KR101178398B1 - Method and apparatus for processing segmentation of imaging in medicine, and recording medium having computer program therefor - Google Patents

Abstract

The present invention relates to a medical image segment processing apparatus, a method and a recording medium on which a computer program is recorded. The disclosed medical image segment processing apparatus includes a medical image input unit configured to receive and load a medical image, and a 3D image segment processing of the medical image. An interface unit providing an interface for setting parameters for the image, an image divider for performing 3D image segmentation processing on the ROI included in the medical image according to the parameter set through the interface, and the image divider It includes a segmentation visualization unit that visualizes and outputs the dimensional image segmented processed region, and has the advantage of improving the segmentation speed without reducing the accuracy of segmentation in the 3D space by applying semi-automatic segmentation processing for medical images. have.

Description

TECHNICAL AND APPARATUS FOR PROCESSING SEGMENTATION OF IMAGING IN MEDICINE, AND RECORDING MEDIUM HAVING COMPUTER PROGRAM THEREFOR}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to medical image segmentation processing, and more particularly, to segmenting a region of interest from an entire image based on image data obtained by a medical image acquisition device such as CT (Computed Tomography). An apparatus and method for medical image segmentation processing and a recording medium having a computer program capable of performing the medical image segmentation processing method are recorded.

Medical image segmentation technology refers to a process of separating a region of interest from an entire image based on image data acquired by a medical image acquisition device such as computed tomography (CT). As the process proceeds, the target region of interest is divided from the entire image, and then the region of interest is represented as a 2D or 3D image to improve the accuracy or efficiency of the diagnosis or research results.

However, since various organs included in the medical image have a complicated geometric shape, dividing the region of interest according to the purpose requires a high degree of medical image segmentation processing technology.

For example, one of the diseases of the otorhinolaryngology, empyema, refers to a disease caused by inflammation of the mucosa in the paranasal sinus located in the nasal and middle nasal passages of the nasal cavity. This often occurs in the maxillary sinus. The nasal airway, which consists of the nasal and paranasal sinuses, is a nasal airway. The nasal cavity is the first gateway to the respiratory tract in the center of the face, and the part surrounding the nasal cavity forms an empty space that contains air and communicates with the nasal cavity, which is called the sinus.

These nasal breathing vessels are very important in establishing treatment plans and surgical plans for otorhinolaryngology diseases. Morphological knowledge of nasal throats has important clinical value. It may be obtained from the cross-sectional images obtained through a medical image acquisition device such as.

In addition, during the medical image segmentation processing for the nasal breathing tube, the nasal breathing tube is separated through appropriate division in each cross-sectional image.

On the other hand, the automation of the segmentation process is difficult to apply due to the characteristics of the medical image, many researches on the medical image segmentation process is progressing to compensate for this.

However, due to the accuracy of segmentation, clinical practice still uses the manual segmentation method. As described above, the nasal cavity and the paranasal sinuses have complex geometrical shapes in cross-sectional images. to be.

Therefore, it took a long time for the segmentation processing of the medical image, there was a problem that the accuracy of segmentation greatly depends on the skill of the operator performing the segmentation process of the medical image.

The present invention has been proposed to solve such a problem of the prior art, and the procedure requiring high medical knowledge during the segmentation processing of the medical image maintains the manual method, and for the procedure capable of automated processing, The present invention provides a semiautomatic medical image segmentation processing method capable of improving segmentation speed without degrading segmentation accuracy in three-dimensional space by applying segmentation processing.

The present invention also provides a recording medium having a computer program recorded thereon capable of performing a method of semi-automatic medical image segmentation processing using a combination of manual and automatic segmentation.

According to a first aspect of the present invention, a medical image segmentation processing apparatus includes: a medical image input unit configured to receive and load a medical image; an interface unit configured to provide an interface for setting parameters for 3D image segmentation processing of the medical image; An image divider which performs 3D image segmentation processing on the ROI included in the medical image according to the parameter set through the interface, and visualizes and outputs a segmented region which has been processed by the image divider It may include a partition visualization unit.

The image splitter may include a region of interest splitter configured to perform the 3D image segmentation processing on the region of interest using a multiple selection point region extension technique.

The image splitter may include a body region separator that separates a body region from the medical image before performing the 3D image segmentation processing.

The image segmentation unit may perform the 3D image segmentation processing using at least one of a boundary-based technique, a threshold holding technique, and a branch shape analysis company.

According to a second aspect of the present invention, there is provided a medical image segmentation processing method comprising: receiving and loading a medical image, providing an interface for setting parameters for 3D image segmentation processing of the medical image, and And performing a 3D image segmentation process on the ROI included in the medical image according to the parameter set through the parameter, and visualizing and outputting the segmented region by the 3D image segmentation process.

In the performing of the step, the 3D image segmentation may be performed on the ROI using a multi-selection point expansion technique.

The performing may include separating a body region from the medical image before performing the 3D image segmentation processing.

The performing of the 3D image segmentation may be performed using at least one of a boundary-based technique, a threshold holding technique, and a branch form analysis company.

As a third aspect of the present invention, there is provided a recording medium on which a computer program capable of performing any one of the above-described medical image segmentation processing methods is recorded.

According to an embodiment of the present invention, a procedure requiring high medical knowledge during a segmentation process for a medical image may be maintained in a manual manner, and an automatic segmentation process for a medical image may be applied to a procedure capable of automatic processing. It is possible to improve the splitting speed without degrading the accuracy of splitting in the phase. For example, when the body region separation and the 3D multi-point selection region expansion technique that excludes the background from the medical image before the 3D image segmentation processing are applied, there is an effect of providing a more accurate image segmentation without noise.

1 is a block diagram of a medical image processing apparatus according to an exemplary embodiment of the present invention.
2 is an example of an interface screen provided by a medical image processing apparatus according to an embodiment of the present invention.
3 is a flowchart illustrating a medical image segmentation processing method according to an embodiment of the present invention.
4 is a diagram for comparing before and after separating a body region from a medical image according to an exemplary embodiment of the present invention.
5 is a diagram illustrating an example of marking a region of interest according to an exemplary embodiment of the present invention.
6 is a conceptual diagram of a 3D region extension technique using a plurality of selection points according to an embodiment of the present invention.
FIG. 7 is a diagram for comparing a result by various image segmentation techniques according to an exemplary embodiment of the present invention. FIG.
8 is a diagram illustrating a result of visualization by dividing the nasal respiratory tract among medical images according to an embodiment of the present invention.
9 is an example in which an image segmentation result according to an exemplary embodiment of the present invention is overlapped with an image segmentation result according to the related art.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

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. The following terms are defined in consideration of the functions in the embodiments of the present invention, which may vary depending on the intention of the user, the intention or the custom of the operator. Therefore, the definition should be based on the contents throughout this specification.

Each block of the accompanying block diagrams and combinations of steps of the flowchart may be performed by computer program instructions. These computer program instructions may be loaded into a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus so that the instructions, which may be executed by a processor of a computer or other programmable data processing apparatus, And means for performing the functions described in each step are created. These computer program instructions may be stored in a computer usable or computer readable memory that can be directed to a computer or other programmable data processing equipment to implement functionality in a particular manner, and thus the computer usable or computer readable memory. It is also possible for the instructions stored in to produce an article of manufacture containing instruction means for performing the functions described in each block or flowchart of each step of the block diagram. Computer program instructions may also be mounted on a computer or other programmable data processing equipment, such that a series of operating steps may be performed on the computer or other programmable data processing equipment to create a computer-implemented process to create a computer or other programmable data. Instructions that perform processing equipment may also provide steps for performing the functions described in each block of the block diagram and in each step of the flowchart.

Also, each block or each step may represent a module, segment, or portion of code that includes one or more executable instructions for executing the specified logical function (s). It should also be noted that in some alternative embodiments, the functions noted in the blocks or steps may occur out of order. For example, two blocks or steps shown in succession may in fact be performed substantially concurrently, or the blocks or steps may sometimes be performed in reverse order according to the corresponding function.

1 is a configuration diagram of a medical image segmentation processing apparatus capable of performing a medical image segmentation processing method according to an exemplary embodiment.

As described above, the medical image segment processing apparatus 100 may include a medical image input unit 110, an image divider 120, an interface unit 130, a segmented image visualization unit 140, and the like. 120 includes a body region separator 121, an ROI divider 123, and the like.

The medical image input unit 110 receives and loads a medical image such as a CT digital image communication in medicine (DICOM) dataset.

The interface unit 130 provides an interface for setting parameters for 3D image segmentation processing of the medical image.

The image splitter 120 performs 3D image segmentation processing on the ROI included in the medical image according to a parameter set through the interface. The image splitter 120 splits the 3D image by using the body region separator 121 that separates the body region from the medical image and a multiple selection point region extension technique for the ROI before performing the 3D image segmentation processing. The ROI divider 123 may perform a processing. For example, the ROI divider 123 may perform 3D image segmentation using any one of a boundary-based technique, a threshold holding technique, and a branch form analysis company in addition to the multiple selection point region expansion technique.

The divided region visualization unit 140 visualizes and outputs the divided region processed by the image divider 120 in three-dimensional image division.

2 is an exemplary view of an interface screen provided by a medical image segment processing apparatus according to an exemplary embodiment.

Reference numeral 1 is a menu for selecting a style of a line to be displayed as the region of interest and selecting a line of the region of interest, and reference numeral 2 is a menu for selecting a plurality of specific picking points. 3 is a menu that sets the range value that determines how much to extend the difference from a specific point value selected for the area extension application, and 4 is a menu for controlling the threshold for applying body area separation. Reference numeral 5 denotes a menu for designating a range of axial slices necessary for medical images, and reference numeral 6 denotes a menu for selecting image segmentation after completing parameter setting through a menu of 1 to 5. .

Here, the selection may be deselected so as not to apply the body region separation through the menu of reference numeral 4. For example, an initial default value of the threshold may be applied as an average value of each cross section.

Hereinafter, the medical image segmentation processing by the medical image segmentation processing apparatus according to an embodiment of the present invention will be described in time series with reference to FIGS. 1 to 9. In the following description, a process of dividing the nasal respiratory tract among medical images for illustrative purposes will be described.

First, when the medical image is input and loaded through the medical image input unit 110 (S301), a procedure of setting various parameters for image segmentation processing by the image splitter 120 through the interface unit 130 must be performed. . The embodiment of FIG. 3 illustrates a parameter setting process for 3D region extended image segmentation processing. As described above with reference to FIG. 2, the region of interest is selected after selecting a region of interest line style using various menus of an interface screen. A plurality of feature points may be selected, an image segmentation comparison allowance value may be set, a body region separation may be selected, a threshold value may be set, and an image processing range may be set (S303).

Computed tomography data, which is a medical image input and loaded through the medical image input unit 110, is an image representing information of air, soft-tissue, and bone. In addition to the data in the image, the air present in the computed tomography apparatus is included, and in some cases, the data serves as a noise, thereby reducing the accuracy of the segmentation result. In the medical image processing according to the prior art, the noise outside the body that may be included in the acquisition of the image does not give meaning to the occurrence, but the body region separator 121 of the image divider 120 may provide more accurate segmentation results. In the medical image, only the body region is separated by removing the region other than the body, that is, the background (S305).

Referring to the above-described body region separation process, a binary image s (x, y) is first generated as an average value of each image. This may be expressed as Equation 1 below.

[Equation 1]

And, if not, the current position s (x _i, y _j) of the above s (x _i, y _j-1) and left-s (x _i-1, y _j) value is present and the ( "0" of the reference ), If a value exists at the current coordinate s (x _i , y _j ) (if not "0"), sets the value of s (x _i , y _j ) to "1", and the current coordinate s (x _i , y _j ) If no value exists ("0"), set the value to "0.5". This may be expressed as Equation 2 below. Here, setting to "1" means marking the body region found correctly, and setting to "0.5" means marking the background region that has not been separated.

&Quot; (2) "

Then, similarly, the search starts from the end of the image so that the following g (x _i , y _{j +1} ) and right g (x _{i +1} , y _j ) values do not exist ("0"), and the current coordinate g If the value of (x _i , y _j ) is "0.5", the value of "0" is input to R (x _i , y _j ) of the image output image to remove the protruding edge portion outside the image shown in FIG. Only body parts can be split. This may be expressed as Equation 3 below.

&Quot; (3) "

4 is a diagram for comparing before and after separating the body region in the medical image according to an embodiment of the present invention, (a) is before separating the body region, (b) is a binary image as an average value of each image And (c) after separating the body region. (d) shows the result of applying the body region separation process according to the embodiment of the present invention to another medical image.

Next, a procedure of marking a region of interest using the interface unit 130 is performed on the medical image in which the body region is separated. For example, the interface screen illustrated in FIG. 2 may be used. 5 shows an example of marking a region of interest according to an embodiment of the present invention, (a) is an example of marking for splitting the right and left nasal breathing tubes together, and (b) is for dividing the right nasal breathing tubes. This is an example of marking.

In this way, a boundary to apply 3D image segmentation is set by marking a region of interest, and by marking a region of interest, the left and right cavities are divided or extracted at once according to the purpose of use of the volume to be separated, or the sinus region is extracted. You can divide by including.

Thereafter, the ROI divider 123 of the image divider 120 performs 3D image segmentation on the marked ROI of the medical image (S307).

Here, the ROI divider 123 may use a segmentation method based on discontinuity or a segmentation method based on similarity in order to segment an ROI of the medical image. The image segmentation algorithm is an algorithm for dividing an area or an object. In general, important parts of an image are distinguished by contrast. Therefore, most segmentation techniques are based on one of two basic properties of contrast: discontinuity or similarity. The discontinuity-based method is a technique for segmenting an image based on a sharp change in contrast, such as an outline of an image, and is used for separation of boundary features of segmented regions such as points, lines, and edges. Segmentation consists of processing these boundary features to separate the desired area. The important methods based on similarity are to collect points with similar contrast and divide them into similar areas, such as thresholding, region growing, region splitting and merging. Techniques, branch point morphological watershed techniques, and so on.

Among these image segmentation techniques, the region extension method provides good results for separating regions with similar values. For example, the nasal respiratory tract is a route of connected respiration with similar values, so using this feature to obtain an accurate segmentation result can be achieved by using the region expansion method. For example, a method of extending a seeded region growing technique, which is one of region expansion methods, is used to store a voxel value of a multi-seed point instead of a single point in a three-dimensional space. Multi-Seed Region Growing (MSRG), which extends to areas with similar values, can provide more accurate partitioning results. This technique can be separated without loss of thin plates, such as the bony septa of ethmoidal sinus, and remembers multi-seed picking coordinates for each image, thus making it possible to select points in multiple images. Picks are possible.

This three-dimensional extended image segmentation process compares the pixel value at one point with the neighboring pixel value, and if the identity measure (ie having similar pixel values) is satisfied, the pixel belongs to the same classification with the neighboring pixels. . To this end, the selection point (Seed Point), which is a starting position for measuring the identity, must be selected first. In detail, the coordinate values (Seed _{1 to n} ) of the plurality of selected points are _first stored, and the maximum value Max.rv of the area to be extended is set. When the current value is g (A) = seed _i (where A is the current voxel), the 8-neighborhood comparison in the x, y, and z directions at each seed _i Calculate the mean of (mean (B)), where B is the progressed voxels. Then, the present value is compared with the average value, and the difference (| g (A) -mean (B) |) is compared with the maximum value Max.rv to check whether it is included in the area. Here, if the difference between the current value and the average value is less than the maximum value (Max.rv) and the coordinates that have not been visited yet, it is included in the nasal breathing area, otherwise it is not included in the nasal breathing area. Repeat these processes in three-dimensional space to expand the area.

6 is a conceptual diagram of a 3D region extension technique using a plurality of selection points according to an embodiment of the present invention. As shown in FIG. 6, when a maximum boundary is specified by marking a region of interest to be extended and a plurality of selection points are specified to pick an inside of a region to be separated, a multi-seed value is shown in FIG. 6. By using it, it expands in the x, y, and z directions, allowing accurate image separation.

Meanwhile, the image splitter 120 may segment the medical image using another image segmentation technique in addition to the region extension technique described above. For example, a boundary-based technique using discontinuity, a threshold holding technique using similarity, and a morphological watersheds technique may be used.

The boundary-based technique takes advantage of the property that the pixel values change rapidly at the boundary of two regions. High-contrast values in this filter process become edge candidates that form a closed curve that marks the boundaries between the regions.

The threshold holding technique is a simple method of classifying regions based on a threshold value, and a method of classifying a region based on a threshold value in a global or local region may be used in a medical image.

The branch point shape analysis technique expresses the concept of watershed based on the shape of three-dimensional space. In general, such a split point division method is applied to a gradient of an image so that a boundary between regions is shown as a topographic map in the form of contour lines. The division is made in the same shape as the topographical map as water flows in the form of region boundaries from the starting point to the other starting point.

However, unlike the above-described area extension method, when the threhold holding method is applied to the nasal breathing tube segmentation, it may include noise as shown in FIG. 7 (a), and when a small threshold is used to reduce the noise, FIG. Too small a region, such as (b), can cause problems in accuracy. In addition, even when the point of interest technique and the threshold holding technique are used together to improve the problems of the two cases, the accurate separation result may not be obtained as shown in FIG. Fig. 7 (d) shows the result of the three-dimensional area extension division processing using the point of interest.

Next, the divided image visualization unit 140 visualizes and outputs the medical image divided by the image division unit 120 (S309). For example, the segmented image visualization unit 140 may apply a shear-warp rendering technique among volume rendering techniques. Figure 8 illustrates the result of visualization by dividing the nasal breathing tube according to the practice of the present invention. 9 illustrates an example in which an image segmentation result according to an exemplary embodiment of the present invention and an image segmentation result according to the prior art are overlapped and compared.

The medical image segmentation processing apparatus and method according to an embodiment of the present invention have been clinically applied. According to an expert's manual work, when a computer tomography image dataset composed of 181 slides of 512 * 512 resolution is image-segmented, it is 5 hours. In the case of applying the embodiment of the present invention, which took a long time, the image segmentation process could be completed within a maximum of 3 minutes.

The medical image segmentation processing method according to the embodiment of the present invention described above may be created by a computer program. Codes and code segments constituting this computer program can be easily inferred by a computer programmer in the art. In addition, the computer program is stored in a computer readable medium, and read and executed by a computer or a medical image processing apparatus according to an embodiment of the present invention to implement a medical image segment processing method. Can be. The information storage medium includes a magnetic recording medium, an optical recording medium and a carrier wave medium.

100: medical image input unit 120: image segmentation unit
121: body region separator 123: region of interest divider
130: interface unit 140: segmented image visualization unit

Claims (9)

delete A medical image input unit configured to receive and load a medical image;
An interface unit for providing an interface for setting parameters for 3D image segmentation processing of the medical image;
An image divider which performs 3D image segmentation processing on the ROI included in the medical image according to the parameter set through the interface;
A division region visualization unit for visualizing and outputting a division region processed by the image division unit by 3D image division;
The image splitter includes a region of interest splitter configured to perform the 3D image segmentation processing on the region of interest using a multiple selection point region extension technique.
Medical image segmentation processing device. The method of claim 2,
The image divider may include a body region separator that separates a body region from the medical image before performing the 3D image segmentation process.
Medical image segmentation processing device. The method of claim 2,
The image segmentation unit may perform the 3D image segmentation processing using at least one of a boundary-based technique, a threshold holding technique, and a branched shape analysis technique.
Medical image segmentation processing device. delete Receiving and loading a medical image;
Providing an interface for setting parameters for 3D image segmentation processing of the medical image;
Performing 3D image segmentation processing on the ROI included in the medical image according to the parameter set through the interface;
And visualizing and outputting the divided region by the 3D image segmentation processing.
The performing may include performing the 3D image segmentation processing on the ROI using a multiple selection point region extension technique.
Medical image segmentation processing method. The method according to claim 6,
The performing may include separating a body region from the medical image before performing the 3D image segmentation processing.
Medical image segmentation processing method. The method according to claim 6,
The performing may include performing the 3D image segmentation using at least one of a boundary-based technique, a threshold holding technique, and a branch shape analysis company.
Medical image segmentation processing method. A recording medium having recorded thereon a computer program for performing the medical image segmentation processing method of any one of claims 6 to 8.

Images