Blood vessel image segmentation method, electronic device, and storage medium

Abstract

The application provides a blood vessel image segmentation method, an electronic device and a storage medium. The method comprises the following steps: using a preset algorithm to explore a center point on a center line of a blood vessel in a to-be-segmented blood vessel image, so as to obtain position information of each center point and cross-section data of a blood vessel cross-section where the center point is located; for each center point, a corresponding limited range is determined according to the position information of the center point; taking the center point as a first initial seed point, taking an upper limit pixel value of the blood vessel cross-section where the center point is located as a first upper limit threshold value of region growing, and taking a lower limit pixel value corresponding to the blood vessel cross-section where the center point is located as a first lower limit threshold value of region growing, region growing is performed in the limited range, so as to obtain a corresponding blood vessel segmentation result; and the blood vessel segmentation results of the limited ranges corresponding to the center points are superimposed, so as to obtain a blood vessel segmentation result corresponding to the to-be-segmented blood vessel image. The application can effectively improve the precision and speed of blood vessel segmentation.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method for segmenting blood vessel images, an electronic device, and a storage medium. Background Technology

[0002] The aorta is the largest and thickest artery in the human systemic circulation. It originates from the left ventricle of the heart, passing through the ascending aorta, aortic arch, and descending aorta in the chest, forming the thoracic aorta. The thoracic aorta descends to the abdominal aorta, where it branches into the celiac trunk, superior mesenteric artery, and left and right renal arteries. The abdominal aorta then descends to the left and right common iliac arteries, branching into the internal and external iliac arteries. The external iliac artery descends in sequence to the femoral artery, popliteal artery, anterior tibial artery, and posterior tibial artery. The aortic arch ascends to the left subclavian artery, left common carotid artery, and brachiocephalic trunk. Finally, the larger arteries of the aorta deliver blood to the smaller arteries throughout the body.

[0003] Common aortic diseases include aortic dissection and aortic aneurysm. The causes and structural characteristics of the aorta differ between the two. Aortic dissection occurs when blood flows from the tear in the aortic intima into the aortic media, causing blood to accumulate in the aortic media. Due to blood flow direction and gravity, this blood extends towards the descending aorta, creating a separation between the true and false lumens of the aorta. An aortic aneurysm is defined as a pathological dilation of the aorta exceeding 50% of its normal diameter; the most common pathological dilation is atherosclerosis. Surgeons aim to simulate aortic diseases to identify the optimal surgical approach, reduce surgical risks, and increase the success rate. To this end, computers need to provide surgeons with quantitative aortic bioparameters, including but not limited to the aortic cross-sectional area, circumference, and maximum and minimum diameters.

[0004] Traditional image segmentation methods involve manually segmenting the target object first, followed by automatic segmentation using computer-aided algorithms, or a combination of manual and computer-aided segmentation for semi-automatic segmentation. Traditional algorithms for aortic segmentation include edge detection, thresholding, and region growing. In recent years, with the rapid development of deep learning, a class of deep learning algorithms based on convolutional structures has emerged for aortic segmentation. These machine learning and deep learning-based aortic segmentation methods have shown good performance in terms of segmentation accuracy, algorithm performance (computation speed, memory usage), and data generalization ability within localized vascular regions such as intracranial arteries, carotid arteries, and coronary arteries.

[0005] However, in large vascular areas, the effect of vascular segmentation is less than satisfactory. The reasons are as follows:

[0006] 1. Existing algorithms suffer from unstable segmentation accuracy when segmenting the aorta. Due to aortic dissection, single or multiple intimal tears often occur. Intimal tears at the same location can present as double, triple, or even multi-lobed shapes. The false lumen of the aorta can severely compress the true lumen, leading to extreme narrowing of the true lumen in 3D imaging space. Single / multiple tears, double / multi-lobed tear shapes, and extreme narrowing of the true lumen all contribute to the instability of the true and false lumen segmentation accuracy obtained by existing algorithms.

[0007] 2. Existing segmentation algorithms suffer from slow computation speed and high memory consumption, primarily due to two factors: First, the three-dimensional aorta has a high dimension and large area. CT and MR image sequences are a series of continuous two-dimensional cross-sectional image sequences, with each cross-section measuring 512*512 pixels. The number of sequences varies depending on the size of the targeted organ. The larger the vascular region of the three-dimensional target organ, the more exponentially the computation time and memory consumption of existing segmentation algorithms increase. As mentioned earlier, the three-dimensional aortic region of interest to doctors can cover the area from the neck to the lower leg, far exceeding the size of intracranial arteries in the brain, carotid arteries in the neck, and coronary arteries in the heart, thus slowing down the aortic segmentation speed and increasing memory consumption of existing algorithms. Second, while deep learning, which has become popular in recent years, is faster in inference tasks, inference requires reading and writing large-scale three-dimensional aortic data between the CPU and GPU, which is slow and memory-intensive. Therefore, segmentation of large-scale three-dimensional aortic vessels requires faster computation speed and lower memory consumption.

[0008] 3. Existing aortic segmentation algorithms lack generalization ability for aortic disease data. The morphologies of aortic dissection and aortic aneurysm differ. Dissection is characterized by a false lumen compressing the true lumen, while aneurysm is characterized by its excessive size in a localized area of ​​the aorta. Segmentation algorithms suitable for aortic dissection data often suffer from poor segmentation speed and accuracy for aortic aneurysms. Furthermore, even for the same disease, such as aortic aneurysm, the computational speed and accuracy of aortic segmentation are unstable due to the large differences in aneurysm volume. While deep learning segmentation algorithms increase the probability of successful vessel segmentation, they cannot directly guarantee the accuracy of the segmentation results.

[0009] It should be noted that the information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0010] The purpose of this invention is to provide a method, electronic device, and storage medium for segmenting blood vessel images, which can effectively improve the accuracy and speed of blood vessel segmentation.

[0011] To achieve the above objectives, the present invention provides a method for segmenting blood vessel images, comprising:

[0012] A preset algorithm is used to explore the center point of the blood vessel centerline in the blood vessel region of the acquired blood vessel image to be segmented, so as to obtain the position information of each center point and the cross-sectional data of the blood vessel cross section where the center point is located. The cross-sectional data includes the upper limit pixel value and the lower limit pixel value corresponding to the blood vessel cross section.

[0013] For each center point, a defined range corresponding to the center point is determined on the blood vessel image to be segmented based on the position information of the center point.

[0014] For each center point, the center point is used as the first initial seed point, the upper limit pixel value corresponding to the blood vessel cross section where the center point is located is used as the first upper limit threshold for region growth, and the lower limit pixel value corresponding to the blood vessel cross section where the center point is located is used as the first lower limit threshold for region growth. Region growth is performed within the limited range corresponding to the center point to obtain the blood vessel segmentation result within the limited range corresponding to the center point.

[0015] The blood vessel segmentation results corresponding to the defined range of each center point are superimposed to obtain the blood vessel segmentation result corresponding to the blood vessel image to be segmented.

[0016] Optionally, the vascular region includes one main branch vessel and two first branch vessels, wherein the two first branch vessels converge at the terminal region of the main branch vessel.

[0017] The process of exploring the center point of the blood vessel centerline in the acquired blood vessel image to be segmented using a preset algorithm includes:

[0018] The first preset algorithm is used to explore the center point on the center line of the two first branch vessels simultaneously until the first termination condition is met. Then, the one of the termination center points of the two first branch vessels located at the higher position is taken as the first intersection point of the center line of the two first branch vessels and replaces the termination center point of the other first branch vessel.

[0019] The second preset algorithm is used to explore the center point on the center line of the main branch vessel until the second termination condition is met, then the first intersection point is taken as the termination center point of the main branch vessel.

[0020] Optionally, when the center point is located on the centerline of the main branch vessel, determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes:

[0021] Using the location of the center point as the bottom center of the first cuboid, the first preset length as the length of the first cuboid, the first preset width as the width of the first cuboid, and the distance between the center point and the adjacent center point located above the center point as the height of the first cuboid, the defined range of the first cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

[0022] Optionally, the blood vessel image to be segmented is an aortic blood vessel image, and the main branch blood vessel includes the aortic arch. When the center point is located at the highest position on the midline of the main branch blood vessel, determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes:

[0023] Based on the center point, the center of the bottom surface of the second cuboid is determined, the second preset length is used as the length of the second cuboid, the second preset width is used as the width of the second cuboid, and the first preset height is used as the height of the second cuboid. The defined range of the second cuboid corresponding to the center point is then determined on the blood vessel image to be segmented.

[0024] Optionally, the blood vessel image to be segmented is an aortic blood vessel image. When the center point is the termination center point of the main branch blood vessel, determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes:

[0025] Using the location of the center point as the center of the bottom surface of the third cuboid, the third preset length as the length of the third cuboid, the third preset width as the width of the third cuboid, and the third preset height as the height of the third cuboid, the defined range of the third cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

[0026] Optionally, the vascular region further includes at least one second branch vessel intersecting with the main branch vessel, and the step of exploring the center point of the vascular centerline of the vascular region in the acquired vascular image to be segmented using a preset algorithm further includes:

[0027] The third preset algorithm is used to explore the center point on the center line of the second branch vessel until the third termination condition is met. Then, the center point where the distance between the termination center point of the main vessel and the termination center point of the second branch vessel is less than the third threshold is determined as the second intersection point corresponding to the second branch vessel and replaces the termination center point of the second branch vessel.

[0028] Optionally, when the center point is located on the centerline of the first branch vessel or the second branch vessel, the cross-sectional data of the vessel cross-section where the center point is located also includes the maximum radius of the vessel cross-section. Determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes:

[0029] Using the location of the center point as the center of the cube and twice the maximum radius of the blood vessel cross-section where the center point is located as the side length of the cube, the defined range of the cube corresponding to the center point is determined on the blood vessel image to be segmented.

[0030] Optionally, the step of using a first preset algorithm to explore the center point on the centerline of the first branch vessel until a first termination condition is met includes:

[0031] Step A1: Determine the initial center point on the center line of the first branch blood vessel based on the first endpoint obtained in the first branch blood vessel, and use the initial center point as the current forward point, and use the unit vector parallel to the Z axis as the current forward direction vector.

[0032] Step A2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the first branch blood vessel, and take the explored new center point as the current center point of the first branch blood vessel;

[0033] Step A3: Determine whether the distance between the current center point of the first branch vessel and the current center point of another first branch vessel is less than a first threshold.

[0034] If yes, proceed to step A4; otherwise, proceed to step A5.

[0035] Step A4: Take the current center point of the first branch vessel as the termination center point of the first branch vessel and end the exploration of the center point on the center line of the first branch vessel.

[0036] Step A5: Take the current center point of the first branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step A2.

[0037] Optionally, the step of exploring the next new center point on the centerline of the first branch vessel based on the current forward point and the current forward direction vector includes:

[0038] Based on the current forward point and the current forward direction vector, extract the current two-dimensional cross-section of interest image from the medical image to be segmented;

[0039] The current two-dimensional cross-sectional image of interest is segmented to obtain the current two-dimensional blood vessel region;

[0040] The current two-dimensional blood vessel region is segmented into true and false lumens to segment out the current two-dimensional blood vessel true lumen region;

[0041] The new center point of the first branch vessel is obtained based on the center point of the current two-dimensional vascular lumen region.

[0042] Optionally, extracting the current two-dimensional section of interest image from the medical image to be segmented based on the current forward point and the current forward direction vector includes:

[0043] Based on the current forward point and the current forward direction vector, determine the center point of the current two-dimensional section of interest and the normal vector of the current two-dimensional section of interest, wherein the normal vector of the current two-dimensional section of interest is the current forward direction vector;

[0044] Based on the preset size, the center point of the current two-dimensional section of interest, and the normal vector of the current two-dimensional section of interest, the current two-dimensional section of interest image is extracted from the medical image to be segmented.

[0045] Optionally, determining the center point of the current two-dimensional cross-section of interest based on the current forward point and the current forward direction vector includes:

[0046] The current forward speed is determined based on the angle between the current forward direction vector and the previous forward direction vector;

[0047] The center point of the current two-dimensional cross-section of interest is determined based on the current forward point, the current forward direction vector, and the current forward velocity.

[0048] Optionally, the method further includes:

[0049] The maximum radius of the current two-dimensional vascular lumen region is determined based on the center point of the current two-dimensional vascular lumen region.

[0050] Based on the maximum radius of the current two-dimensional vascular lumen region, determine the locking radius corresponding to the current two-dimensional vascular lumen region;

[0051] The step of segmenting the blood vessel region in the current two-dimensional cross-sectional image of interest to segment the current two-dimensional blood vessel region includes:

[0052] Based on the average gray value of the pixels in the two-dimensional blood vessel lumen region, determine the second upper limit threshold and the second lower limit threshold for region growth.

[0053] Based on the locking radius corresponding to the previous two-dimensional true lumen region of the blood vessel and the center point of the current two-dimensional section of interest, the current circular region is determined in the current two-dimensional section of interest image;

[0054] In the current circular domain, a pixel with a value greater than the lower threshold and less than the upper threshold is selected as the second initial seed point for region growth;

[0055] Based on the second initial seed point, the second upper threshold, and the second lower threshold, the region growing method is used to segment the blood vessel region of the current two-dimensional cross-sectional image of interest, so as to segment the current two-dimensional blood vessel region.

[0056] Optionally, determining the locking radius corresponding to the current two-dimensional vascular lumen region based on the maximum radius of the current two-dimensional vascular lumen region includes:

[0057] Determine whether the ratio of the square of the maximum radius of the previous two-dimensional vascular lumen region to the square of the maximum radius of the current two-dimensional vascular lumen region is greater than a preset ratio.

[0058] If so, the maximum radius of the previous two-dimensional true lumen region is used as the locking radius corresponding to the current two-dimensional true lumen region, so as to lock the locking radius of the current two-dimensional true lumen region.

[0059] If not, then the maximum radius of the current two-dimensional vascular lumen region is taken as the locking radius corresponding to the previous two-dimensional vascular lumen region.

[0060] Optionally, the step of segmenting the current two-dimensional blood vessel region into true and false lumens to segment out the current two-dimensional blood vessel true lumen region includes:

[0061] Perform connected component analysis on the current two-dimensional blood vessel region, and extract the connected component closest to the center point of the current two-dimensional cross-section of interest as the current two-dimensional blood vessel true lumen region.

[0062] Optionally, the upper and lower pixel values ​​corresponding to the blood vessel cross section where the center point is located are determined based on the average gray value of the pixels in the two-dimensional true lumen blood vessel region where the center point is located.

[0063] Optionally, determining the initial center point on the centerline of the first branch vessel based on the obtained first endpoint within the first branch vessel includes:

[0064] Using the first endpoint as the center point of the initial two-dimensional section of interest, and the unit vector parallel to the Z-axis as the normal vector of the initial two-dimensional section of interest, the initial two-dimensional section of interest image is extracted from the medical image to be segmented.

[0065] The initial two-dimensional cross-sectional image of interest is segmented into a vascular region to extract the initial two-dimensional vascular region;

[0066] The initial two-dimensional vascular region is segmented into true and false lumens to separate the initial two-dimensional vascular true lumen region;

[0067] The initial center point of the first branch vessel is obtained based on the center point of the initial two-dimensional vascular lumen region.

[0068] Optionally, the step of using a second preset algorithm to explore the center point on the centerline of the main branch vessel until the second termination condition is met includes:

[0069] Step B1: Determine the initial center point on the center line of the main branch vessel based on the obtained second endpoint within the main branch vessel, and use the initial center point as the current forward point, and use the unit vector parallel to the Z-axis as the current forward direction vector.

[0070] Step B2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the main branch vessel, and take the explored new center point as the current center point of the main branch vessel.

[0071] Step B3: Determine whether the distance between the current center point of the main branch vessel and the first intersection point is less than the second threshold.

[0072] If yes, proceed to step B4; otherwise, proceed to step B5.

[0073] Step B4: Replace the current center point with the first intersection point as the termination center point of the main branch vessel and end the exploration of the center point on the center line of the main branch vessel.

[0074] Step B5: Take the current center point of the main branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step B2.

[0075] Optionally, the step of using a third preset algorithm to explore the center point on the centerline of the second branch vessel until a third termination condition is met includes:

[0076] Step C1: Determine the initial center point on the center line of the second branch vessel based on the obtained third endpoint within the second branch vessel, and use the initial center point as the current forward point, and use the unit vector pointing from the third endpoint to the fourth endpoint corresponding to the second branch vessel as the current forward direction vector, wherein the fourth endpoint is the center point on the center line of the main branch vessel that is closest to the third endpoint of the second branch vessel.

[0077] Step C2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the second branch blood vessel, and take the explored new center point as the current center point of the second branch blood vessel;

[0078] Step C3: Determine whether the minimum distance between the current center point of the second branch vessel and each center point of the main branch vessel is less than the third threshold.

[0079] If yes, proceed to step C4; otherwise, proceed to step C5.

[0080] Step C4: Take the current center point as the termination center point of the second branch vessel and end the exploration of the center point on the center line of the second branch vessel;

[0081] Step C5: Take the current center point of the second branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step C2.

[0082] To achieve the above objectives, the present invention also provides an electronic device, including a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, it implements the blood vessel image segmentation method described above.

[0083] To achieve the above objectives, the present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, implements the blood vessel image segmentation method described above.

[0084] Compared with existing technologies, the blood vessel image segmentation method, electronic device, and storage medium provided by this invention have the following advantages:

[0085] The vascular image segmentation method provided by this invention first explores the center points on the center lines of the vascular regions in the acquired vascular image to be segmented using a preset algorithm to obtain the position information of each center point and the cross-sectional data of the vascular cross section where the center point is located. The cross-sectional data includes the upper limit pixel value and the lower limit pixel value corresponding to the vascular cross section. Then, for each center point, a limited range corresponding to the center point is determined on the vascular image to be segmented based on the position information of the center point. Then, for each center point, region growth is performed within the limited range corresponding to the center point, using the center point as a first initial seed point, the upper limit pixel value corresponding to the vascular cross section where the center point is located as a first upper limit threshold for region growth, and the lower limit pixel value corresponding to the vascular cross section where the center point is located as a first lower limit threshold for region growth, to obtain the vascular segmentation result within the limited range corresponding to the center point. Finally, the vascular segmentation results within the limited ranges corresponding to each center point are superimposed to obtain the vascular segmentation result corresponding to the vascular image to be segmented. Therefore, this invention can transform three-dimensional blood vessel (e.g., aorta) segmentation into blood vessel segmentation within defined ranges guided by the blood vessel centerline, thereby achieving rapid and convenient blood vessel segmentation while significantly improving the accuracy of segmentation. Compared to existing deep learning algorithms, this invention greatly reduces memory usage during blood vessel segmentation, significantly increasing segmentation speed. This invention can achieve rapid and convenient blood vessel segmentation over large areas, while also improving the generalization of data processing. Furthermore, since the first upper and lower thresholds for region growing within the defined range corresponding to each center point are different, the segmentation accuracy of the blood vessel image segmentation method provided by this invention can be further effectively improved.

[0086] Since the electronic device and storage medium provided by this invention belong to the same inventive concept as the blood vessel image segmentation method provided by this invention, the electronic device and storage medium provided by this invention have all the advantages of the blood vessel image segmentation method provided by this invention. Therefore, the beneficial effects of the electronic device and storage medium provided by this invention will not be described in detail here. Attached Figure Description

[0087] Figure 1 This is a flowchart illustrating a blood vessel image segmentation method according to an embodiment of the present invention.

[0088] Figure 2 This is a schematic diagram of aortic segmentation including aortic dissection, provided as a specific example of the present invention;

[0089] Figure 3 This is a schematic diagram of aortic segmentation including an aortic aneurysm, provided as a specific example of the present invention;

[0090] Figure 4 A schematic diagram illustrating the exploration process of the center point on the centerline of the first branch blood vessel according to an embodiment of the present invention;

[0091] Figure 5 A schematic diagram illustrating the principle of exploring the centerline of a vascular region, provided for a first specific example of the present invention;

[0092] Figure 6a This is a schematic diagram of a one- or two-dimensional cross-sectional image of interest.

[0093] Figure 6b A schematic diagram illustrating the determination of the circular region based on the lock radius;

[0094] Figure 6c This is a schematic diagram of a two-dimensional vascular region segmented using the region growing method;

[0095] Figure 6d A schematic diagram of the segmented two-dimensional true lumen region of the blood vessel;

[0096] Figure 7 A schematic diagram illustrating the process of exploring the center point on the center line of the main blood vessel according to one embodiment of the present invention;

[0097] Figure 8 A schematic diagram illustrating the principle of exploring the centerline of a vascular region, provided as a second specific example of the present invention;

[0098] Figure 9 A schematic diagram illustrating the exploration of the next new center point of the main branch vessel according to one embodiment of the present invention;

[0099] Figure 10 A schematic diagram of the centerline exploration process for a second branch vessel provided in one embodiment of the present invention;

[0100] Figure 11 A schematic diagram illustrating the centerline exploration principle of a second branch vessel according to an embodiment of the present invention;

[0101] Figure 12 A schematic diagram illustrating the initial direction for determining the centerline of a second branch vessel according to an embodiment of the present invention;

[0102] Figure 13 This is a block diagram of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0103] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, further illustrates the blood vessel image segmentation method, electronic device, and storage medium proposed in this invention. The advantages and features of this invention will become clearer from the following description. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, intended only to facilitate and clearly illustrate the embodiments of this invention. Please refer to the accompanying drawings to make the objectives, features, and advantages of this invention more apparent and understandable. It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are only for illustrative purposes and to enable those skilled in the art to understand and read them, and are not intended to limit the implementation conditions of this invention. Any modifications to the structure, changes in proportions, or adjustments to the size, provided they produce the same or similar effects and achieve the same objectives as this invention, should still fall within the scope of the technical content disclosed in this invention.

[0104] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0105] Furthermore, in the description of this specification, the reference to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., means that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0106] The core idea of ​​this invention is to provide a blood vessel image segmentation method, electronic device, and storage medium, which can effectively improve the accuracy and speed of blood vessel segmentation.

[0107] It should be noted that the vascular image segmentation method of this invention can be applied to the electronic device provided in this invention. This electronic device can be a personal computer, a mobile terminal, etc., and the mobile terminal can be a mobile phone, tablet computer, or other hardware device with various operating systems. Furthermore, it should be noted that, as those skilled in the art will understand, although this invention uses an aortic vascular image as an example of the vascular image to be segmented, this does not constitute a limitation of the invention. This invention can also be applied to the segmentation of vascular images such as the carotid artery, intracranial artery, and coronary artery.

[0108] It should also be noted that the blood vessel images to be segmented in this invention can be acquired by scanning with various modal imaging systems, or transmitted through internal or external storage systems such as Picture Archiving and Communication Systems (PACS). These modalities include, but are not limited to, one or more combinations of magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), computed tomography (CT), and positron emission tomography (PET).

[0109] To achieve the above-mentioned goals, this invention provides a method for segmenting blood vessel images. Please refer to [the relevant documentation]. Figure 1 The diagram illustrates a flowchart of a blood vessel image segmentation method according to an embodiment of the present invention. Figure 1 As shown, the blood vessel image segmentation method provided by the present invention includes the following steps:

[0110] Step S100: Use a preset algorithm to explore the center point on the center line of the blood vessel region in the acquired blood vessel image to be segmented, so as to obtain the position information of each center point and the cross-sectional data of the blood vessel cross section where the center point is located. The cross-sectional data includes the upper limit pixel value and the lower limit pixel value corresponding to the blood vessel cross section.

[0111] Step S200: For each center point, determine the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point.

[0112] Step S300: For each center point, take the center point as the first initial seed point, take the upper limit pixel value corresponding to the blood vessel cross section where the center point is located as the first upper limit threshold for region growth, and take the lower limit pixel value corresponding to the blood vessel cross section where the center point is located as the first lower limit threshold for region growth, and perform region growth within the limited range corresponding to the center point to obtain the blood vessel segmentation result within the limited range corresponding to the center point.

[0113] Step S400: Superimpose the blood vessel segmentation results corresponding to the defined range of each center point to obtain the blood vessel segmentation result corresponding to the blood vessel image to be segmented.

[0114] Therefore, this invention can transform the segmentation of three-dimensional blood vessels (e.g., the aorta) into segmentation of blood vessels within defined ranges guided by the vessel centerline, thereby achieving rapid and convenient blood vessel segmentation while significantly improving the accuracy of segmentation. Compared to existing deep learning algorithms, this invention greatly reduces memory usage during blood vessel segmentation and significantly increases segmentation speed. This invention can achieve rapid and convenient blood vessel segmentation over large vascular regions while improving the generalization of data processing. Furthermore, since the first upper limit threshold and the first lower limit threshold for region growing within the defined range corresponding to each center point are different, the segmentation accuracy of the blood vessel image segmentation method provided by this invention can be further effectively improved. It should be noted that, as those skilled in the art will understand, the blood vessel segmentation results within the defined ranges corresponding to each center point can be superimposed sequentially according to the arrangement order of the center points on the vessel centerline.

[0115] In one exemplary embodiment, for each center point, after completing the region growing within the defined range corresponding to that center point, the blood vessel image segmentation method further includes performing a morphological opening operation on the region growing result corresponding to that center point. Therefore, by performing a morphological opening operation on the region growing result, the adhesion between the blood vessel region and the background region can be reduced, further improving the accuracy of blood vessel segmentation.

[0116] In one exemplary embodiment, the vascular region includes a main branch vessel and two first branch vessels, the two first branch vessels converging at the terminal region of the main branch vessel.

[0117] The process of exploring the center point of the blood vessel centerline in the acquired blood vessel image to be segmented using a preset algorithm includes:

[0118] The first preset algorithm is used to explore the center point on the center line of the two first branch vessels simultaneously until the first termination condition is met. Then, the one of the termination center points of the two first branch vessels located at the higher position is taken as the first intersection point of the center line of the two first branch vessels and replaces the termination center point of the other first branch vessel.

[0119] The second preset algorithm is used to explore the center point on the center line of the main branch vessel until the second termination condition is met, then the first intersection point is taken as the termination center point of the main branch vessel.

[0120] Therefore, the preset algorithm provided by the present invention does not require segmenting the entire blood vessel region from the blood vessel image to be segmented first, thus eliminating the need for a heavy deep learning algorithm. This effectively avoids the problems of slow speed and high memory consumption caused by the prior art of segmenting a large area of ​​three-dimensional blood vessels first and then extracting the blood vessel centerline. At the same time, it can also avoid the problem of insufficient accuracy of the calculated blood vessel centerline due to the low quality of the three-dimensional blood vessel segmentation results.

[0121] It should be noted that, as those skilled in the art will understand, the two first branch vessels exhibit left-right anatomical characteristics. Specifically, when one of the first branch vessels is the left common iliac artery, left internal iliac artery, left external iliac artery, left femoral artery, left popliteal artery, left anterior tibial artery, or left posterior tibial artery, the other first branch vessel corresponds to the right common iliac artery, right internal iliac artery, right external iliac artery, right femoral artery, right popliteal artery, right anterior tibial artery, or right posterior tibial artery. Furthermore, when the image of the vessel to be segmented is a medical image of the aorta of a normal person or a patient suspected of having aortic dissection, the vascular region to be segmented consists at least of the thoracic aorta, abdominal aorta, and left and right iliac arteries; when the image of the vessel to be segmented is a medical image of the aorta of a patient suspected of having an abdominal aortic aneurysm, the vascular region to be segmented consists at least of the abdominal aorta and left and right iliac arteries.

[0122] In one exemplary embodiment, when the center point is located on the centerline of the first branch vessel, the cross-sectional data of the vessel cross-section where the center point is located also includes the maximum radius of the vessel cross-section. Determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes:

[0123] Using the location of the center point as the center of the first cube and twice the maximum radius of the blood vessel cross-section where the center point is located as the side length of the first cube, the defined range of the first cube corresponding to the center point is determined on the blood vessel image to be segmented.

[0124] Since the radius of the cross-section of the first branch vessel is relatively small, a first cube is defined on the image of the vessel to be segmented, with the center point (excluding the first intersection point) on the centerline of the first branch vessel as the center and twice the maximum radius of the cross-section of the vessel at the center point as the side length. The segmentation method described above is then used to segment the vessel region within the first cube. This can effectively improve the segmentation speed of the first branch vessel. At the same time, since the segmentation range of each first cube is relatively small, the superposition speed of the segmentation results of the vessel within each subsequent segmentation range of the first cube can be effectively improved.

[0125] Further, please refer to Figure 2 and Figure 3 ,in Figure 2 A schematic diagram of aortic segmentation with aortic dissection is provided in a specific example of the present invention. Figure 3 A schematic diagram illustrating aortic segmentation including an aortic aneurysm is provided in a specific example of the present invention. For example... Figure 2 and Figure 3 As shown, for each of the first branch vessels 120, the vessel region can be segmented sequentially along the direction from its initial center point to its termination center point, within the first cube-defined area around each center point on the centerline of the first branch vessel 120, to complete the segmentation of the first branch vessel 120. By superimposing the vessel segmentation results of each first cube-defined area, a complete segmentation result of the first branch vessel 120 can be obtained. It should be noted that, as those skilled in the art will understand, after obtaining the vessel segmentation result within the first cube-defined area corresponding to each center point, the vessel segmentation result corresponding to that center point can be mapped onto the image of the vessel to be segmented.

[0126] In one exemplary embodiment, when the center point is located on the center line of the main branch vessel 110, determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes:

[0127] Using the location of the center point as the bottom center of the first cuboid 111, the first preset length as the length of the first cuboid 111, the first preset width as the width of the first cuboid 111, and the distance between the center point and the adjacent center point located above the center point as the height of the first cuboid 111, the defined range of the first cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

[0128] Because the radius of the cross-section of the main branch vessel 110 is relatively large, a first cuboid is defined on the image of the vessel to be segmented. This first cuboid is used as the center point on the centerline of the main branch vessel 110, with the center point as the center of the bottom surface of the first cuboid 111. A first preset length (e.g., 128) is used as the length, a first preset width (e.g., 128) as the width, and the distance between the center point and the adjacent center point located above the center point (i.e., with a Z-coordinate higher than the center point) as the height. The segmentation method described above is then used to segment the vessel region within this first cuboid. This prevents significant overlap between the first cuboid segments corresponding to the center points on the centerline of the main branch vessel 110, thereby effectively improving the segmentation speed of the main branch vessel 110 and also facilitating the superposition of subsequent segmentation results from the various first cuboid segments. It should be noted that, as those skilled in the art will understand, the first preset length and the first preset width can be set according to actual conditions, and this invention does not limit them. Furthermore, it should be noted that, as those skilled in the art will understand, the first preset length and the first preset width can be the same value or different values.

[0129] Further, please refer to Figure 2 and Figure 3 ,like Figure 2 and Figure 3 As shown, for the main branch vessel 110, the vessel region can be segmented sequentially along the direction from its initial center point to its terminating center point within a first cuboid-defined area around each center point on the centerline of the main branch vessel 110 to complete the segmentation of the main branch vessel 110. By superimposing the vessel segmentation results of each first cuboid-defined area, a complete segmentation result of the main branch vessel 110 can be obtained. It should be noted that, as those skilled in the art will understand, after obtaining the vessel segmentation result of the first cuboid-defined area corresponding to each center point, the vessel segmentation result corresponding to that center point can be mapped onto the image of the vessel to be segmented.

[0130] In one exemplary embodiment, the blood vessel image to be segmented is an aortic blood vessel image, and the main branch 110 includes the aortic arch. When the center point is located at the highest position on the midline of the main branch 110, determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes:

[0131] Based on the center point, the bottom center of the second cuboid 112 is determined, the second preset length is used as the length of the second cuboid 112, the second preset width is used as the width of the second cuboid 112, and the first preset height is used as the height of the second cuboid 112. The defined range of the second cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

[0132] Since the common origin of aortic dissection and the trident structure, which is of great interest to doctors, are both located in the aortic arch, the segmentation of the main aortic branch 110, which includes aortic dissection, requires a larger segmentation near the highest center point of the aortic arch to isolate important structures such as the aortic dissection and the trident. Please continue to refer to [reference needed]. Figure 2 ,like Figure 2 As shown, the highest center point of the main branch of the aorta 110 can be shifted to the left by a preset amount to serve as the bottom center, and a second preset length, a second preset width, and a first preset height can be used to determine the range of the second cuboid on the image of the blood vessel to be segmented.

[0133] It should be noted that, as those skilled in the art will understand, the second preset length, the second preset width, and the first preset height can be set according to actual conditions. For example, the second preset length can be set to 192, the second preset width to 192, and the first preset width to 96. This invention does not limit this. Furthermore, it should be noted that, as those skilled in the art will understand, the second preset length and the second preset width can be the same value or different values.

[0134] In one exemplary embodiment, the blood vessel image to be segmented is an aortic blood vessel image. When the center point is the termination center point of the main branch 110, determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes:

[0135] Using the location of the center point as the center of the bottom surface of the third cuboid 113, the third preset length as the length of the third cuboid 113, the third preset width as the width of the third cuboid 113, and the third preset height as the height of the third cuboid 113, the defined range of the third cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

[0136] Please continue to refer to this. Figure 3 ,like Figure 3As shown, since the aortic aneurysm occurs at the first junction N0 of the first branch vessel 120 and the main branch vessel 110, the segmentation of the aortic main branch vessel 110 containing the aortic aneurysm requires a larger segmentation near the first junction N0 to completely segment the aortic aneurysm. It should be noted that, as those skilled in the art will understand, the third preset length, the third preset width, and the second preset height can be set according to actual conditions. For example, the third preset length can be set to 256, the third preset width to 256, and the second preset width to 96. This invention does not limit this. Furthermore, it should be noted that, as those skilled in the art will understand, the third preset length and the third preset width can be the same value or different values.

[0137] Furthermore, such as Figure 2 and Figure 3 As shown, for other center points on the centerline of the main aortic branch 110 (other center points besides the highest center point on the aortic arch and the first intersection point), the location of the center point can be used as the center of the bottom surface of the first cuboid 111, the first preset length can be used as the length of the first cuboid 111, the first preset width can be used as the width of the first cuboid 111, and the distance between the center point and the adjacent center point above the center point can be used as the height of the first cuboid 111. The defined range of the first cuboid corresponding to the center point can be determined on the image of the blood vessel to be segmented.

[0138] Please continue to refer to this. Figure 4 The diagram illustrates the process of exploring the center point on the center line of the first branch vessel 120 provided in one embodiment of the present invention. Figure 4 As shown, in one exemplary embodiment, the step of using a first preset algorithm to explore the center point on the centerline of the first branch vessel 120 until a first termination condition is met includes:

[0139] Step A1: Determine the initial center point on the center line of the first branch vessel 120 based on the first endpoint within the first branch vessel 120, and use the initial center point as the current forward point, and use the unit vector parallel to the Z-axis as the current forward direction vector.

[0140] Step A2: Based on the current forward point and the current forward direction vector, explore the next new center point on the center line of the first branch blood vessel 120, and take the explored new center point as the current center point of the first branch blood vessel 120;

[0141] Step A3: Determine whether the distance between the current center point of the first branch vessel 120 and the current center point of another first branch vessel 120 is less than a first threshold.

[0142] If yes, proceed to step A4; otherwise, proceed to step A5.

[0143] Step A4: Take the current center point of the first branch vessel 120 as the termination center point of the first branch vessel 120 and end the exploration of the center point on the center line of the first branch vessel 120.

[0144] Step A5: Take the current center point of the first branch blood vessel 120 as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step A2.

[0145] It should be noted that, as those skilled in the art will understand, the purpose of selecting the first endpoint is to initialize the exploration path of the centerline of the first branch vessel 120 and to constrain the exploration path. The first endpoint within each of the first branch vessels 120 can be selected manually or by a computer according to a pre-set algorithm; this invention does not impose any limitations on this. Please continue to refer to [the relevant documentation / reference]. Figure 5 The diagram illustrates the principle of centerline exploration of a vascular region according to a first specific example of the present invention. Figure 5 As shown, the first endpoints P2 and P3 are preferably located close to the end of the first branch vessel 120. The first endpoints P2 and P3 do not need to be close to the center of the first branch vessel 120, as long as the first endpoints P2 and P3 are located within the first branch vessel 120.

[0146] In one exemplary embodiment, a first preset algorithm is used simultaneously to alternately explore the center points on the center lines of the two first branch vessels 120, following the principle of prioritizing lower-order points. Thus, by alternately exploring the center lines of the left and right first branch vessels 120 using the principle of prioritizing lower-order points, it can be ensured that the exploration of the center lines of the left and right first branch vessels 120 is carried out synchronously. Specifically, the low-priority principle means that the exploration of the center point on the centerline of the first branch vessel 120 (represented as first branch vessel A for ease of distinction), which is located at the first endpoint farther from the main vessel, is carried out first. After a center point on the centerline of the first branch vessel A is found to be above the first endpoint of another first branch vessel 120 (represented as first branch vessel B), the exploration of the center point on the centerline of the first branch vessel B begins. After a center point on the centerline of the first branch vessel B is found, it is determined whether the distance between the two most recently found center points of the two first branch vessels 120 is less than the first threshold. If not, the exploration of the center point on the centerline of the first branch vessel A is carried out again. After a center point on the centerline of the first branch vessel A is found, the exploration of the center point on the centerline of the first branch vessel B is carried out again. After a center point on the centerline of the first branch vessel B is found, it is determined whether the distance between the two most recently found center points of the two first branch vessels 120 is less than the first threshold, and so on. By alternately exploring the center points on the centerlines of the two first branch vessels 120, the synchronous exploration of the centerlines of the two first branch vessels 120 can be achieved. It should be noted that, as those skilled in the art will understand, when the coordinates of each pixel in the image of the blood vessel to be segmented are represented by positional information in the LPS anatomical coordinate system (X-axis direction from right to left, Y-axis direction from front to back, Z-axis direction from bottom to top), the first endpoint with the smaller Z-coordinate is the first endpoint farther from the main blood vessel. Furthermore, it should be noted that, as those skilled in the art will understand, the one with the larger Z-coordinate among the termination center points of the two first branch blood vessels 120 is the one located at the higher position; that is, the one with the larger Z-coordinate among the termination center points of the two first branch blood vessels 120 is the first intersection point N0.

[0147] Specifically, assume the coordinates of the previous center point are c. j-1 The current center point coordinates are c j Then the current forward direction vector The calculation formula is as follows:

[0148]

[0149] In one exemplary implementation, the step of exploring the next new center point on the centerline of the first branch vessel 120 based on the current forward point and the current forward direction vector includes:

[0150] Based on the current forward point and the current forward direction vector, extract the current two-dimensional cross-section of interest image from the medical image to be segmented;

[0151] The current two-dimensional cross-sectional image of interest is segmented to obtain the current two-dimensional blood vessel region;

[0152] The current two-dimensional blood vessel region is segmented into true and false lumens to segment out the current two-dimensional blood vessel true lumen region;

[0153] The new center point of the first branch vessel 120 is obtained based on the center point of the current two-dimensional vascular lumen region.

[0154] Therefore, by reducing the dimensionality of the three-dimensional blood vessel image to be segmented into multiple consecutive two-dimensional cross-sectional images of interest, and exploring the center point on the blood vessel centerline based on the extracted two-dimensional cross-sectional images of interest, the extraction speed of the three-dimensional blood vessel centerline can be effectively improved, memory usage can be reduced, and a high-precision three-dimensional blood vessel centerline extraction function with strong data generalization can be achieved. Furthermore, since the center point on the blood vessel centerline in this invention is the center point of the corresponding two-dimensional true lumen region of the blood vessel, it can be ensured that the finally extracted blood vessel centerline is located within the true lumen of the blood vessel, thus meeting the extraction needs of aortic blood vessel centerlines for patients with diseases such as aortic dissection and aortic aneurysm. It should be noted that, as those skilled in the art will understand, after segmenting the current two-dimensional true lumen region of the blood vessel, the maximum radius of the current two-dimensional true lumen region of the blood vessel is taken as the maximum radius of the blood vessel cross section where its center point is located. The upper limit pixel value and lower limit pixel value corresponding to the blood vessel cross section where its center point is located are determined based on the gray average value of the pixels in the current two-dimensional true lumen region of the blood vessel. Specifically, assuming that the gray average value of the pixels in the current two-dimensional true lumen region of the blood vessel is h, the upper limit pixel value corresponding to the blood vessel cross section where its center point is located is h / λ1, and the lower limit pixel value is λ1h, where λ1 is a constant greater than 0 and less than 1.

[0155] Specifically, after segmenting the current two-dimensional vascular lumen region, the center point of this region can be calculated using the extreme erosion method. Therefore, by employing the extreme erosion method, the center point of the current two-dimensional vascular lumen region can be accurately calculated, ensuring that the calculated center point is located within the vascular lumen region. It should be noted that, as those skilled in the art will understand, further details regarding the calculation of the center point using the extreme erosion method can be found in existing technologies and will not be elaborated upon here.

[0156] In one exemplary implementation, the step of extracting the current two-dimensional section of interest image from the medical image to be segmented based on the current forward point and the current forward direction vector includes:

[0157] Based on the current forward point and the current forward direction vector, determine the center point of the current two-dimensional section of interest and the normal vector of the current two-dimensional section of interest, wherein the normal vector of the current two-dimensional section of interest is the current forward direction vector;

[0158] Based on the preset size, the center point of the current two-dimensional section of interest, and the normal vector of the current two-dimensional section of interest, the current two-dimensional section of interest image is extracted from the medical image to be segmented.

[0159] Therefore, by using the calculated center point of the current two-dimensional section of interest (ROI), the normal vector of the current two-dimensional ROI, and the preset size, the current two-dimensional ROI image can be accurately extracted from the blood vessel image to be segmented. It should be noted that, as those skilled in the art will understand, the preset size includes a width dimension and a height dimension; for ease of calculation, the width dimension is equal to the height dimension.

[0160] In one exemplary implementation, determining the center point of the current two-dimensional section of interest based on the current forward point and the current forward direction vector includes:

[0161] The current forward speed is determined based on the angle between the current forward direction vector (i.e., the current two-dimensional cross-section of interest normal vector) and the previous forward direction vector (i.e., the previous two-dimensional cross-section of interest normal vector).

[0162] The center point of the current two-dimensional cross-section of interest is determined based on the current forward point, the current forward direction vector, and the current forward velocity.

[0163] Specifically, assume the current forward direction vector is The previous forward direction vector is Then the current forward direction vector and the previous forward direction vector The formula for calculating the included angle θ between them is as follows:

[0164]

[0165] After calculating the current forward direction vector and the previous forward direction vector After determining the included angle θ, the current forward velocity v can be calculated using any one of the following formulas (3) to (6):

[0166] v=k*α*θ+b,θ∈[0,π] (3)

[0167] v=k*α*(cosθ+1)+b,θ∈[0,π] (4)

[0168]

[0169]

[0170] Where k, b, and β are constants, and α is the ratio of the area of ​​the previous two-dimensional true lumen region to the area of ​​the two-dimensional true lumen region where the starting center point is located.

[0171] It should be noted that, as those skilled in the art can understand, k and b in formulas (3) to (6) are pre-set, specifically calculated based on a large number of vascular centerlines (e.g., aortic centerlines), and β is an amplification factor used to control the maximum and minimum values ​​of the curve.

[0172] Therefore, by calculating the forward direction vector of the current forward point (i.e., the current forward direction vector) according to formula (2) and in combination with any one of formulas (3) to (6), the present invention can achieve adaptive adjustment of the forward speed of the exploration of the blood vessel centerline. That is, when the deflection angle is small or the blood vessel is thick, a larger forward speed is used to improve the exploration efficiency, and when the deflection angle is large or the blood vessel is thin, a smaller speed is used to improve the accuracy of the exploration of the blood vessel centerline.

[0173] After calculating the current forward velocity, the center point sc of the current two-dimensional section of interest can be calculated using the following formula. j :

[0174]

[0175] Among them, c j-1 As the current starting point, This is the current forward direction vector.

[0176] In one exemplary embodiment, the method further includes:

[0177] The maximum radius of the current two-dimensional vascular lumen region is determined based on the center point of the current two-dimensional vascular lumen region.

[0178] Based on the maximum radius of the current two-dimensional vascular lumen region, determine the locking radius corresponding to the current two-dimensional vascular lumen region;

[0179] The step of segmenting the blood vessel region in the current two-dimensional cross-sectional image of interest to segment the current two-dimensional blood vessel region includes:

[0180] Based on the average gray value of the pixels in the two-dimensional blood vessel lumen region, determine the second upper limit threshold and the second lower limit threshold for region growth.

[0181] Based on the locking radius corresponding to the previous two-dimensional true lumen region of the blood vessel and the center point of the current two-dimensional section of interest, the current circular region is determined in the current two-dimensional section of interest image;

[0182] In the current circular domain, a pixel with a value greater than the lower threshold and less than the upper threshold is selected as the second initial seed point for region growth;

[0183] Based on the second initial seed point, the second upper threshold, and the second lower threshold, the region growing method is used to segment the blood vessel region of the current two-dimensional cross-sectional image of interest, so as to segment the current two-dimensional blood vessel region.

[0184] In one exemplary embodiment, the method further includes:

[0185] The maximum radius of the current two-dimensional vascular lumen region is determined based on the center point of the current two-dimensional vascular lumen region.

[0186] The locking radius corresponding to the current two-dimensional blood vessel lumen region is determined based on the maximum radius of the current two-dimensional blood vessel lumen region.

[0187] Since the maximum radius of the explored artery suddenly decreases, such as when a dissection occurs or a sharp turn is encountered, continuing to explore the center point is prone to errors. Therefore, by setting a lock radius, the range of the second initial seed point used for region growth can be limited, thereby ensuring that the results obtained from region growth accurately include blood vessels.

[0188] Specifically, assuming the average grayscale value of the pixels in the two-dimensional true lumen region of the blood vessel is h j-1 Then the upper limit threshold is h j-1 / λ2, the lower threshold is λ2h j-1 , where λ² is a constant greater than 0 and less than 1. Please refer to [reference needed]. Figures 6a to 6d ,in, Figure 6a A schematic diagram of a two-dimensional cross-sectional image of interest is given; Figure 6b A schematic diagram illustrating the determination of the circular domain based on the lock radius (Ir in the diagram) is given. j-1 That is, the locking radius corresponding to the true lumen region of the previous two-dimensional blood vessel; Figure 6c A schematic diagram of a two-dimensional vascular region segmented using the region growing method is shown (the white elliptical area in the figure represents the two-dimensional vascular region). Figure 6dA schematic diagram of the segmented two-dimensional true lumen region of a blood vessel is shown (the white area in the diagram represents the two-dimensional true lumen region of the blood vessel). Figures 6a to 6d As shown, since there is a certain continuity between blood vessels in adjacent sections, this invention adds a circular domain and a dual threshold constraint including an upper threshold and a lower threshold to the basic region growing method. Using the position and pixel values ​​of the blood vessels in the previous section as prior information, interference from irrelevant tissues can be reduced, improving the accuracy of blood vessel region growing. It should be noted that, as those skilled in the art will understand, the relevant techniques for segmenting the blood vessel region of the current two-dimensional cross-section image of interest using the region growing method based on the second initial seed point, the second upper threshold, and the second lower threshold can be found in existing region growing methods, and therefore will not be elaborated upon here.

[0189] In one exemplary embodiment, determining the locking radius corresponding to the current two-dimensional vascular lumen region based on the maximum radius of the current two-dimensional vascular lumen region includes:

[0190] Determine whether the ratio of the square of the maximum radius of the previous two-dimensional vascular lumen region to the square of the maximum radius of the current two-dimensional vascular lumen region is greater than a preset ratio.

[0191] If so, the maximum radius of the previous two-dimensional true lumen region is used as the locking radius corresponding to the current two-dimensional true lumen region, so as to lock the locking radius of the current two-dimensional true lumen region.

[0192] If not, then the maximum radius of the current two-dimensional vascular lumen region is taken as the locking radius corresponding to the previous two-dimensional vascular lumen region.

[0193] Specifically, assuming the maximum radius of the true lumen region of the two-dimensional blood vessel is mr j-1 The maximum radius of the current two-dimensional vascular lumen region is mr. j The preset ratio is γ, γ≥1, if (mr j-1 ) 2 / (mr j If )2>γ, then the locking radius Ir of the current two-dimensional vascular true lumen region is... j For mr j-1 ;(mr j-1 ) 2 / (mr j If 2 ≤ γ, then the locking radius Ir of the current two-dimensional vascular true lumen region is... j For mr j It should be noted that, as those skilled in the art will understand, the specific value of the preset ratio γ can be set according to the actual situation, and the present invention does not limit it in this regard.

[0194] In one exemplary embodiment, if the locking radius of the previous two-dimensional vascular lumen region is locked, the method further includes, before segmenting the current two-dimensional vascular region into true and false lumens:

[0195] The current two-dimensional blood vessel region is eroded to remove interfering areas within it.

[0196] Specifically, if the locking radius Ir of the upper two-dimensional vascular true lumen region j-1 The maximum radius mr of the two-dimensional true lumen region of the blood vessel is above. j-2 If the vascular centerline deviates significantly, it indicates that the exploration has entered the interlayer or has deviated significantly. Therefore, by performing minor erosion on the current two-dimensional vascular region, the adhesion between the blood vessel and other tissues can be removed, thereby effectively removing the interference area in the current two-dimensional vascular region and laying a good foundation for obtaining a high-precision vascular centerline.

[0197] In one exemplary embodiment, the segmentation of the current two-dimensional vascular region into true and false lumens to segment out the current two-dimensional vascular true lumen region includes:

[0198] Perform connected component analysis on the current two-dimensional blood vessel region, and extract the connected component closest to the center point of the current two-dimensional cross-section of interest as the current two-dimensional blood vessel true lumen region.

[0199] Since the current two-dimensional true lumen vessel region and the previous two-dimensional true lumen vessel region are continuous under the premise that the cross-sectional direction of the previous two-dimensional true lumen vessel region is correct, the current two-dimensional true lumen vessel region should be close to the center point of the current two-dimensional cross-section of interest. Therefore, the connected region closest to the center point of the current two-dimensional cross-section of interest should be the current two-dimensional true lumen vessel region.

[0200] In one exemplary embodiment, after extracting a current two-dimensional section of interest image from the blood vessel image to be segmented, and before segmenting the blood vessel region in the current two-dimensional section of interest image, the method further includes:

[0201] Map the current two-dimensional cross-sectional image of interest onto the two-dimensional plane with Z=0.

[0202] Correspondingly, the segmentation of the blood vessel region in the current two-dimensional cross-sectional image of interest to segment the current two-dimensional blood vessel region includes:

[0203] The current two-dimensional cross-sectional image of interest, mapped to a two-dimensional plane of Z=0, is segmented to extract the current two-dimensional blood vessel region.

[0204] Therefore, by mapping the current two-dimensional cross-sectional image of interest to the plane with Z=0, it can be ensured that the Z coordinate of each pixel in the mapped current two-dimensional cross-sectional image of interest is 0, which can effectively simplify the subsequent segmentation steps.

[0205] Specifically, suppose the coordinate matrix formed by the homogeneous coordinates of a pixel in the current two-dimensional cross-sectional image of interest in the three-dimensional coordinate system corresponding to the blood vessel image to be segmented is an m×4 matrix X. 3d , where X 3d Each row in the matrix corresponds to the homogeneous coordinates of a pixel in the current 2D cross-sectional image of interest in the 3D coordinate system corresponding to the blood vessel image to be segmented; the coordinate matrix formed by the homogeneous coordinates (where the Z coordinate is 0) of all pixels in the current 2D cross-sectional image of interest mapped to the Z=0 2D plane in the 2D cross-sectional coordinate system (the 2D coordinate system with the center of the current 2D cross-sectional image of interest mapped to the Z=0 2D plane as the origin) of the 2D cross-sectional coordinate system is an m×4 matrix X. 2d , where X 2d Each row in the table corresponds to the homogeneous coordinates of a pixel in the current two-dimensional cross-sectional image of interest in the two-dimensional cross-sectional coordinate system. Then X... 3d and X 2d The following relationship exists between them:

[0206] X 3d =X 2d R y R x R T (8)

[0207] Among them, R x Let R be the rotation matrix along the X-axis. y Let R be the rotation matrix along the Y-axis. T Translation matrix.

[0208] Furthermore, suppose that the center point sc of the two-dimensional section of interest corresponding to the current two-dimensional section of interest image is... j The homogeneous coordinates in the three-dimensional coordinate system are: The normal vector of the current two-dimensional section of interest corresponding to the current two-dimensional section of interest image. (i.e., the current forward direction vector) The homogeneous coordinates in a three-dimensional coordinate system are represented as follows: Then R x R y and R T The calculation formula is as follows:

[0209]

[0210]

[0211]

[0212]

[0213]

[0214]

[0215] sinβ=-x nj (15)

[0216] It should be noted that, as those skilled in the art will understand, offset is a pre-set parameter, and the specific value of offset can be 1e-6. Therefore, by setting the offset, it is possible to prevent [the problem caused by] y nj and z nj When the value is zero, it causes the problem that cosα and sinα cannot be calculated.

[0217] Therefore, after calculating the rotation matrix R x R y and the translation matrix R T Then, according to the above formula (8), the current two-dimensional cross-section image of interest can be extracted from the blood vessel image to be segmented and transformed from the three-dimensional coordinate system to the two-dimensional cross-section coordinate system. After calculating the coordinates of the center point in the two-dimensional cross-section coordinate system, the coordinates of the center point in the three-dimensional coordinate system can be obtained by formula (8).

[0218] In one exemplary embodiment, if the determination result of step A3 is negative, then before executing step A5, the method further includes:

[0219] Determine whether the exploration direction of the current center point of the first branch vessel 120 has backtracked;

[0220] If so, stop exploring the center point on the center line of the first branch vessel 120 and send a message indicating that the center line exploration has failed.

[0221] Specifically, when the exploration direction of the blood vessel centerline reverses, it indicates that the exploration direction has deviated backward, meaning an error has occurred. Therefore, by determining whether the exploration direction of the current center point has reversed, we can promptly avoid continuing in the wrong direction.

[0222] Furthermore, the following steps can be used to determine whether the exploration direction of the current center point of the first branch vessel 120 has backtracked:

[0223] Determine whether the distance between the current center point and any of the most recently explored center points (of the first preset number) is less than a first preset distance threshold. If yes, backtracking has occurred; otherwise, no backtracking has occurred.

[0224] Since the exploration direction of the vessel centerline has deviated backward when the distance between the current center point and any of the most recently explored center points is less than the first preset distance threshold, it can be accurately determined whether the exploration direction of the current center point of the first branch vessel 120 has deviated backward by judging whether the distance between the current center point and any of the most recently explored center points is less than the first preset distance threshold.

[0225] In one exemplary embodiment, if the exploration direction of the current center point of the first branch vessel 120 has not been backtracked, then before executing step A5, the method further includes:

[0226] Determine whether the selected position of the first endpoint of the first branch vessel 120 meets the first preset condition;

[0227] If not, the search for the center point on the center line of the first branch vessel 120 is stopped, and a prompt message is sent to reselect the first endpoint of the first branch vessel 120.

[0228] Specifically, when the first endpoint selected by the user is not within the first branch vessel 120, or when the vessel at the location of the first endpoint selected by the user is relatively thin, that is, when the position of the selected first endpoint does not meet the first preset condition, the centerline exploration will be incorrect. Therefore, by judging whether the position of the selected first endpoint meets the first preset condition, the continued incorrect exploration can be avoided in time.

[0229] Furthermore, the following steps can be used to determine whether the selected position of the first endpoint of the first branch vessel 120 meets the first preset condition:

[0230] Determine whether the average maximum radius of the second preset number of recently extracted two-dimensional vascular lumen regions (including the current two-dimensional vascular lumen region) is less than a first preset radius threshold. If yes, determine that the selection position of the first endpoint does not meet the first preset condition; otherwise, determine that the selection position of the first endpoint meets the first preset condition.

[0231] Since the average maximum radius of the second preset number of two-dimensional vascular lumen regions (including the current two-dimensional vascular lumen region) is less than the first preset radius threshold, it indicates that the first endpoint is not located within the first branch vessel 120. Therefore, by determining whether the average maximum radius of the second preset number of two-dimensional vascular lumen regions (including the current two-dimensional vascular lumen region) is less than the first preset radius threshold, it is possible to accurately determine whether the selection position of the first endpoint of the first branch vessel 120 meets the first preset condition.

[0232] In one exemplary embodiment, if the distance between the current center point of the first branch vessel 120 and the current center point of another first branch vessel 120 is less than the first threshold, then before performing step A4, the method further includes:

[0233] If the number of explored center points is greater than a first preset value, the exploration of center points on the center line of the first branch vessel 120 is terminated; otherwise, a prompt message indicating that the center line exploration has failed is sent.

[0234] Specifically, when the selected position of the first endpoint does not meet the preset conditions, the exploration of the centerline of the first branch vessel 120 may stop too early. Therefore, by further determining whether the number of explored center points is greater than the first preset value when the distance between the current center point of the first branch vessel 120 and the current center point of another first branch vessel 120 is less than the first threshold, the problem of the exploration of the centerline of the first branch vessel 120 stopping too early due to the selected position of the first endpoint not meeting the preset conditions can be effectively avoided.

[0235] In one exemplary embodiment, determining the initial center point on the centerline of the first branch vessel 120 based on the acquired first endpoint within the first branch vessel 120 includes:

[0236] Using the first endpoint as the center point of the initial two-dimensional section of interest, and the unit vector parallel to the Z-axis as the normal vector of the initial two-dimensional section of interest, the initial two-dimensional section of interest image is extracted from the medical image to be segmented.

[0237] The initial two-dimensional cross-sectional image of interest is segmented into a vascular region to extract the initial two-dimensional vascular region.

[0238] The initial two-dimensional vascular region is segmented into true and false lumens to separate the initial two-dimensional vascular true lumen region;

[0239] The initial center point of the first branch vessel 120 is obtained based on the center point of the initial two-dimensional vascular lumen region.

[0240] Specifically, such as Figure 5 As shown, the initial two-dimensional cross-section normal vector of the first branch vessel 120 is... The coordinates are represented as (0,0,1). For details on how to segment the vascular region in the initial two-dimensional cross-sectional image of interest and how to segment the true and false lumens of the initial two-dimensional vascular region, please refer to the relevant descriptions above. It should be noted that, as those skilled in the art will understand, the upper limit threshold for region growing the initial two-dimensional vascular true lumen region is h0 / λ, and the lower limit threshold is λh0, where h0 is the pixel value of the first endpoint. Further, when the side length of the initial two-dimensional cross-sectional image of interest is r, the radius of the initial circular domain of the second initial seed point used to define the region growing of the initial two-dimensional vascular true lumen region is r / 2.

[0241] Please continue to refer to this. Figure 7 The diagram illustrates the process of exploring the center point on the central line of the main blood vessel according to one embodiment of the present invention. Figure 7 As shown, in one exemplary embodiment, the step of using a second preset algorithm to explore the center point on the centerline of the main branch vessel 110 until a second termination condition is met includes:

[0242] Step B1: Determine the initial center point on the center line of the main branch vessel 110 based on the second endpoint obtained within the main branch vessel 110, and use the initial center point as the current forward point, and use the unit vector parallel to the Z-axis as the current forward direction vector.

[0243] Step B2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the main branch vessel 110, and take the explored new center point as the current center point of the main branch vessel 110.

[0244] Step B3: Determine whether the distance between the current center point of the main branch vessel 110 and the first intersection point is less than the second threshold.

[0245] If yes, proceed to step B4; otherwise, proceed to step B5.

[0246] Step B4: Replace the current center point with the first intersection point as the termination center point of the main branch vessel 110 and end the exploration of the center point on the center line of the main branch vessel 110.

[0247] Step B5: Take the current center point of the main branch vessel 110 as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step B2.

[0248] It should be noted that, as those skilled in the art will understand, the purpose of selecting the second endpoint is to initialize and constrain the exploration path of the centerline of the main branch vessel 110. The second endpoint can be selected manually or by a computer according to a pre-set algorithm; this invention does not impose any limitations on this. Furthermore, the location of the second endpoint P1 is preferably close to the main branch vessel 110 (see...). Figure 5 The second endpoint P1 does not need to be close to the center of the main branch vessel 110, as long as it is located within the main branch vessel 110. It should be noted that, as those skilled in the art will understand, the second endpoint P1 can be selected manually or by a computer according to a pre-set algorithm, and the present invention does not impose any limitations on this. Furthermore, when the image of the vessel to be segmented is a medical image including the aorta, the selectable region of the second endpoint P1 includes the ascending aorta, aortic arch, descending aorta, and abdominal aorta of the thoracic aorta segment.

[0249] Please continue to refer to 5 and... Figure 8 ,in Figure 8 A schematic diagram illustrating the principle of centerline exploration of a vascular region provided in a second specific example of the present invention is given. For example... Figure 5 As shown, when the second endpoint P1 is located within a blood vessel segment extending towards the head (e.g., within the ascending aorta), the unit vector parallel to the Z-axis in step B1 should be a positive unit vector parallel to the Z-axis, meaning the coordinates of the unit vector parallel to the Z-axis in A1 are (0,0,1). Figure 8 As shown, when the second endpoint P1 is located in a blood vessel segment extending towards the foot (e.g., in the descending aorta), the unit vector parallel to the Z-axis in step B1 should be a negative unit vector parallel to the Z-axis, that is, the coordinates of the unit vector parallel to the Z-axis in A1 are represented as (0,0,-1).

[0250] In one exemplary embodiment, the step of exploring the next new center point on the centerline of the main branch vessel 110 based on the current forward point and the current forward direction vector, and using the explored new center point as the current center point of the main branch vessel 110, includes:

[0251] Based on the current forward point and the current forward direction vector, extract the current two-dimensional cross-section image of interest from the blood vessel image to be segmented;

[0252] The current two-dimensional cross-sectional image of interest is segmented to obtain the current two-dimensional blood vessel region;

[0253] The current two-dimensional blood vessel region is segmented into true and false lumens to segment out the current two-dimensional blood vessel true lumen region;

[0254] The new center point of the main branch vessel 110 is obtained based on the center point of the current two-dimensional vascular lumen region.

[0255] Please continue to refer to this. Figure 9 This schematically illustrates a diagram of exploring the next new center point of the main branch vessel 110 according to an embodiment of the present invention. Figure 9 As shown, after exploring the center point c j-1 Then, center point c j-1 As the current point of progress, it will be from the center point c. j-2 Pointing to the center point c j-1 The unit vector is used as the current forward vector. Then, the center point sc of the current two-dimensional cross-section of interest is determined according to formula (7) above. j Then, based on the preset size, the current two-dimensional cross-sectional image of interest can be extracted from the blood vessel image to be segmented. It should be noted that, as those skilled in the art will understand, further details regarding how to extract the current two-dimensional cross-sectional image of interest from the blood vessel image to be segmented based on the current forward point and the current forward direction vector, how to segment the blood vessel region from the current two-dimensional cross-sectional image of interest to segment the current two-dimensional blood vessel region, and how to segment the true and false lumens from the current two-dimensional blood vessel region to segment the true lumen region of the current two-dimensional blood vessel, can be found in the specific content regarding the exploration of the center point of the first segment 120 mentioned above, and will not be repeated here.

[0256] In one exemplary embodiment, if the determination result of step B3 is negative, then before executing step B5, the method further includes:

[0257] Determine whether the exploration direction of the current center point of the main branch vessel 110 has backtracked;

[0258] If so, the search for the center point on the center line of the main branch vessel 110 will be stopped, and a message indicating that the center line search has failed will be sent.

[0259] Furthermore, the following steps can be used to determine whether the exploration direction of the current center point of the main branch vessel 110 has backtracked:

[0260] Determine whether the distance between the current center point and any of the most recently explored center points (of the first preset number) is less than a second preset distance threshold. If yes, backtracking has occurred; otherwise, no backtracking has occurred.

[0261] In one exemplary embodiment, if the exploration direction of the current center point of the main branch vessel 110 has not been backtracked, then before executing step B5, the method further includes:

[0262] Determine whether the selected position of the second endpoint of the main branch vessel 110 meets the second preset condition;

[0263] If not, the search for the center point on the center line of the main branch vessel 110 is stopped, and a prompt message is sent to reselect the second endpoint of the main branch vessel 110.

[0264] Furthermore, the selection position of the second endpoint of the main branch vessel 110 can be determined through the following steps to see if it meets the second preset condition:

[0265] Determine whether the average maximum radius of the second preset number of recently extracted two-dimensional vascular lumen regions (including the current two-dimensional vascular lumen region) is less than the second preset radius threshold. If yes, determine that the selection position of the second endpoint does not meet the second preset condition; otherwise, determine that the selection position of the second endpoint meets the second preset condition.

[0266] In one exemplary embodiment, if the distance between the current center point of the main branch vessel 110 and the first junction point is less than the second threshold, then before performing step B4, the method further includes:

[0267] If the number of explored center points is greater than a second preset value, the exploration of center points on the center line of the main branch vessel 110 is terminated; otherwise, a prompt message indicating failure of center line exploration is sent.

[0268] In one exemplary embodiment, determining the initial center point on the centerline of the main branch vessel 110 based on the second endpoint of the main branch vessel 110 includes:

[0269] Using the second endpoint as the center point of the initial two-dimensional section of interest, and the unit vector parallel to the Z-axis as the normal vector of the initial two-dimensional section of interest, the initial two-dimensional section of interest image is extracted from the blood vessel image to be segmented.

[0270] The initial two-dimensional cross-sectional image of interest is segmented into a vascular region to extract the initial two-dimensional vascular region;

[0271] The initial two-dimensional vascular region is segmented into true and false lumens to separate the initial two-dimensional vascular true lumen region;

[0272] The initial center point of the main branch vessel 110 is obtained based on the center point of the initial two-dimensional true lumen region of the vessel.

[0273] Specifically, such as Figure 2 and Figure 8 As shown, when the second endpoint P1 is located within a segment of the blood vessel extending towards the head (e.g., within the ascending aorta), the initial two-dimensional cross-sectional normal vector of the main branch vessel 110 is... The coordinates are represented as (0,0,1); when the second endpoint P1 is located within a segment of the blood vessel extending towards the foot (e.g., within the descending aorta), the initial two-dimensional cross-section normal vector of the main branch vessel 110 is... The coordinates are represented as (0,0,-1).

[0274] In one exemplary embodiment, the vascular region further includes at least one second branch vessel 130 intersecting with the main branch vessel 110, and the step of exploring the center point of the vascular centerline of the vascular region in the acquired vascular image to be segmented using a preset algorithm further includes:

[0275] The third preset algorithm is used to explore the center point on the center line of the second branch vessel 130 until the third termination condition is met. Then, the center point where the distance between the main vessel and the termination center point of the second branch vessel 130 is less than the third threshold is determined as the second intersection point corresponding to the second branch vessel 130 and replaces the termination center point of the second branch vessel 130.

[0276] Furthermore, for each center point on the center line of the second branch vessel 130, the location of the center point is taken as the center of the second cube, and the side length of the second cube is taken as twice the maximum radius of the vessel cross section where the center point is located. The defined range of the second cube corresponding to the center point is determined on the image of the vessel to be segmented.

[0277] Please continue to refer to this. Figure 10 The diagram illustrates the centerline exploration process of the second branch vessel 130 provided in one embodiment of the present invention. Figure 10 As shown, in one exemplary embodiment, the step of using a third preset algorithm to explore the center point on the centerline of the second branch vessel 130 until a third termination condition is met includes:

[0278] Step C1: Determine the initial center point on the center line of the second branch vessel 130 based on the obtained third endpoint within the second branch vessel 130, and use the initial center point as the current forward point, and use the unit vector pointing from the third endpoint to the fourth endpoint corresponding to the second branch vessel 130 as the current forward direction vector, wherein the fourth endpoint is the center point on the center line of the main branch vessel 110 that is closest to the third endpoint of the second branch vessel 130.

[0279] Step C2: Based on the current forward point and the current forward direction vector, explore the next new center point on the center line of the second branch blood vessel 130, and take the explored new center point as the current center point of the second branch blood vessel 130;

[0280] Step C3: Determine whether the minimum distance between the current center point of the second branch vessel 130 and each center point of the main branch vessel 110 is less than a third threshold.

[0281] If yes, proceed to step C4; otherwise, proceed to step C5.

[0282] Step C4: Take the current center point as the termination center point of the second branch vessel 130 and end the exploration of the center point on the center line of the second branch vessel;

[0283] Step C5: Take the current center point of the second branch blood vessel 130 as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step C2.

[0284] For details, please refer to Figure 11 and Figure 12 ,in Figure 11 A schematic diagram illustrating the centerline exploration principle of the second branch vessel 130 provided in one embodiment of the present invention is shown. Figure 12 A schematic diagram illustrating the initial direction for determining the centerline of the second branch vessel 130 according to an embodiment of the present invention is shown. Figure 11 As shown in the figure, points P4, P5, P6, P7, P8, P9, and P... 10 This refers to the third endpoint corresponding to each of the second branch vessels 130. It should be noted that, as those skilled in the art will understand, the third endpoint of each second branch vessel 130 can be selected manually or by a computer according to a pre-set algorithm; this invention does not impose any limitations on this. Furthermore, it should be noted that, as those skilled in the art will understand, for each second branch vessel 130, the purpose of selecting the third endpoint is to initialize and constrain the exploration path of the centerline of that second branch vessel 130. The selected position of the third endpoint is preferably close to the end of the second branch vessel 130; the third endpoint does not need to be close to the center of the second branch vessel 130, as long as it is located within the second branch vessel 130. Figure 12 As shown in the figure, point P0 is the fourth endpoint corresponding to the second branch vessel 130 where point P6 is located. The initial forward direction vector of the second branch vessel 130 where point P6 is located is the unit vector pointing from point P6 to point P0.

[0285] It should be noted that, as those skilled in the art will understand, for each second branch vessel 130, by determining the second junction point corresponding to that second branch vessel 130, the center point of that second branch vessel 130 and the center line of the main branch vessel 110 can be connected at the second junction point corresponding to that second branch vessel 130, thereby allowing for the rapid and accurate acquisition of the center line of the entire vascular region. Please continue to refer to... Figure 11 ,like Figure 11 As shown, the second junction point corresponding to the second branch vessel 130 where point P4 is located is N1; the second junction point corresponding to the second branch vessel 130 where point P5 is located is N2; the second junction point corresponding to the second branch vessel 130 where point P6 is located is N3; the second junction point corresponding to the second branch vessel 130 where point P7 is located is N4; the second junction point corresponding to the second branch vessel 130 where point P8 is located is N5; the second junction point corresponding to the second branch vessel 130 where point P9 is located is N6; and point P... 10 The second junction point corresponding to the second branch vessel 130 is N7.

[0286] In one exemplary embodiment, the step of exploring the next new center point on the centerline of the second branch vessel 130 based on the current forward point and the current forward direction vector, and using the explored new center point as the current center point of the second branch vessel 130, includes:

[0287] Based on the current forward point and the current forward direction vector, extract the current two-dimensional cross-section image of interest from the medical image to be extracted;

[0288] The current two-dimensional cross-sectional image of interest is segmented to obtain the current two-dimensional blood vessel region;

[0289] The current two-dimensional blood vessel region is segmented into true and false lumens to segment out the current two-dimensional blood vessel true lumen region;

[0290] The new center point of the second branch vessel 130 is obtained based on the center point of the current two-dimensional vascular lumen region.

[0291] Specifically, for more details on how to extract the current two-dimensional cross-section image of interest from the medical image to be extracted based on the current forward point and the current forward direction vector, how to segment the vascular region of the current two-dimensional cross-section image of interest to segment the current two-dimensional vascular region, and how to segment the true and false lumens of the current two-dimensional vascular region to segment the true lumen region of the current two-dimensional vascular region, please refer to the specific content on the exploration of the center point of the first branch vessel 120 above, which will not be repeated here.

[0292] In one exemplary embodiment, if the determination result of step C3 is negative, then before executing step C5, the method further includes:

[0293] Determine whether the exploration direction of the current center point of the second branch vessel 130 has backtracked;

[0294] If so, then stop exploring the center point on the center line of the second branch vessel 130.

[0295] Furthermore, the following steps can be used to determine whether the exploration direction of the current center point of the second branch vessel 130 has backtracked:

[0296] Determine whether the distance between the current center point and any of the most recently explored center points (of the first preset number) is less than a preset distance threshold. If yes, backtracking has occurred; otherwise, no backtracking has occurred.

[0297] In one exemplary embodiment, if the exploration direction of the current center point of the second branch vessel 130 has not been traced back, then before performing step C5, the method further includes:

[0298] Determine whether the selected position of the third endpoint of the second branch vessel 130 meets the third preset condition;

[0299] If not, the search for the center point on the center line of the second branch vessel 130 is stopped, and a prompt message is sent to reselect the third endpoint of the second branch vessel 130.

[0300] Furthermore, the selection position of the third endpoint of the second branch vessel 130 can be determined through the following steps to see if it meets the third preset condition:

[0301] Determine whether the average maximum radius of the second preset number of recently extracted two-dimensional vascular lumen regions (including the current two-dimensional vascular lumen region) is less than a third preset radius threshold. If yes, determine that the selection position of the third endpoint does not meet the third preset condition; otherwise, determine that the selection position of the third endpoint meets the third preset condition.

[0302] In one exemplary embodiment, if the distance between the current center point of the second branch vessel 130 and the first junction point is less than the third threshold, then before performing step C4, the method further includes:

[0303] If the number of explored center points is greater than a third preset value, the exploration of center points on the center line of the second branch vessel 130 is terminated; otherwise, a prompt message indicating that the center line exploration has failed is sent.

[0304] In one exemplary embodiment, determining the initial center point on the centerline of the second branch vessel 130 based on the third endpoint of the second branch vessel 130 includes:

[0305] Using the third endpoint as the center point of the initial two-dimensional section of interest, and the unit vector pointing from the third endpoint to the fourth endpoint corresponding to the second branch blood vessel 130 as the normal vector of the initial two-dimensional section of interest, the initial two-dimensional section of interest image is extracted from the medical image to be extracted.

[0306] The initial two-dimensional cross-sectional image of interest is segmented into a vascular region to extract the initial two-dimensional vascular region;

[0307] The initial two-dimensional vascular region is segmented into true and false lumens to separate the initial two-dimensional vascular true lumen region;

[0308] The initial center point of the second branch vessel 130 is obtained based on the center point of the initial two-dimensional vascular lumen region.

[0309] Based on the same inventive concept, the present invention also provides an electronic device, please refer to... Figure 13 A block diagram illustrating an embodiment of the electronic device provided by the present invention is shown. Figure 13 As shown, the electronic device includes a processor 210 and a memory 230. The memory 230 stores a computer program, which, when executed by the processor 210, implements the blood vessel image segmentation method described above. Since the electronic device provided by this invention and the blood vessel image segmentation method provided by this invention belong to the same inventive concept, the electronic device provided by this invention possesses all the advantages of the blood vessel image segmentation method provided by this invention. Therefore, the beneficial effects of the electronic device provided by this invention will not be elaborated upon here.

[0310] like Figure 13 As shown, the electronic device also includes a communication interface 220 and a communication bus 240, wherein the processor 210, the communication interface 220, and the memory 230 communicate with each other via the communication bus 240. The communication bus 240 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 240 can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used to represent it in the figure, but this does not indicate that there is only one bus or one type of bus. The communication interface 220 is used for communication between the aforementioned electronic device and other devices.

[0311] The processor 210 referred to in this invention can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 210 is the control center of the electronic device, connecting various parts of the electronic device through various interfaces and lines.

[0312] The memory 230 can be used to store the computer program. The processor 210 implements various functions of the electronic device by running or executing the computer program stored in the memory 230 and calling the data stored in the memory 230.

[0313] The memory 230 may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0314] This invention also provides a readable storage medium storing a computer program, which, when executed by a processor, can implement the blood vessel image segmentation method described above. Since the readable storage medium and the blood vessel image segmentation method provided by this invention belong to the same inventive concept, the readable storage medium provided by this invention possesses all the advantages of the blood vessel image segmentation method provided by this invention. Therefore, the beneficial effects of the readable storage medium provided by this invention will not be elaborated upon here.

[0315] The readable storage medium of embodiments of the present invention can be any combination of one or more computer-readable media. The readable medium can be a computer-readable signal medium or a computer-readable storage medium. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections having one or more wires, portable computer hard disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device.

[0316] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of transmitting, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber, RF, etc., or any suitable combination thereof.

[0317] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0318] In summary, compared with the prior art, the blood vessel image segmentation method, electronic device, and storage medium provided by the present invention have the following advantages:

[0319] The vascular image segmentation method provided by this invention first explores the center points on the center lines of the vascular regions in the acquired vascular image to be segmented using a preset algorithm to obtain the position information of each center point and the cross-sectional data of the vascular cross section where the center point is located. The cross-sectional data includes the upper limit pixel value and the lower limit pixel value corresponding to the vascular cross section. Then, for each center point, a limited range corresponding to the center point is determined on the vascular image to be segmented based on the position information of the center point. Then, for each center point, region growth is performed within the limited range corresponding to the center point, using the center point as a first initial seed point, the upper limit pixel value corresponding to the vascular cross section where the center point is located as a first upper limit threshold for region growth, and the lower limit pixel value corresponding to the vascular cross section where the center point is located as a first lower limit threshold for region growth, to obtain the vascular segmentation result within the limited range corresponding to the center point. Finally, the vascular segmentation results within the limited ranges corresponding to each center point are superimposed to obtain the vascular segmentation result corresponding to the vascular image to be segmented. Therefore, this invention can transform three-dimensional blood vessel (e.g., aorta) segmentation into blood vessel segmentation within defined ranges guided by the blood vessel centerline, thereby achieving rapid and convenient blood vessel segmentation while significantly improving the accuracy of segmentation. Compared to existing deep learning algorithms, this invention greatly reduces memory usage during blood vessel segmentation, significantly increasing segmentation speed. This invention can achieve rapid and convenient blood vessel segmentation over large areas, while also improving the generalization of data processing. Furthermore, since the first upper and lower thresholds for region growing within the defined range corresponding to each center point are different, the segmentation accuracy of the blood vessel image segmentation method provided by this invention can be further effectively improved.

[0320] Since the electronic device and storage medium provided by this invention belong to the same inventive concept as the blood vessel image segmentation method provided by this invention, the electronic device and storage medium provided by this invention have all the advantages of the blood vessel image segmentation method provided by this invention. Therefore, the beneficial effects of the electronic device and storage medium provided by this invention will not be described in detail here.

[0321] It should be noted that the apparatus and methods disclosed in the embodiments herein can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments herein. In this regard, each block in a flowchart or block diagram may represent a module, program, or part of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system to perform the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0322] In addition, the functional modules in the various embodiments of this article can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0323] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure are within the protection scope of the present invention. Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A method for segmenting blood vessel images, characterized in that, include: A preset algorithm is used to explore the center point of the blood vessel centerline in the blood vessel region of the acquired blood vessel image to be segmented, so as to obtain the position information of each center point and the cross-sectional data of the blood vessel cross section where the center point is located. The cross-sectional data includes the upper limit pixel value and the lower limit pixel value corresponding to the blood vessel cross section. For each center point, a defined range corresponding to the center point is determined on the blood vessel image to be segmented based on the position information of the center point. For each center point, the center point is used as the first initial seed point, the upper limit pixel value corresponding to the blood vessel cross section where the center point is located is used as the first upper limit threshold for region growth, and the lower limit pixel value corresponding to the blood vessel cross section where the center point is located is used as the first lower limit threshold for region growth. Region growth is performed within the limited range corresponding to the center point to obtain the blood vessel segmentation result within the limited range corresponding to the center point. The blood vessel segmentation results corresponding to the defined range of each center point are superimposed to obtain the blood vessel segmentation result corresponding to the blood vessel image to be segmented. The vascular region includes one main branch vessel and two first branch vessels, with the two first branch vessels converging at the terminal region of the main branch vessel. The process of exploring the center point of the blood vessel centerline in the acquired blood vessel image to be segmented using a preset algorithm includes: The first preset algorithm is used to explore the center point on the center line of the two first branch vessels simultaneously until the first termination condition is met. Then, the one of the termination center points of the two first branch vessels located at the higher position is taken as the first intersection point of the center line of the two first branch vessels and replaces the termination center point of the other first branch vessel. The second preset algorithm is used to explore the center point on the center line of the main branch vessel until the second termination condition is met, then the first intersection point is taken as the termination center point of the main branch vessel.

2. The blood vessel image segmentation method according to claim 1, characterized in that, When the center point is located on the centerline of the main branch vessel, determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes: Using the location of the center point as the bottom center of the first cuboid, the first preset length as the length of the first cuboid, the first preset width as the width of the first cuboid, and the distance between the center point and the adjacent center point located above the center point as the height of the first cuboid, the defined range of the first cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

3. The blood vessel image segmentation method according to claim 1, characterized in that, The blood vessel image to be segmented is an aortic blood vessel image, and the main branch blood vessel includes the aortic arch. When the center point is located at the highest position on the centerline of the main branch blood vessel, the step of determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes: Based on the center point, the center of the bottom surface of the second cuboid is determined, the second preset length is used as the length of the second cuboid, the second preset width is used as the width of the second cuboid, and the first preset height is used as the height of the second cuboid. The defined range of the second cuboid corresponding to the center point is then determined on the blood vessel image to be segmented.

4. The blood vessel image segmentation method according to claim 1, characterized in that, The blood vessel image to be segmented is an aortic blood vessel image. When the center point is the termination center point of the main branch blood vessel, the step of determining the defined range corresponding to the center point on the blood vessel image to be segmented based on the position information of the center point includes: Using the location of the center point as the center of the bottom surface of the third cuboid, the third preset length as the length of the third cuboid, the third preset width as the width of the third cuboid, and the third preset height as the height of the third cuboid, the defined range of the third cuboid corresponding to the center point is determined on the blood vessel image to be segmented.

5. The blood vessel image segmentation method according to claim 1, characterized in that, The vascular region also includes at least one second branch vessel intersecting with the main branch vessel. The step of exploring the center point of the vascular centerline in the acquired vascular image to be segmented using a preset algorithm further includes: The third preset algorithm is used to explore the center point on the center line of the second branch vessel until the third termination condition is met. Then, the center point where the distance between the termination center point of the main branch vessel and the termination center point of the second branch vessel is less than the third threshold is determined as the second intersection point corresponding to the second branch vessel and replaces the termination center point of the second branch vessel.

6. The blood vessel image segmentation method according to claim 5, characterized in that, When the center point is located on the centerline of the first branch vessel or the second branch vessel, the cross-sectional data of the vessel cross-section where the center point is located also includes the maximum radius of the vessel cross-section. Determining the defined range corresponding to the center point on the vessel image to be segmented based on the position information of the center point includes: Using the location of the center point as the center of the cube and twice the maximum radius of the blood vessel cross-section where the center point is located as the side length of the cube, the defined range of the cube corresponding to the center point is determined on the blood vessel image to be segmented.

7. The blood vessel image segmentation method according to claim 1, characterized in that, The step of simultaneously exploring the center point on the centerline of the two first branch vessels using a first preset algorithm until a first termination condition is met includes: Step A1: Determine the initial center point on the center line of the first branch blood vessel based on the first endpoint obtained in the first branch blood vessel, and use the initial center point as the current forward point, and use the unit vector parallel to the Z axis as the current forward direction vector. Step A2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the first branch blood vessel, and take the explored new center point as the current center point of the first branch blood vessel; Step A3: Determine whether the distance between the current center point of the first branch vessel and the current center point of another first branch vessel is less than a first threshold. If yes, proceed to step A4; otherwise, proceed to step A5. Step A4: Take the current center point of the first branch vessel as the termination center point of the first branch vessel and end the exploration of the center point on the center line of the first branch vessel. Step A5: Take the current center point of the first branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step A2.

8. The blood vessel image segmentation method according to claim 7, characterized in that, The step of exploring the next new center point on the centerline of the first branch vessel based on the current forward point and the current forward direction vector includes: Based on the current forward point and the current forward direction vector, extract the current two-dimensional cross-section image of interest from the blood vessel image to be segmented; The current two-dimensional cross-sectional image of interest is segmented to obtain the current two-dimensional blood vessel region; The current two-dimensional blood vessel region is segmented into true and false lumens to segment out the current two-dimensional blood vessel true lumen region; The new center point of the first branch vessel is obtained based on the center point of the current two-dimensional vascular lumen region.

9. The blood vessel image segmentation method according to claim 8, characterized in that, The step of extracting the current two-dimensional cross-sectional image of interest from the blood vessel image to be segmented based on the current forward point and the current forward direction vector includes: Based on the current forward point and the current forward direction vector, determine the center point of the current two-dimensional section of interest and the normal vector of the current two-dimensional section of interest, wherein the normal vector of the current two-dimensional section of interest is the current forward direction vector; Based on the preset size, the center point of the current two-dimensional section of interest, and the normal vector of the current two-dimensional section of interest, the current two-dimensional section of interest image is extracted from the blood vessel image to be segmented.

10. The blood vessel image segmentation method according to claim 9, characterized in that, Determining the center point of the current two-dimensional section of interest based on the current forward point and the current forward direction vector includes: The current forward speed is determined based on the angle between the current forward direction vector and the previous forward direction vector; The center point of the current two-dimensional cross-section of interest is determined based on the current forward point, the current forward direction vector, and the current forward velocity.

11. The blood vessel image segmentation method according to claim 9, characterized in that, The method further includes: The maximum radius of the current two-dimensional vascular lumen region is determined based on the center point of the current two-dimensional vascular lumen region. Based on the maximum radius of the current two-dimensional vascular lumen region, determine the locking radius corresponding to the current two-dimensional vascular lumen region; The step of segmenting the blood vessel region in the current two-dimensional cross-sectional image of interest to segment the current two-dimensional blood vessel region includes: Based on the average gray value of the pixels in the two-dimensional blood vessel lumen region, determine the second upper limit threshold and the second lower limit threshold for region growth. Based on the locking radius corresponding to the previous two-dimensional true lumen region of the blood vessel and the center point of the current two-dimensional section of interest, the current circular region is determined in the current two-dimensional section of interest image; In the current circular domain, a pixel with a value greater than the lower threshold and less than the upper threshold is selected as the second initial seed point for region growth; Based on the second initial seed point, the second upper threshold, and the second lower threshold, the region growing method is used to segment the blood vessel region of the current two-dimensional cross-sectional image of interest, so as to segment the current two-dimensional blood vessel region.

12. The blood vessel image segmentation method according to claim 11, characterized in that, The step of determining the locking radius corresponding to the current two-dimensional vascular lumen region based on the maximum radius of the current two-dimensional vascular lumen region includes: Determine whether the ratio of the square of the maximum radius of the previous two-dimensional vascular lumen region to the square of the maximum radius of the current two-dimensional vascular lumen region is greater than a preset ratio. If so, the maximum radius of the previous two-dimensional true lumen region is used as the locking radius corresponding to the current two-dimensional true lumen region, so as to lock the locking radius of the current two-dimensional true lumen region. If not, then the maximum radius of the current two-dimensional vascular lumen region is taken as the locking radius corresponding to the previous two-dimensional vascular lumen region.

13. The blood vessel image segmentation method according to claim 9, characterized in that, The step of segmenting the current two-dimensional blood vessel region into true and false lumens to segment out the current two-dimensional blood vessel true lumen region includes: Perform connected component analysis on the current two-dimensional blood vessel region, and extract the connected component closest to the center point of the current two-dimensional cross-section of interest as the current two-dimensional blood vessel true lumen region.

14. The blood vessel image segmentation method according to claim 8, characterized in that, Based on the average grayscale value of the pixels within the two-dimensional true lumen blood vessel region where the center point is located, the upper and lower pixel values ​​corresponding to the blood vessel cross section where the center point is located are determined.

15. The blood vessel image segmentation method according to claim 7, characterized in that, The step of determining the initial center point on the centerline of the first branch vessel based on the obtained first endpoint within the first branch vessel includes: Using the first endpoint as the center point of the initial two-dimensional section of interest, and the unit vector parallel to the Z-axis as the normal vector of the initial two-dimensional section of interest, the initial two-dimensional section of interest image is extracted from the blood vessel image to be segmented. The initial two-dimensional cross-sectional image of interest is segmented into a vascular region to extract the initial two-dimensional vascular region; The initial two-dimensional vascular region is segmented into true and false lumens to separate the initial two-dimensional vascular true lumen region; The initial center point of the first branch vessel is obtained based on the center point of the initial two-dimensional vascular lumen region.

16. The blood vessel image segmentation method according to claim 1, characterized in that, The process of using a second preset algorithm to explore the center point on the centerline of the main branch vessel until a second termination condition is met includes: Step B1: Determine the initial center point on the center line of the main branch vessel based on the obtained second endpoint within the main branch vessel, and use the initial center point as the current forward point, and use the unit vector parallel to the Z-axis as the current forward direction vector. Step B2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the main branch vessel, and take the explored new center point as the current center point of the main branch vessel; Step B3: Determine whether the distance between the current center point of the main branch vessel and the first intersection point is less than the second threshold. If yes, proceed to step B4; otherwise, proceed to step B5. Step B4: Replace the current center point with the first intersection point as the termination center point of the main branch vessel and end the exploration of the center point on the center line of the main branch vessel. Step B5: Take the current center point of the main branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step B2.

17. The blood vessel image segmentation method according to claim 5, characterized in that, The process of using a third preset algorithm to explore the center point on the centerline of the second branch vessel until a third termination condition is met includes: Step C1: Determine the initial center point on the center line of the second branch vessel based on the obtained third endpoint within the second branch vessel, and use the initial center point as the current forward point, and use the unit vector pointing from the third endpoint to the fourth endpoint corresponding to the second branch vessel as the current forward direction vector, wherein the fourth endpoint is the center point on the center line of the main branch vessel that is closest to the third endpoint of the second branch vessel. Step C2: Based on the current forward point and the current forward direction vector, explore the next new center point on the centerline of the second branch blood vessel, and take the explored new center point as the current center point of the second branch blood vessel; Step C3: Determine whether the minimum distance between the current center point of the second branch vessel and each center point of the main branch vessel is less than the third threshold. If yes, proceed to step C4; otherwise, proceed to step C5. Step C4: Take the current center point as the termination center point of the second branch vessel and end the exploration of the center point on the center line of the second branch vessel; Step C5: Take the current center point of the second branch blood vessel as the current forward point, take the unit vector pointing from the previous center point to the current center point as the current forward direction vector, and return to execute step C2.

18. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, which, when executed by the processor, implements the blood vessel image segmentation method according to any one of claims 1 to 17.

19. A readable storage medium, characterized in that, The readable storage medium stores a computer program, which, when executed by a processor, implements the blood vessel image segmentation method according to any one of claims 1 to 17.

Images