instructions, and a processor executing the one or more instructions stored in the memory, wherein the processor executes the one or more instructions by executing the one or more instructions. Obtaining a head image comprising: dividing the head image into a plurality of regions based on the brain structure; .

The computer program according to an embodiment of the present invention is stored in a computer-readable recording medium so as to be combined with a computer, which is hardware, to perform the brain image correction method using the brain structure.

According to the present invention, it is possible to generate a more accurate brain image divided into brain regions by dividing the brain into regions according to the layer arrangement structure of the brain and performing correction for each divided brain region.

According to the present invention, by dividing the brain into regions using the layer arrangement structure of the brain, it is possible to obtain a brain image classified into a brain region more suitable for the actual brain structure.

According to the present invention, by acquiring a brain image corrected for each brain region to match the layer structure of the brain, it is possible to obtain a target point to apply a more accurate electrical stimulation during electrical brain stimulation performed to treat various brain diseases. have. In addition, the therapeutic effect can also be improved by obtaining an accurate electrical stimulation point.

According to the present invention, each brain region can be effectively classified in a brain image by using a learning model in segmenting each brain region according to the layer structure of the brain.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a flowchart schematically illustrating a brain image correction method using a brain structure according to an embodiment of the present invention.
2 to 7 show an example of dividing the brain based on the brain structure and labeling each brain region according to an embodiment of the present invention.
8 is a diagram illustrating examples of results of dividing a head image into a plurality of regions according to an embodiment of the present invention.
9 is a diagram schematically showing the configuration of an apparatus 200 for performing a brain image correction method using a brain structure according to an embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully understand the scope of the present invention to those skilled in the art, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. As used herein, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless specifically defined explicitly.

As used herein, the term “unit” or “module” refers to a hardware component such as software, FPGA, or ASIC, and “unit” or “module” performs certain roles. However, “part” or “module” is not meant to be limited to software or hardware. A “unit” or “module” may be configured to reside on an addressable storage medium or to reproduce one or more processors. Thus, by way of example, “part” or “module” refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, Includes procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Components and functionality provided within “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules” or as additional components and “parts” or “modules”. can be further separated.

As used herein, the term “computer” includes various devices capable of providing a result to a user by performing arithmetic processing. For example, computers include desktop PCs and notebooks (Note Books) as well as smart phones, tablet PCs, cellular phones, PCS phones (Personal Communication Service phones), synchronous/asynchronous A mobile terminal of International Mobile Telecommunication-2000 (IMT-2000), a Palm Personal Computer (PC), a Personal Digital Assistant (PDA), and the like may also be applicable. Also, when a head mounted display (HMD) device includes a computing function, the HMD device may be a computer. In addition, the computer may correspond to a server that receives a request from a client and performs information processing.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a flowchart schematically illustrating a brain image correction method using a brain structure according to an embodiment of the present invention.

Although the method of FIG. 1 is described as being performed by a computer for convenience of description, the subject performing each step is not limited to a specific device and may be used to encompass devices capable of performing computing processing. That is, in the present embodiment, the computer may refer to a device capable of performing the brain image correction method using the brain structure according to the embodiment of the present invention.

Referring to FIG. 1 , the method for correcting a brain image using a brain structure according to an embodiment of the present invention includes acquiring a head image including the brain of a subject ( S100 ), and a plurality of the head images based on the brain structure. It may include dividing into regions of (S200), and performing correction using the layer arrangement condition of the brain for the plurality of regions (S300). Hereinafter, detailed description of each step is described.

The computer may acquire a head image including the subject's brain (S100).

Here, the object may include a person or an animal, or a part of a person or a part of an animal.

The head image refers to a medical image obtained by photographing the head including the brain of a subject. For example, a computed tomography (CT) image and a magnetic resonance imaging (MRI) image taken with a medical imaging device. , positron emission tomography (PET) images, and the like.

Also, the head image may be a medical image captured including the subject's brain as well as the subject's skull and scalp (skin).

The computer may segment the image of the head of the object into a plurality of regions based on the brain structure (S200).

As an embodiment, the computer may divide the head image of the object for each brain region based on the brain structure.

Here, the brain structure may have a layered structure, for example, a scalp (skin), a skull, cerebrospinal fluid, and an inner brain region. In addition, the brain inner region may include the cerebrum, cerebellum, and ventricle, and in the case of the cerebrum and cerebellum, it may be formed of a more subdivided layer structure into gray matter and white matter.

That is, the computer divides the head image of the subject into a plurality of regions corresponding to the layer structure of the brain based on the scalp, skull, cerebrospinal fluid, and internal regions of the brain (gray/white matter of the cerebrum, gray matter/white matter of the cerebellum, ventricle) can do.

Here, the layer structure of the brain has been described as consisting of the scalp, skull, cerebrospinal fluid, and brain internal regions (gray matter/white matter of the cerebrum, gray matter/white matter of the cerebellum, ventricle), but this is only an example, and the present invention is limited thereto it is not Depending on the embodiment, the structure of the brain may be classified in different ways, and the types of brain regions may be different depending on the classification method.

Also, the computer may divide the image of the head of the object into a plurality of regions through learning. As an embodiment, the computer may divide the brain included in the head image of the object into a plurality of regions by using a learning model that labels the brain for each region based on the brain structure. Here, the learning model may be derived through learning using deep learning.

For example, the computer may acquire medical images obtained by photographing the head including the brain from a plurality of objects, and perform learning (eg, convolutional neural network; learning using CNN) by using them as learning data. In this case, the computer may divide the brain into a plurality of regions based on the layer structure of the brain, and learn the learning data to label each of the plurality of divided regions. And, the computer can derive a learning model by learning the learning data. That is, the computer performs learning based on the learning data, and based on the layer structure of the brain, a learning model for labeling the brain for each region can be built in advance. Accordingly, the computer may input a head image of a specific object to the learning model as an input value, and obtain an output value obtained by dividing the head image of a specific object into a plurality of regions from the learning model. In this case, the plurality of output regions may correspond to each brain region labeled based on the layer structure of the brain through the learning model.

2 to 7 show an example of dividing the brain based on the brain structure and labeling each brain region according to an embodiment of the present invention.

As described above, the brain structure may be formed of a layer structure arranged in the order of the scalp, skull, cerebrospinal fluid, and brain internal regions (gray matter/white matter of the cerebellum, gray matter/white matter of the cerebellum, and ventricle). Therefore, the computer divides the brain into a plurality of brain regions as shown in FIGS. 2 to 7 based on this layer arrangement structure of the brain, and assigns a label value (ie, an identification value) to each divided region to label. can

In an embodiment, the computer may subdivide brain internal regions into cerebrum, cerebellum, and ventricle, and may assign a specific label value to each subdivided region.

Figure 2 (a) shows a region 10 corresponding to cerebral gray matter (Cerebral Gray Matter) in a head image including the brain, (b) of Figure 2 is a region corresponding to cerebral white matter (Cerebral White Matter) (11) is shown. For example, the computer may set the label value to 1 for the cerebral gray matter region 10, which is the outer region of the cerebrum, and set the label value to 2 for the cerebral white matter region 11 filling the inner region of the cerebrum. have.

Figure 3 (a) shows the region 20 corresponding to the cerebellar gray matter (Cerebellar Gray Matter) in the head image including the brain, Figure 3 (b) is the region corresponding to the cerebellar white matter (Cerebellar White Matter) (21) is shown. For example, the computer may set the label value to 3 for the cerebellar gray matter region 20, which is an outer region of the cerebellum, and set the label value to 4 for the cerebellar white matter region 21 filling the inner region of the cerebellum. have.

Figure 4 shows the region 30 corresponding to the ventricles (Ventricles; Lateral Ventricles) in the head image including the brain. For example, the computer may set the label value to 5 for the ventricular region 30 corresponding to an empty space connected to each other in the brain internal region.

5 shows a region 40 corresponding to cerebrospinal fluid (CSF) in a head image including the brain. Here, the cerebrospinal fluid 40 refers to the liquid filling the space between the outside of the cerebral / cerebellar gray matter and the skull. According to an embodiment, the region 40 corresponding to the cerebrospinal fluid may include a superior saggital sinus and a transverse sinus. In addition, the sickle membrane may be classified as part of the cerebrospinal fluid. For example, the computer may set the label value to 6 for the cerebrospinal fluid region 40 .

6 shows a region 50 corresponding to a skull in a head image including a brain. For example, the computer may set the label value to 7 for the cranial region 50 .

7 shows a region 60 corresponding to the skin in the head image including the brain. For example, the computer may set the label value to 8 for the skin region 60 .

2 to 7 , the computer may divide the brain into eight brain regions labeled with label values 1 to 8 based on the brain structure. Here, the division into eight brain regions may be based on a clinical layer structure. Accordingly, the computer may divide the head image of the specific object obtained in step S100 into a plurality of regions to correspond to the brain region having a preset label value.

In this case, the computer may derive a learning model divided into eight brain regions labeled with label values 1 to 8 as shown in FIGS. 2 to 7 through learning using deep learning as described above. In this case, the computer may input the head image of the specific object obtained in step S100 to the learning model, and as a result obtain a head image divided into eight brain regions labeled with label values 1 to 8.

Referring back to FIG. 1 , the computer may perform correction for the plurality of regions divided in step S200 by using the layer arrangement condition of the brain ( S300 ).

In one embodiment, the computer determines whether the brain layer arrangement condition is satisfied with respect to the head image divided into a plurality of regions, and performs correction by extracting a region that does not meet the layer arrangement condition of the brain from among the plurality of regions can do.

Here, as described above, the layer arrangement condition of the brain is based on the layer structure of the brain in which the brain is arranged in the order of the skin, skull, cerebrospinal fluid, and inner brain regions (gray matter/white matter of the cerebellum, gray matter/white matter of the cerebellum, ventricle). It may be a condition set to . For example, as shown in FIGS. 2 to 7 , the computer may set a condition to have an arrangement relationship consistent with the layer structure of the brain divided into eight brain regions labeled with label values 1 to 8 .

As an embodiment, the layer arrangement condition of the brain may include at least one of the following first to fifth conditions.

The first condition is that a layer disposed outside of the layer corresponding to the skin should not exist. For example, an area having any label value cannot be disposed outside the area corresponding to the label value 8 (skin in FIG. 7 ).

The second condition is that the layer corresponding to the cerebrospinal fluid should not come into contact with the layer corresponding to the skin. For example, the region corresponding to the label value 6 (cerebrospinal fluid in FIG. 5) should be disposed so as not to directly contact the region corresponding to the label value 8 (skin in FIG. 7).

The third condition is that the layer corresponding to the inner region of the brain does not come into contact with the layer corresponding to the skull or skin. For example, a region corresponding to a label value of 1 to 5 (cerebrum in FIG. 2, cerebellum in FIG. 3, ventricle in FIG. 4) is labeled value 7 (skull in FIG. 6) or label value 8 (skin in FIG. 7). It should be placed so that it does not come into direct contact with the relevant area.

The fourth condition is that the layer corresponding to the white matter in the inner brain region does not exist outside the layer corresponding to the gray matter (cortex) in the inner brain region. For example, the region corresponding to the label value 2 (cerebral white matter of Fig. 2 (b)) should not be disposed outside the region corresponding to the label value 1 (cerebral gray matter of Fig. 2 (a)). In addition, the region corresponding to the label value 4 (cerebellar white matter of Fig. 3 (b)) should not be disposed outside the region corresponding to the label value 3 (cerebellar gray matter of Fig. 3 (a)).

The fifth condition is that the arrangement distribution of all layers in the layer structure of the brain must exist within a certain range. For example, it is necessary to minimize all labels (label values 1 to 8) and label values that are not distributed within a certain range and are distributed outside a certain range.

That is, the computer has an arrangement relationship that matches the layer structure of the brain (eg, the label values shown in FIGS. 2 to 7 ) for the head image divided into a plurality of regions based on the first to fifth conditions described above. It can be determined whether The computer may extract a region having an arrangement relationship that does not match the layer structure of the brain from the head image according to the determination result, and correct the extracted region. For example, when correcting a region that does not satisfy the layer arrangement condition of the brain, the computer may rearrange the corresponding region to match the order of the skin, skull, cerebrospinal fluid, and internal regions of the brain.

8 is a diagram illustrating examples of results of dividing a head image into a plurality of regions according to an embodiment of the present invention. For example, FIG. 8 may be a result of performing step S200.

As described above, after the computer divides the image of the head of the object into a plurality of brain regions, the computer may determine whether the brain layer arrangement conditions (first to fifth conditions) are satisfied for each of the divided brain regions. .

As an example, in the case of the head image 100 divided into a plurality of regions as shown in FIG. It may be recognized that another layer (eg, skull) 101 exists outside the layer corresponding to , and thus it may be determined that the first condition is not satisfied. In this case, the computer may correct the layer region (eg, skull) 101 that does not satisfy the first condition to match the layer structure of the brain. For example, the computer may correct the head image so that the layer region (eg, the skull) 101 that does not satisfy the first condition is disposed as the inner region of the skin.

At this time, in determining whether the first condition is satisfied as an embodiment, the computer first acquires a skin layer area including a difference area by subtracting the original skull area before expansion from the area in which the skull is dilated (binary dilation) can do. At this time, the skin layer region completely covers the skull. Next, the computer may extract a part that newly surrounds the skull from among the skin layer regions including the difference region, and determine the extracted part as a 'region existing outside the skin'. In this case, by extracting the 'region existing outside the skin' from the head image, the computer can determine that the first condition in which a layer disposed outside of a layer corresponding to the skin does not exist is not satisfied.

As another example, in the case of the head image 110 divided into a plurality of regions as shown in (b) of FIG. 8 , the computer calculates the cerebrospinal fluid from the head image 110 based on the layer arrangement conditions (first to fifth conditions) of the brain. It is recognized that a portion 113 in which a partial region of the layer 111 corresponding to the skin is in contact with a partial region of the layer 112 corresponding to the skin exists, and thus it can be determined that the second condition is not satisfied. . In this case, the computer may correct the contact portion 113 that does not satisfy the second condition to match the layer structure of the brain. For example, the computer may correct the head image so that the contact portion 113 is disposed as an inner region of the skin.

At this time, in determining whether the second condition is satisfied as an embodiment, the computer first obtains a region of the skull layer including a difference region by subtracting the original cerebrospinal fluid region before dilation from the region in which the cerebrospinal fluid is dilated (binary dilation). can do. At this time, the skull layer region completely covers the cerebrospinal fluid. Next, the computer may extract a part that newly surrounds the cerebrospinal fluid from among the skull layer regions including the difference region, and determine that the extracted part is 'the part where the cerebrospinal fluid is in direct contact with the skin'. In this case, the computer extracts the 'part where the cerebrospinal fluid is in direct contact with the skin' from the head image, so that the layer corresponding to the cerebrospinal fluid does not satisfy the second condition that the layer corresponding to the skin should not be in contact.

As another example, the computer determines that a part of the layer corresponding to the inner brain region (eg, the brain region corresponding to the label value 1 to 5) in the head image is the skull based on the layer arrangement conditions (1st to 5th conditions) of the brain. Alternatively, if it is recognized that there is a part in contact with a part of the skin, it may be determined that the third condition is not satisfied. In this case, the computer may correct the contact portion that does not satisfy the third condition to fit the layer structure of the brain. For example, the computer may calibrate the head image to place the contact portion on the inside of the skull and/or skin.

At this time, in determining whether the third condition is satisfied as an embodiment, the computer first subtracts the original brain internal region before expansion from the region in which the entire brain internal region (cerebrum, cerebellum, ventricle) is expanded (binary dilation). It is possible to obtain a cerebrospinal fluid layer region including a difference region of . At this time, the cerebrospinal fluid layer region completely surrounds the brain inner region. Next, the computer extracts the part that newly surrounds the inner brain region from among the cerebrospinal fluid layer regions including the difference region, and determines that the extracted part is 'the part where the internal brain region is in direct contact with the skull and/or skin'. have. In this case, the computer extracts the 'part in which the internal brain region directly contacts the skull and/or skin' from the head image, so that the layer corresponding to the internal brain region must not come into contact with the layer corresponding to the skull or skin. It can be seen that , is not satisfied.

As another example, in the case of the head image 120 divided into a plurality of regions as shown in FIG. It may be recognized that the partial region 121 of the layer corresponding to the cerebral white matter exists outside the layer corresponding to the cerebral gray matter, and thus it may be determined that the fourth condition is not satisfied. In this case, the computer may correct the partial region 121 that does not satisfy the fourth condition to match the layer structure of the brain. For example, the computer may correct the head image so that the partial region 121 that does not satisfy the fourth condition is placed inside the gray matter of the brain.

At this time, in determining whether or not the fourth condition is satisfied as an embodiment, the computer first obtains a first difference region equal to the amount obtained by subtracting the original white matter region before expansion from the region in which the white matter is dilated (binary dilation), and the gray matter A second difference region may be obtained by subtracting the original gray matter region before dilation from the binary dilation region. Next, the computer may extract a common part (ie, an intersection part) of the first difference region and the second difference region, and determine the extracted common part as 'the part where the gray matter does not cover the white matter'. In this case, the computer extracts the 'part where the gray matter does not cover the white matter' from the head image, so that the layer corresponding to the white matter in the inner brain region should not exist outside the layer corresponding to the gray matter (cortex) in the inner brain region. It can be seen that the fourth condition is not satisfied.

As another example, in the case of the head image 130 divided into a plurality of areas as shown in FIG. It is recognized that there is a portion 131 that is not distributed as a lump with the entire brain region and is distributed at a location outside a certain range, and thus it can be determined that the fifth condition is not satisfied. In this case, the computer may correct the portion 131 that does not satisfy the fifth condition to fit the layer structure of the brain. For example, the computer may perform the connected component-based noise rejection to remove the portion 131 that does not satisfy the fifth condition.

At this time, in determining whether or not the fifth condition is satisfied as an embodiment, the computer performs a binary fill hole operation on the remaining layer regions except for the ventricle, and thereafter, the connection element-based noise Removal can be performed to extract the largest chunk of the entire brain region. In this case, it is possible to determine the regions removed through the connection element-based noise removal among the entire brain regions as 'the entire layer region and the portion not distributed within a certain range'. Here, for the ventricular region, if there is a ventricular region that does not reach the reference size (for example, 30% of the size of the largest mass) based on the size of the region consisting of the largest mass, the 'full layer' for the ventricular region It can be judged as a 'part that is not distributed within the area and a certain range'.

Here, when determining whether the fifth condition is satisfied, since a binary fill hole is performed for the entire brain region except for the ventricle, the inside of each brain region is filled, so the brain layer order must be rearranged accordingly. In this case, the computer first rearranges each brain region according to the order of the layers of the brain (eg, the order of the skin, skull, cerebrospinal fluid, inner brain region), and for each of the rearranged brain regions, the region that still does not satisfy the fifth condition is Whether or not it exists can be judged through connection component-based denoising. Next, the computer may set an NxNxN-sized cube area for the area that still does not satisfy the fifth condition, and check a layer (eg, a label value) to which the neighboring areas of the set cube area belong. The computer may select a layer (eg, a label value) to which the neighboring regions most belong according to the result of the check, and correct the region that does not satisfy the fifth condition to belong to the selected layer region.

As described above, after the computer divides the image of the head of the object into a plurality of brain regions, the computer may determine whether the brain layer arrangement conditions (first to fifth conditions) are satisfied for each of the divided brain regions. .

In this case, the computer may determine whether there is an arrangement relationship that matches the layer structure of the brain in the order of the fifth condition, the fourth condition, the third condition, the second condition, the first condition, and then the fifth condition.

In an embodiment, first, the computer may determine whether a portion of the head image divided into a plurality of regions is distributed at a location outside a certain range rather than clustered and distributed as a lump with the entire brain region. That is, the computer can detect a brain region that is separate from the entire brain region. In the process of determining whether the fifth condition is satisfied, a hole in each layer (eg, regions specified by label values 1 to 8 by dividing the brain into regions based on the brain structure) is filled with binary fill holes ) operation, and extracting only the region consisting of the largest mass of the layer region excluding the ventricle through the connection component label technique (connection component-based noise removal).

Through this process, the head image divided into a plurality of regions may be processed in a state in which a hole is filled and a brain region separated from the entire brain region is removed. Therefore, the computer may proceed with the following process by using the image of the head in the thus-processed state.

Next, the computer determines that the layer corresponding to the cerebral white matter should not exist outside the layer corresponding to the gray matter of the cerebrum in the head image after the hole filling operation and the connection component-based noise removal operation as described above are performed. It may be determined whether the fourth condition is satisfied. Since the process of determining whether the fourth condition is satisfied has been described in detail with reference to FIG. 8 , a detailed description thereof will be omitted herein. As a result of the determination of the fourth condition, the computer may detect whether there is a 'part in which gray matter does not cover white matter' from the head image, and may correct it.

Next, the computer calculates that, in the head image after the hole filling operation and the connection component-based noise removal operation as described above, the layer corresponding to the inner brain region (cerebrum, cerebellum, ventricle) is in contact with the skull or / and skin It may be determined whether the third condition not to be satisfied is satisfied. Since the process of determining whether the third condition is satisfied has been described in detail with reference to FIG. 8 , a detailed description thereof will be omitted herein. As a result of the determination of the third condition, the computer may detect whether there is a 'part in which the brain internal region is in direct contact with the skull and/or skin', and may correct it.

Next, the computer determines whether the second condition that the layer corresponding to the cerebrospinal fluid should not come into contact with the skin in the head image after the hole filling operation and the connection element-based noise removal operation as described above is satisfied. can Since the process of determining whether the second condition is satisfied has been described in detail with reference to FIG. 8 , a detailed description thereof will be omitted herein. As a result of the determination of the second condition, the computer may detect whether a 'part in which the cerebrospinal fluid is in direct contact with the skin' exists, and may correct it.

Next, the computer satisfies the first condition that, in the head image after the hole filling operation and the connection component-based noise removal operation as described above, there should not be another layer disposed outside of the layer corresponding to the skin. You can decide whether to do it or not. Since the process of determining whether the first condition is satisfied has been described in detail with reference to FIG. 8 , a detailed description thereof will be omitted herein. As a result of the determination of the first condition, the computer may detect whether a 'region other than the skin outside the skin' exists, and may correct it.

Next, the computer may rearrange each layer region processed in the order of the above process, that is, the fourth condition, the third condition, the second condition, and the first condition, according to the brain structure. At this time, since the inside of each layer region that has gone through the above process is filled through the hole filling operation, each layer must be rearranged to fit the brain structure. Thus, the computer can rearrange each layer in the following order: skin, skull, cerebrospinal fluid, and regions inside the brain.

After rearranging each layer, the computer can check whether there is still a region separated from the entire brain region for each rearranged layer. In this case, the computer may apply a connection component labeling technique to each rearranged layer. Thereafter, when a region distant from the entire brain region exists, the computer may set a cube region of size NxNxN and extract a region surrounding the set cube region. In this case, the computer may identify a layer (eg, a label value) to which the surrounding areas extracted from the cube area belong, and derive the layer most distributed around the cube. Accordingly, the computer can perform correction by changing a region distant from the entire brain region (the corresponding cube region) to the layer region most distributed around it.

That is, the computer has an arrangement relationship that matches the layer structure of the brain in the order of the fifth condition, the fourth condition, the third condition, the second condition, the first condition, and the fifth condition as described above for the head image. By judging whether or not and correcting a layer that does not conform to the layer structure of the brain, a brain image conforming to the clinical brain structure can be finally derived.

As described with reference to FIGS. 1 to 8 , the computer may acquire a corrected head image using the layer arrangement condition of the brain. Accordingly, it is possible to derive a more accurate head image divided into a plurality of brain regions.

Also, according to an embodiment of the present invention, the computer may generate a 3D brain image of the object by performing 3D brain modeling of the object based on the corrected head image. In this case, it is possible to obtain a more accurate three-dimensional brain model that fits the layer structure of the clinical brain.

Also, according to an embodiment of the present invention, the computer may simulate electrical stimulation to the brain of the subject based on the corrected head image. Alternatively, the computer may simulate electrical stimulation of the subject's brain using a three-dimensional brain image generated based on the corrected head image.

As in an embodiment of the present invention, instead of using a head image in which each brain region is corrected to match the layer structure of the brain, a general head image (ie, a head image divided into wrong areas) is used for brain modeling or brain When simulating the transmission process of electrical stimulation to the brain, voltage and electric field values may be too high or low if electrical stimulation is applied to an incorrectly divided brain part. That is, when a specific electrical stimulus is applied to a point on the head of the subject and a path through which the specific electric stimulus is propagated in the brain of the subject is simulated, an accurate simulation result cannot be derived.

However, in an embodiment of the present invention, since the head image in which each brain region is corrected to match the layer structure of the brain is used, a specific point on the head image of the object to which the electrical stimulation is applied can be more accurately designated. Accordingly, the accuracy of the simulation result can also be improved.

9 is a diagram schematically showing the configuration of an apparatus 200 for performing a brain image correction method using a brain structure according to an embodiment of the present invention.

Referring to FIG. 9 , the processor 210 includes one or more cores (not shown) and a graphic processing unit (not shown) and/or a connection path (eg, a bus) for transmitting and receiving signals with other components. ) may be included.

The processor 210 according to an embodiment executes one or more instructions stored in the memory 220 to perform a brain image correction method using the brain structure described with reference to FIGS. 1 to 8 .

In one example, the processor 210 executes one or more instructions stored in the memory 220 to obtain a head image including the subject's brain, dividing the head image into a plurality of regions based on the brain structure; and performing correction using the layer arrangement condition of the brain for the plurality of regions.

On the other hand, the processor 210 is a RAM (Random Access Memory, not shown) and ROM (Read-Only Memory: ROM) for temporarily and / or permanently storing signals (or data) processed inside the processor 210 . , not shown) may be further included. Also, the processor 210 may be implemented in the form of a system on chip (SoC) including at least one of a graphic processing unit, a RAM, and a ROM.

The memory 220 may store programs (one or more instructions) for processing and controlling the processor 210 . Programs stored in the memory 220 may be divided into a plurality of modules according to functions.

The brain image correction method using the brain structure according to an embodiment of the present invention described above may be implemented as a program (or application) and stored in a medium to be executed in combination with a computer, which is hardware.

The above-described program is C, C++, JAVA, machine language, etc. that a processor (CPU) of the computer can read through a device interface of the computer in order for the computer to read the program and execute the methods implemented as a program It may include code (Code) coded in the computer language of Such code may include functional code related to functions defining functions necessary for executing the methods, etc. can do. In addition, the code may further include additional information necessary for the processor of the computer to execute the functions or code related to memory reference for which location (address address) in the internal or external memory of the computer to be referenced. have. In addition, when the processor of the computer needs to communicate with any other computer or server located remotely in order to execute the above functions, the code uses the communication module of the computer to determine how to communicate with any other computer or server remotely. It may further include a communication-related code for whether to communicate and what information or media to transmit and receive during communication.

The storage medium is not a medium that stores data for a short moment, such as a register, a cache, a memory, etc., but a medium that stores data semi-permanently and can be read by a device. Specifically, examples of the storage medium include, but are not limited to, ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage device. That is, the program may be stored in various recording media on various servers accessible by the computer or in various recording media on the computer of the user. In addition, the medium may be distributed in a computer system connected by a network, and a computer readable code may be stored in a distributed manner.

The steps of a method or algorithm described in relation to an embodiment of the present invention may be implemented directly in hardware, as a software module executed by hardware, or by a combination thereof. A software module may contain random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can realize that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

Claims (19)

A method for segmenting a medical image, comprising:
determining one or more first segmentation regions that are a set of pixels or voxels having a labeling value indicating a first region;
determining one or more second segmentation regions that are a set of pixels or voxels having a labeling value indicating a second region;
determining the arrangement relationship between the first segmentation area and the second segmentation area; wherein the determining of the arrangement relationship comprises expanding the first segmentation area and forming the expanded first segmentation area and the first segmentation area. acquiring a first difference region that is a difference region to obtain a second difference region that is a difference region between the first difference region and the second segmentation region;
When the second difference region exists, changing the labeling value of the second difference region to a labeling value indicating the second region;
How to segment medical images.
The method of claim 1,
The medical image relates to the skull and brain,
The first segmentation region and the second segmentation region is a region included in at least one of the skull and the brain
How to segment medical images.
3. The method of claim 2,
The second segmentation region is a region related to at least any one of internal regions of the brain including skin, skull, cerebrospinal fluid and cerebrum, cerebellum, and ventricle,
Wherein the first segmentation region relates to a region located inside the region related to the first segmentation region
How to segment medical images.
The method of claim 1,
The medical image is an MRI image
How to segment medical images.
The method of claim 1,
The first segmentation area and the second segmentation area are adjacent to each other.
How to segment medical images.
The method of claim 1,
changing the labeling values of the remaining regions except for the region including the most pixels or voxels among the first segmentation regions;
How to segment medical images.
The method of claim 1,
changing the labeling values of the remaining regions except for the region in which pixels or voxels are included the most among the second segmentation regions;
How to segment medical images.
The method of claim 1,
The difference region is defined as a region excluding a common region from any one segmentation region in a relationship between two or more different segmentation regions.
How to segment medical images.
The method of claim 1,
determining one or more third segmentation regions that are a set of pixels or voxels having a labeling value indicating a third region;
determining a positional arrangement relationship between the second and third segmentation regions; wherein the determining of the arrangement relationship includes expanding a second segmentation region, and forming the second segmentation region and the third segmentation region acquiring a third difference region that is a difference region to obtain a fourth difference region that is a difference region between the third difference region and the third segmentation region;
When the fourth difference region exists, changing the labeling value of the fourth difference region to a labeling value indicating a third region;
How to segment medical images.

A device for segmenting a medical image, comprising:
a medical image acquisition module configured to acquire a medical image of the object; and
A processor for segmenting the obtained medical image; including,
The processor determines one or more first segmentation regions that are a set of pixels or voxels having a labeling value indicating a first region in the obtained medical image, and one that is a set of pixels or voxels having a labeling value indicating a second region determining the above second segmentation area,
determining the arrangement relationship between the first segmentation area and the second segmentation area, extending the first segmentation area, and obtaining a first difference area that is a difference area between the expanded first segmentation area and the first segmentation area , obtain a second difference region that is a difference region between the first difference region and the second segmentation region,
When the second difference region exists, the labeling value of the second difference region is changed to a labeling value indicating the second region.
A device for segmenting medical images.
11. The method of claim 10,
The medical image relates to the skull and brain,
The first segmentation region and the second segmentation region is a region included in at least one of the skull and the brain
A device for segmenting medical images.
12. The method of claim 11,
The second segmentation region is a region related to at least any one of internal regions of the brain including skin, skull, cerebrospinal fluid and cerebrum, cerebellum, and ventricle,
Wherein the first segmentation region relates to a region located inside the region related to the first segmentation region
A device for segmenting medical images.
11. The method of claim 10,
The medical image is an MRI image
A device for segmenting medical images.
11. The method of claim 10,
The processor determines a third segmentation region representing the third region,
The second segmentation area is located between the first and third segmentation areas.
A device for segmenting medical images.
11. The method of claim 10,
The processor is configured to change the labeling value of an area other than an area including the most pixels or voxels among the first segmentation areas.
A device for segmenting medical images.
11. The method of claim 10,
and the processor changes the labeling value of the remaining regions except for the region including the most pixels or voxels among the second segmentation regions.
A device for segmenting medical images.
11. The method of claim 10,
The difference region is defined as a region excluding a common region from any one segmentation region in a relationship between two or more different segmentation regions.
Device for segmenting medical images
11. The method of claim 10,
The processor further determines one or more third segmentation regions that are a set of pixels or voxels having a labeling value indicating the third region,
In the determining of the disposition relationship, a second segmentation area is expanded, a third difference area that is a difference area between the extended second segmentation area and the third segmentation area is obtained, and the third difference area and the third determining a positional arrangement relationship between the second and third segmentation regions by obtaining a fourth difference region that is a difference region of the segmentation region;
When the fourth difference region exists, the labeling value of the fourth difference region is changed to a labeling value indicating the third region.
A device for segmenting medical images.
A non-transitory electronic recording medium in which a computer program for driving the method of any one of claims 1 to 9 is recorded.

Images