KR102690206B1 - Method and device to analyze medical image for detecting calcified part in vessel - Google Patents

Abstract

A medical image analysis device according to an embodiment includes an image acquisition unit that acquires a medical image having consecutive frame images; A vessel part is determined from contrast medium frame images in which a contrast medium is observed among the consecutive frame images, and a vessel part is located around the vessel part from non-contrast frame images in which a contrast medium is not observed among the continuous frame images. Determining a located at least one calcium candidate region, wherein the at least one calcium candidate region is determined based on a comparison between a first direction of movement of the at least one calcium candidate region due to a heartbeat and a second direction of movement of the vascular site. It may include a processor that determines whether a calcification part is included.

Description

Medical image analysis device and method for detecting calcified areas in blood vessels {METHOD AND DEVICE TO ANALYZE MEDICAL IMAGE FOR DETECTING CALCIFIED PART IN VESSEL}

Hereinafter, a technology for detecting calcified lesions located in blood vessels subject to analysis in medical images within angiography images is provided.

Interventional procedures that involve inserting stents using catheters to treat cardiovascular, cerebrovascular, and peripheral blood vessels are widely available. Before proceeding with the procedure, the severity of the patient's lesion is assessed using cardiovascular angiography. When confirmed through angiography, the treatment policy may vary depending on the characteristics of the atherosclerotic plaque. In particular, if there is a calcified lesion, myocardial infarction may occur as the rupture of the lesion causes calcified fragments to block the end of the branch. Also, referring to the SYNTAX score, in cases where there are multiple lesions and many calcified lesions in the coronary artery, surgical methods may be more advantageous for prognosis. In this regard, there is a need for a technology to detect calcified lesions based on X-ray angiography images instead of computer tomography (CT).

The background technology described above is possessed or acquired by the inventor in the process of deriving the disclosure of the present application, and cannot necessarily be said to be known technology disclosed to the general public before this application.

A medical image analysis device according to an embodiment includes an image acquisition unit that acquires a medical image having consecutive frame images; and determining a vessel part from the contrast medium frame images in which the contrast medium is observed among the consecutive frame images, and the periphery of the vessel part from the non-contrast frame images in which the contrast medium is not observed among the continuous frame images. Determine at least one calcium candidate region located in the at least one calcium candidate region based on a comparison between a first direction of movement of the at least one calcium candidate region due to a heartbeat and a second direction of movement of the vascular site. It may include a processor that determines whether it contains a calcification part.

The processor detects a blood vessel area from frame images of the medical image, determines a lesion portion among the detected blood vessel areas, and selects the lesion within a region of interest (ROI) including the lesion. A blood vessel site and the at least one calcium candidate region can be determined.

The processor extracts a visual feature including at least one of a borderline, a center line, and a blood vessel segment into which the blood vessel region is divided, and determines the visual feature and a potentially calcified region. The target calcium site can be determined based on the quantitative characteristics of the part).

The processor is configured to, based on a correlation score between a first motion vector calculated for the at least one calcium candidate region and a second motion vector calculated for the blood vessel region, exceed a threshold, the at least one It can be determined that the calcium candidate region includes the potential calcified site.

The processor may determine whether the potential calcified site is a target calcium site based on at least one of the length and width of the potential calcified site in the quantitative characteristic.

The processor may determine whether the potential calcified site is a target calcium site based on the extension direction of the potential calcified site and the central axis direction of the blood vessel site in the visual feature.

The processor is configured to determine the location of the at least one calcium candidate region in a second non-contrast frame after the first non-contrast frame from the position of the at least one calcium candidate region in a first non-contrast frame among the non-contrast frame images. A motion vector representing the motion of the robot can be calculated as the first movement direction.

The processor moves the second motion vector representing motion from the location of the blood vessel portion in a first contrast medium frame among the contrast medium frame images to the position of the blood vessel portion in a second contrast medium frame after the first contrast medium frame. It can be calculated as a direction.

The processor may exclude a calcium candidate region detected at a position exceeding a critical distance from the blood vessel region among the at least one calcium candidate region from calcium site determination.

The processor may determine the first movement direction and the second movement direction due to the heartbeat based on the period of the patient's ECG signal.

When the ratio of the movement distance of the at least one calcium candidate region to the movement distance of the blood vessel site is less than or equal to a threshold ratio, the processor may determine that the at least one calcium candidate region does not include the target calcium site.

The medical image may be an X-ray-based Cardiac Angiography (CAG) image.

A medical image analysis method implemented with a processor according to an embodiment includes acquiring a medical image having consecutive frame images; determining a vessel part from contrast medium frame images in which the contrast medium is observed among the consecutive frame images; determining at least one calcium candidate area located around the blood vessel area from non-contrast frame images in which no contrast medium is observed among the consecutive frame images; and whether the at least one calcium candidate region includes a calcification part based on a comparison between the first movement direction of the at least one calcium candidate area and the second movement direction of the blood vessel portion by heartbeat. It may include a step of determining.

1 shows a medical image analysis device according to an embodiment.
Figures 2 and 3 are flowcharts showing a medical image analysis method according to an embodiment.
Figure 4 explains the operation of detecting a lesion (lesion portion) and extracting visual features of a blood vessel portion according to an embodiment.
Figure 5 illustrates detection of a calcium candidate region according to one embodiment.
Figures 6 and 7 illustrate the first movement direction of the calcium candidate region and the second movement direction of the blood vessel site according to one embodiment.
Figure 8 is a diagram explaining the determination of a calcified part according to an embodiment.

Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be changed and implemented in various forms. Accordingly, the actual implementation form is not limited to the specific disclosed embodiments, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical idea described in the embodiments.

Terms such as first or second may be used to describe various components, but these terms should be interpreted only for the purpose of distinguishing one component from another component. For example, a first component may be named a second component, and similarly, the second component may also be named a first component.

When a component is referred to as being “connected” to another component, it should be understood that it may be directly connected or connected to the other component, but that other components may exist in between.

Singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, terms such as "comprise" or "have" are intended to designate the presence of the described features, numbers, steps, operations, components, parts, or combinations thereof, but are not intended to indicate the presence of one or more other features or numbers, It should be understood that this does not exclude in advance the possibility of the presence or addition of steps, operations, components, parts, or combinations thereof.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art. Terms as defined in commonly used dictionaries should be interpreted as having meanings consistent with the meanings they have in the context of the related technology, and unless clearly defined in this specification, should not be interpreted in an idealized or overly formal sense. No.

Hereinafter, embodiments will be described in detail with reference to the attached drawings. In the description with reference to the accompanying drawings, identical components will be assigned the same reference numerals regardless of the reference numerals, and overlapping descriptions thereof will be omitted.

1 shows a medical image analysis device according to an embodiment.

The medical image analysis device 100 according to one embodiment is a device that analyzes medical images. The medical image analysis device 100 may include an image acquisition unit, a processor, and memory.

The image acquisition unit 130 may acquire a medical image 131. The medical image 131 in which the blood vessels of the target object (eg, the recipient 190) are imaged may also be referred to as a blood vessel image. For example, the image acquisition unit 130 includes X-ray imaging equipment, and produces a medical image 131 (e.g., X-ray-based CAG image) through ) can be taken.

The medical image 131 may include a plurality of consecutive frames. To image blood vessels, a contrast agent may be injected into the blood vessels of the patient 190, and the blood vessels of the patient 190 may be imaged while the contrast agent is maintained. If the injected contrast medium is absorbed into the body outside the blood vessels and disappears, the blood vessels are not imaged, and parts with a high X-ray absorption rate, including the bones and calcified parts of the subject 190, may be imaged. Among consecutive frames of the medical image 131, a frame in which the contrast agent is not observed may be referred to as a non-contrast frame, and a frame in which the contrast agent is observed may be referred to as a contrast agent frame. Consecutive frames are a series of non-contrast frames before the contrast medium is injected (e.g., a non-contrast frame set), a series of contrast frames after the contrast medium is injected (e.g., a contrast frame set), and a series of frames after the contrast medium is absorbed and disappears. May include non-contrast frames. When contrast agent is repeatedly injected, non-contrast agent frame sets and contrast agent frame sets may appear repeatedly. An image in a non-contrast frame can be referred to as a non-contrast frame image, and an image in a contrast frame can be referred to as a contrast frame image. In a CAG image based on X-rays, the intensity value of each pixel is the intensity due to the transmission of

For example, in a non-contrast frame image, pixels corresponding to parts with high X-ray absorption may have low intensity values. In non-contrast frame images, bones and calcified lesions mainly appear dark, and blood vessels may appear bright. In non-contrast framed images, blood vessels may have a transmittance similar to that of other organs in the human body. In non-contrast framed images, blood vessels may not be distinguishable from other organs. In a contrast agent frame image, pixels corresponding to blood vessels where the contrast agent was injected and points where calcium is located may have low intensity values. In contrast frame images, blood vessels may have a transmittance similar to that of calcium. In contrast frame images, blood vessels can be distinguished from other organs. However, when a calcified lesion exists within a blood vessel, the point corresponding to the calcified area and the point corresponding to the blood vessel are adjacent to or overlap each other, so it may be difficult to distinguish between the blood vessel and calcium in the contrast frame image.

For reference, an example in which the image acquisition unit 130 captures the medical image 131 with an It includes a communication module for, and may receive a blood vessel image (e.g., CAG image based on X-rays) from an external imaging device through the communication module.

The processor 110 may detect calcified lesions based on the contrast medium frame image and the non-contrast frame image. For example, calcified areas caused by calcified lesions that cause vascular stenosis occur on the inner wall of blood vessels, so the calcified areas can move with the blood vessels. The movement of blood vessels caused by the heartbeat and the movement of calcified areas occurring in the blood vessels may be the same or similar. Therefore, the processor 110 determines whether the calcium candidate region includes a calcified region based on the movement of the calcium candidate region detected in the non-contrast frame image and the movement of the blood vessel region detected in the contrast agent frame image. It can be determined. The operation of the processor 110 is explained in FIGS. 2 to 8 below.

The output (not shown) allows visualization of confirmed calcified sites. For example, the output unit (not shown) may include a display. The processor 110 may visualize (eg, overlay) a graphical representation indicating a calcified area in the medical image 131 through the display.

The memory 120 may store at least some frames or all frames of the medical image 131 . The memory 120 may store information about the detected calcified site (eg, location and size of the calcified site). Additionally, the memory 120 may temporarily and/or permanently store data and/or information required to perform a method for analyzing the medical image 131.

Figures 2 and 3 are flowcharts showing a medical image analysis method according to an embodiment.

First, a medical image analysis device (e.g., the medical image analysis device 100 of FIG. 1) may acquire a medical image having consecutive frame images through an image acquisition unit, as described above with reference to FIG. 1. The medical image analysis device may capture a medical image of the blood vessels of the patient receiving the treatment, or may receive the medical image from an external imaging device that captured the medical image.

And in step 210, the medical image analysis device may determine the blood vessel area in the contrast medium frame image. For example, in step 311, the medical image analysis device may detect a blood vessel region in a contrast medium frame of an angiography image. The blood vessel area may represent an area corresponding to a blood vessel in an angiography image. A medical image analysis device can determine a vessel part from contrast medium frame images in which a contrast medium is observed among consecutive frame images. The blood vessel area is a area corresponding to a blood vessel in the contrast medium frame image, and the medical image analysis device can extract at least a portion of the blood vessel area as the blood vessel area.

In step 302, the medical image analysis device may detect the lesion. A medical image analysis device can detect a blood vessel area from frame images of a medical image and determine a lesion portion of the detected blood vessel area. The lesion may represent a region in which the lesion is located among the vascular regions. For example, in step 302-1, the medical image analysis device may determine visual characteristics. The visual feature is a feature related to the visual appearance of the blood vessel region and/or blood vessel portion, and may include at least one of the boundary line of the blood vessel region and/or blood vessel portion, the center line, and the blood vessel segment into which the blood vessel portion is divided. there is. In step 302-2, the medical image analysis device may detect a lesion (lesion portion) through a trend line of diameter. Determination of visual features, including borders, midlines, and vessel segments, and detection of lesions are described in Figure 4 below. A medical image analysis device may set a region of interest (ROI) including a lesion for frame images of a medical image. For reference, the medical image analysis device may select a blood vessel area based on a set region of interest. For example, a medical image analysis device may extract a region included in a region of interest among blood vessel regions as a blood vessel region.

Next, in step 220, the medical image analysis device may determine a calcium candidate region in the non-contrast frame image. The medical image analysis apparatus may determine at least one calcium candidate region located around a blood vessel area from non-contrast frame images in which no contrast medium is observed among consecutive frame images. A calcium candidate region may be a candidate region where a calcified lesion is likely to exist. Determination of the calcium candidate region is explained in Figure 5 below.

And in step 230, the medical image analysis device may determine the calcified part based on comparison between the first movement direction of the calcium candidate area and the second movement direction of the blood vessel area. For example, the medical image analysis device may determine a calcification part of at least one calcium candidate region based on comparison between the first movement direction of the at least one calcium candidate area and the second movement direction of the blood vessel area due to the heartbeat. ) can be determined. In step 331, the medical image analysis device may determine whether the calcium candidate region includes a potential calcium site based on the correlation between the first movement direction and the second movement direction. Determination of potential calcium sites is explained in Figures 6 and 7 below.

Subsequently, in step 340, the medical image analysis device may determine the target calcium part based on the visual characteristics of the blood vessel part and the quantitative characteristics of the potentially calcification part. Quantitative characteristics of a potential calcified region may include the length and width of the calcified region. For example, the medical image analysis device may determine whether a potential calcified site is a target calcium site based on at least one of the length and width of the potential calcified site in quantitative characteristics. The medical image analysis device may determine that the potential calcium site is a target calcium site based on the length being within a predetermined length range. The medical image analysis device may determine the potential calcium site to be a target calcium site based on the fact that the width of the potential calcium site is less than or equal to a predetermined threshold width. The predetermined critical width may be illustratively determined as the width of a blood vessel adjacent to a potential calcified site (eg, a blood vessel site closest to a potential calcified site among blood vessel sites). The width of a potential calcified site may represent the length along an axis perpendicular to the longitudinal direction of the potential calcified site. The width of a blood vessel and/or blood vessel segment may represent the length along an axis perpendicular to the central axis of the blood vessel. As another example, a medical image analysis device may determine whether a potential calcified site is a target calcium site based on the extension direction of the potential calcified site and the central axis direction of the vascular site in the visual features. there is. Identification of target calcium sites based on the direction of extension of potential calcified sites is explained in Figure 8 below.

Figure 4 explains the operation of detecting a lesion (lesion portion) and extracting visual features of a blood vessel portion according to an embodiment.

According to one embodiment, a medical image analysis device (e.g., the medical image analysis device 100 of FIG. 1) may identify a blood vessel region from a contrast agent frame image of a medical image. For example, a medical image analysis device may detect points (eg, pixels) corresponding to blood vessels for each contrast medium frame image. For example, a medical image analysis device may extract a pixel mass (eg, a blob) with an intensity value less than a threshold brightness among the pixels of each contrast frame image. A medical image analysis device may determine a blood vessel area in a medical image based on a chunk of pixels having an intensity value less than a threshold brightness. As shown in Figure 4, blood vessels into which contrast agent has been injected may appear dark.

The medical image analysis device can identify the boundary 410 between the blood vessel area and the remaining area. The medical image analysis device may identify the boundary 410 between the blood vessel area and the remaining area based on pixel intensity. For example, a medical image analysis device may extract the boundary 410 between a blood vessel area and the remaining area using a machine learning model from each frame image of a medical image (e.g., blood vessel image). The machine learning model is a model designed to output pixels corresponding to the boundary 410 among the pixels of the frame image, and may include a neural network (e.g., deep neural network (DNN)) as an example, but is not limited to this. no. For another example, a medical image analysis device may calculate an edge based on a gradient value for each pixel of the medical image and identify the boundary 410 based on the calculated edge. Additionally, the medical image analysis device can extract the center line 420 of the blood vessel area. The center line 420 may represent a line connecting central points of the blood vessel (eg, center points) along the longitudinal direction of the blood vessel. The medical image analysis device may segment the blood vessel segment 430 among the blood vessel areas. The blood vessel segment 430 may represent a blood vessel region and/or blood vessel portion divided into unit sizes based on branches. The medical image analysis device may calculate a trend line 440 for the diameter of the blood vessel based on the center line 420. In the trend line 440 shown in FIG. 4, the horizontal axis may represent the length along the center line 420 of the blood vessel from the starting point of the blood vessel (e.g., the point where the catheter for contrast agent injection is inserted). The vertical axis may represent the diameter of the blood vessel (e.g., blood vessel width) at the corresponding point for each length of the center line 420.

A medical image analysis device may detect a lesion based on the trend line 440. For example, when a section 441 in which the diameter along the length of the center line 420 decreases and then increases is detected, the medical image analysis device may detect a portion corresponding to the section 441 as a lesion. As described above, the medical image analysis device can set a region of interest including a lesion. For example, the medical image analysis device may set the area where the lesion is located at the center of the medical image as the area of interest.

Figure 5 illustrates detection of a calcium candidate region according to one embodiment.

A medical image analysis device (e.g., the medical image analysis device 100 of FIG. 1) according to an embodiment detects a calcium candidate region in the non-contrast frame image 501 and detects a blood vessel region 520 in the contrast frame image 502. ) can be detected.

The medical image analysis device can detect a calcium candidate region from the non-contrast frame image 501. The medical image analysis device extracts a chunk of pixels having an intensity value less than a threshold brightness from the non-contrast frame image 501, and selects chunks of pixels located within the blood vessel and/or around the blood vessel among the extracted pixel chunks as calcium candidate areas. You can decide. As shown in FIG. 5 , the calcium candidate region may have a shape that extends parallel to the longitudinal direction of the blood vessel region 520 (eg, the central axis direction of the blood vessel). However, this is purely illustrative and is not limited thereto.

The medical image analysis device may determine whether to perform analysis on the calcium candidate region using the blood vessel region 520 detected in the contrast medium frame image 502. For example, the medical image analysis device may determine a position corresponding to the blood vessel portion 520 in the non-contrast frame image 501 based on the blood vessel detection result in step 210 described above in FIG. 2 . For another example, the medical image analysis device determines a position 512 corresponding to the calcium candidate region in the contrast agent frame image 502 based on the detection result 511 of the calcium candidate region in the non-contrast agent frame image 501. can be decided. The medical image analysis device may exclude a calcium candidate region detected at a position exceeding a critical distance from the blood vessel region 520 among at least one calcium candidate region from calcium site determination. If the distance 530 (e.g., the shortest distance) from the calcium candidate region to the border of the blood vessel exceeds the critical distance, the medical image analysis device may exclude the calcium candidate region from determining the calcium site. The medical image analysis device may perform calcium site determination according to step 230 on the calcium candidate region detected at a position less than or equal to the critical distance from the blood vessel site 520. The calcium candidate area that appears at a location exceeding the critical distance may be a lump of calcium that is not narrowed in a blood vessel and is attached to another organ or moving around in the body. Calcium lumps narrowed in blood vessels can also be called calcified plaques.

The critical distance may be determined based on the diameter of the blood vessel and the frame rate of the medical image. For example, the critical distance may be determined as a distance corresponding to 1 times the diameter of the blood vessel. Since the diameter of the blood vessel varies at each point along the center line of the blood vessel and the critical distance is determined based on the diameter of the blood vessel at that point, the critical distance may also vary depending on the diameter of the blood vessel at the center line point corresponding to the detected calcium candidate region. The critical distance may vary depending on the blood vessel diameter at the point where the straight line having the shortest distance from the calcium candidate area to the center line of the blood vessel intersects the center line of the blood vessel. However, the critical distance is not limited to 1 times the diameter of the blood vessel, and the critical distance may be determined to be a distance of 0.5 to 1 times the diameter of the blood vessel. There may be uncertainty because the frame rate of medical images is lower than the subject's heart rate. If there is no uncertainty, the critical distance may be set to 0.5 times the vessel diameter, but if a margin is secured, the critical distance may be set to 1 times the aforementioned vessel diameter.

For reference, although it is shown in Figure 5 that determination of the calcium candidate area is performed on the entire area of the medical image, the process is not limited to this. The medical image analysis device may detect a calcium candidate region based on a region of interest (ROI) including the lesion detected in step 302 described above in FIG. 3 . For example, the medical image analysis device may determine the blood vessel region 520 and at least one calcium candidate region by performing the operation described in FIG. 5 within the region of interest.

Figures 6 and 7 illustrate the first movement direction of the calcium candidate region and the second movement direction of the blood vessel site according to one embodiment.

Blood vessels are a type of tube, and plaque can grow within their walls. Some of the grown plaque may become calcified. Since calcified plaque is attached to the blood vessel wall, it may exhibit the same and/or similar movement at the same and/or similar frequency as the blood vessel moved by the heartbeat. As shown in FIG. 6, the medical image analysis device (e.g., the medical image analysis device 100 of FIG. 1) determines the first movement direction 615 of the calcium candidate region 610 from the non-contrast frame images 601. can be calculated. The medical image analysis device may calculate the second movement direction 625 of the blood vessel area 620 from the contrast medium frame images 602. The medical image analysis device determines that the calcium candidate area 610 is a calcium lump narrowed in the blood vessel area 620 (e.g., a calcified area) based on the comparison between the first movement direction 615 and the second movement direction 625. It can be determined whether or not it contains .

According to one embodiment, the medical image analysis device may determine the first movement direction and the second movement direction due to the heartbeat based on the period of the patient's electrocardiogram signal. The medical image analysis device may determine a first movement direction and a second movement direction for the same type of cardiac cycle among a plurality of cardiac cycles (eg, atrial systole, ventricular systole, and ventricular diastole) according to heart rate. For example, the medical image analysis device may calculate the first movement direction of the calcium candidate region 610 using two or more non-contrast frames belonging to the target cardiac cycle (e.g., atrial systole) among the plurality of non-contrast frames. there is. Similarly, the medical image analysis device may select two or more contrast agents belonging to a target cardiac cycle (e.g., atrial systole as the same type of cardiac cycle used to calculate the first movement direction of the calcium candidate region described above) among the plurality of contrast agent frames. The second movement direction of the blood vessel portion 620 can be calculated using the frames.

The medical image analysis device is based on the fact that the correlation score between the first motion vector calculated for at least one calcium candidate region 615 and the second motion vector calculated for the blood vessel region 620 exceeds a threshold. Accordingly, it can be determined that at least one calcium candidate region 615 includes a potential calcified site. A potential calcified region is a region estimated to be a calcium lump in non-contrast frames and may represent a region that exhibits the same and/or similar movement as the blood vessel region 620. The correlation score is a score indicating correlation and/or similarity between the first motion vector and the second motion vector, and may be calculated as an Euclidean distance. However, the correlation score is not limited to this. If the correlation score between two motion vectors exceeds the threshold, the two motion vectors may be interpreted as having the same and/or similar movements. The threshold may represent a numerical value that serves as a standard for determining whether the movements of two vectors are the same and/or similar. The first motion vector may be a vector representing the first movement direction 615, and the second motion vector may be a vector representing the second movement direction 625. The first motion vector and the second motion vector are explained in FIG. 7 below.

Figure 7 explains the calculation of motion vectors.

For example, the medical image analysis device may be configured to determine the location of at least one calcium candidate region 710 in the first non-contrast frame 701-1 among the non-contrast frame images. A motion vector 731 representing motion from the second non-contrast frame 701-2 to the position of at least one calcium candidate region 710 may be calculated as the first movement direction. The medical image analysis apparatus may calculate a plurality of first motion vectors and determine the first average motion vector 741 of the calculated plurality of motion vectors as the first movement direction. For example, in FIG. 7 , a motion vector 731 between the second non-contrast frame 701-2 and the third non-contrast frame 701-3 may also be calculated.

The medical image analysis device detects blood vessels in the second contrast medium frame 702-2 after the first contrast medium frame 702-1 from the position of the blood vessel portion 720 in the first contrast medium frame 702-1 among the contrast medium frame images. A motion vector 732 representing motion to the position of the portion 720 may be calculated as the second movement direction. The medical image analysis apparatus may calculate a plurality of second motion vectors and determine the second average motion vector 742 of the calculated plurality of motion vectors as the second movement direction. For example, in FIG. 7 , a motion vector 732 between the second contrast medium frame 702-2 and the third non-contrast frame 702-3 may also be calculated.

The motion vectors described above may be determined based on perfusion. Perfusion can be estimated from the results of tracking the movement of corresponding points (eg, pixels) between consecutive frames (eg, consecutive non-contrast frames or consecutive contrast media frames). However, the determination of the motion vector is not limited to this.

For reference, there may be an intensity difference between contrast frames and non-contrast frames, but this can be compensated for to minimize the overall intensity difference through image preprocessing.

Additionally, the medical image analysis device may, when the ratio of the movement distance of at least one calcium candidate region 710 to the movement distance of the blood vessel region 720 is less than or equal to the threshold ratio, the at least one calcium candidate region 710 is configured to target calcium. You may decide not to include the region.

Figure 8 is a diagram explaining the determination of a calcified part according to an embodiment.

A medical image analysis device (e.g., the medical image analysis device 100 of FIG. 1) according to an embodiment may display visual features in the extension direction of a potential calcified area and the central axis direction 821 of the blood vessel area 820. Based on this, it can be determined whether the potential calcified site shown in the medical image 801 is the target calcium site.

The medical image analysis device may identify the central axis direction 821 of the blood vessel region 820 based on the blood vessel structure data 890. For example, blood vessel structure data 890 may include branching points of blood vessels and connection information of branched blood vessels. Blood vessel structure data 890 may include node data indicating branching points and edge data indicating branched blood vessels. Since a branched blood vessel is connected to two different branch points, two node data can be connected to one edge data. Connection information of branch points and branched blood vessels may be information indicating the connection relationship between branch points and branched blood vessels. Connection information can be created using edge data mapped to node data and node data to which the edge data is connected. The blood vessel structure data 890 may include a tree structure in which nodes and edges are connected with the branch point closest to the blood vessel introduction portion as the root node. The root node is a node corresponding to the highest branch point, and may be, for example, a node corresponding to the starting point. The medical image analysis device identifies the blood vessel branch to which the detected blood vessel area 820 belongs, and determines the central axis direction 821 from the upper branch to the lower branch in the blood vessel area 820 with reference to the above-mentioned blood vessel structure data 890. can be identified.

The medical image analysis device can identify the extension direction 811 of the candidate calcium region. For example, since calcium lumps are formed long within blood vessels, the candidate calcium region may also have a long shape along one axis (eg, the axis corresponding to the extension direction 811). As an example, the medical image analysis device may determine the direction corresponding to the line connecting the center points in the candidate calcium region as the extension direction 811. However, it is not limited to this.

The medical image analysis device can compare whether the extension direction 811 of the candidate calcium area matches the central axis direction 821 of the blood vessel area 820. For example, if the angle difference between the extension direction 811 and the central axis direction 821 is less than or equal to the critical angle, the medical image analysis device may ultimately determine that the candidate calcium area includes a calcified area on the inner wall of the blood vessel. . For another example, if the angle difference between the extension direction 811 and the central axis direction 821 exceeds the critical angle, the medical image analysis device may determine that the candidate calcium area is not a calcified area on the inner wall of the blood vessel. there is. This is because calcium lumps formed on the inner wall of blood vessels cannot be formed in a direction perpendicular to or intersecting the longitudinal direction of the blood vessel (e.g., central axis direction 821).

In addition, the medical image analysis device also considers the quantitative characteristics of the calcium candidate region 810 (e.g., the length and width of the calcium candidate region 810) and ultimately determines whether the calcium candidate region 810 includes a calcified site. can be determined.

Arteriosclerosis is a disease in which the inner diameter through which blood flows within blood vessels narrows, and can be caused by plaque stenosis. The medical image analysis device according to one embodiment can accurately identify the lesion type, whether the lesion is calcium, when there is a lesion in which plaque stenosis has occurred. In addition, the medical image analysis device can accurately detect calcified areas (e.g., calcified areas) even in X-ray-based angiography images, even if they are not CT.

The embodiments described above may be implemented with hardware components, software components, and/or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, and a field programmable gate (FPGA). It may be implemented using a general-purpose computer or a special-purpose computer, such as an array, programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and software applications running on the operating system. Additionally, a processing device may access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, a single processing device may be described as being used; however, those skilled in the art will understand that a processing device includes multiple processing elements and/or multiple types of processing elements. It can be seen that it may include. For example, a processing device may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are possible.

Software may include a computer program, code, instructions, or a combination of one or more of these, which may configure a processing unit to operate as desired, or may be processed independently or collectively. You can command the device. Software and/or data may be used on any type of machine, component, physical device, virtual equipment, computer storage medium or device to be interpreted by or to provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on a computer-readable recording medium.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. A computer-readable medium may include program instructions, data files, data structures, etc., singly or in combination, and the program instructions recorded on the medium may be specially designed and constructed for the embodiment or may be known and available to those skilled in the art of computer software. It may be possible. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -Includes optical media (magneto-optical media) and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine language code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

The hardware devices described above may be configured to operate as one or multiple software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with limited drawings, those skilled in the art can apply various technical modifications and variations based on this. For example, the described techniques are performed in a different order than the described method, and/or components of the described system, structure, device, circuit, etc. are combined or combined in a different form than the described method, or other components are used. Alternatively, appropriate results may be achieved even if substituted or substituted by an equivalent.

Therefore, other implementations, other embodiments, and equivalents of the claims also fall within the scope of the claims described below.

Claims (20)

In a medical image analysis device,
An image acquisition unit that acquires an X-ray-based Cardiac Angiography (CAG) image having consecutive frame images;
A vessel part is determined from contrast medium frame images in which a contrast medium is observed among the consecutive frame images, and a vessel part is located around the vessel part from non-contrast frame images in which a contrast medium is not observed among the continuous frame images. Determining at least one located calcium candidate area, based on a similarity comparison between a first movement direction of the at least one calcium candidate area in the non-contrast frame images and a second movement direction of the blood vessel region in the contrast frame images due to heartbeat A processor that determines whether the at least one calcium candidate region includes a calcification part.
A medical image analysis device comprising: According to paragraph 1,
The processor,
Detecting a blood vessel region from frame images of the angiography image, and determining a lesion among the detected blood vessel regions,
Determining the vascular region and the at least one calcium candidate region within a region of interest (ROI) containing the lesion,
Medical image analysis device. According to paragraph 1,
The processor,
Extracting a visual feature including at least one of the boundary line of the blood vessel region, the center line, and the blood vessel segment into which the blood vessel region is divided,
Determining the target calcium site based on the visual features and quantitative characteristics of the potentially calcification part,
Medical image analysis device. According to paragraph 3,
The processor,
Based on the correlation score between the first motion vector calculated for the at least one calcium candidate region and the second motion vector calculated for the blood vessel region exceeding a threshold, the at least one calcium candidate region Determining that this contains the potential calcified site,
Medical image analysis device. According to paragraph 3,
The processor,
Based on at least one of the length and width of the potential calcified site in the quantitative characteristic, determining whether the potential calcified site is a target calcium site,
Medical image analysis device. According to paragraph 3,
The processor,
Based on the extension direction of the potential calcified site and the central axis direction of the vascular site in the visual feature, determining whether the potential calcified site is a target calcium site,
Medical image analysis device. According to paragraph 1,
The processor,
Motion from the position of the at least one calcium candidate area in a first non-contrast frame among the non-contrast frame images to the position of the at least one calcium candidate area in a second non-contrast frame after the first non-contrast frame ( calculating a motion vector representing motion as the first movement direction,
Medical image analysis device. According to paragraph 1,
The processor,
Calculating a motion vector representing motion from the position of the blood vessel portion in a first contrast medium frame among the contrast medium frame images to the position of the blood vessel portion in a second contrast medium frame after the first contrast medium frame as the second movement direction. ,
Medical image analysis device. According to paragraph 1,
The processor,
Excluding from calcium site determination a calcium candidate region detected at a position exceeding a critical distance from the blood vessel site among the at least one calcium candidate region,
Medical image analysis device. According to paragraph 1,
The processor,
Determining the first movement direction and the second movement direction due to the heartbeat based on the period of the patient's electrocardiogram signal,
Medical image analysis device. According to paragraph 1,
The processor,
When the ratio of the movement distance of the at least one calcium candidate region to the movement distance of the blood vessel site is less than or equal to a threshold ratio, determining that the at least one calcium candidate region does not include a target calcium site,
Medical image analysis device. delete In a medical image analysis method implemented with a processor,
Acquiring an X-ray based angiography image having successive frame images;
determining a vessel part from contrast medium frame images in which the contrast medium is observed among the consecutive frame images;
determining at least one calcium candidate area located around the blood vessel area from non-contrast frame images in which no contrast medium is observed among the consecutive frame images; and
The at least one calcium candidate region is calcified based on similarity comparison between a first movement direction of the at least one calcium candidate region of the non-contrast frame images due to heartbeat and a second movement direction of the blood vessel portion of the contrast frame images. Step to determine whether it contains a calcification part
A medical image analysis method comprising: According to clause 13,
Extracting a visual feature including at least one of a boundary line, a center line, and a blood vessel segment into which the blood vessel region is divided; and
Confirming the target calcium site based on the visual features and quantitative characteristics of the potentially calcification part.
A medical image analysis method further comprising: According to clause 13,
The step of determining whether the at least one calcium candidate region includes a calciumd site,
Motion from the position of the at least one calcium candidate area in a first non-contrast frame among the non-contrast frame images to the position of the at least one calcium candidate area in a second non-contrast frame after the first non-contrast frame ( calculating a motion vector representing motion as the first movement direction,
A medical image analysis method comprising: According to clause 13,
The step of determining whether the at least one calcium candidate region includes a calciumd site,
Calculating a motion vector representing motion from the position of the blood vessel portion in a first contrast medium frame among the contrast medium frame images to the position of the blood vessel portion in a second contrast medium frame after the first contrast medium frame as the second movement direction. step,
A medical image analysis method comprising: According to clause 13,
The step of determining the at least one calcium candidate region includes:
Excluding from calcium site determination a calcium candidate region detected at a position exceeding a critical distance from the blood vessel site among the at least one calcium candidate region.
A medical image analysis method comprising: According to clause 13,
The step of determining whether the at least one calcium candidate region includes a calciumd site,
determining the first movement direction and the second movement direction due to the heartbeat based on the period of the patient's electrocardiogram signal;
A medical image analysis method comprising: delete A computer-readable recording medium storing one or more computer programs including instructions for performing the method of any one of claims 13 to 18.

Images