Coronary vessel deformation calculation method and device in cardiac cycle, medium and equipment

Abstract

The invention provides a coronary vessel deformation calculation method and device in a cardiac cycle, a medium and equipment, and the method comprises the steps: obtaining image data, selecting a frame of image as a key frame from the image data, and enabling the image data to comprise coronary vessel image data of the whole cardiac cycle; positioning a blood vessel section used for deformation analysis in the key frame as a reference analysis section; selecting a plurality of analysis frames from the image data, and positioning a blood vessel section corresponding to the reference analysis section in each analysis frame as a comparison analysis section; and carrying out contour segmentation and registration on the reference analysis section and each comparison analysis section so as to calculate deformation of each position in the blood vessel section to be analyzed. Deformation of each position in a to-be-analyzed blood vessel section is calculated, so that the problem that blood vessel deformation preprocessing is complex is solved, and the simplicity and accuracy of coronary artery diameter change detection are improved.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and more specifically to a method for calculating coronary vessel deformation during a cardiac cycle. Background Art

[0002] Coronary artery disease is a common cardiovascular disease, the main feature of which is coronary artery stenosis or obstruction, leading to myocardial ischemia and hypoxia, and even myocardial infarction in severe cases. The occurrence and development of coronary heart disease is closely related to the stability of plaques in the coronary arteries.

[0003] At present, the degree of stenosis is usually assessed clinically by measuring the diameter of the coronary artery, and the stability of the plaque at the corresponding position is assessed by observing the diameter change within the same cardiac cycle, thereby judging the patient's condition. However, the traditional measurement method is usually manual measurement by doctors, which is not only time-consuming and labor-intensive, but also highly subjective and prone to errors. With the development of computer technology, image processing technology is increasingly used in the medical field. Among them, coronary angiography is a commonly used examination method that can clearly display the morphology and structure of the coronary artery, providing an important basis for doctors' diagnosis.

[0004] Prior art CN 114782358 A discloses a method, device and storage medium for automatic calculation of vascular deformation, including obtaining vascular data; selecting key frame image data in different parts of the cardiac cycle from the image data according to the cardiac cycle; registering the selected key frame image data to obtain the registered key frame image data; calculating relative deformation data based on the registered key frame image data; calculating absolute deformation data based on the registered key frame image data; and obtaining the deformation result of the blood vessel based on the relative deformation data and / or the absolute deformation data. This patent can monitor the lumen deformation size of the same position of the blood vessel at different times in real time, and can know the stability of the plaque. However, this patent requires complex image preprocessing and segmentation, which is not only time-consuming and labor-intensive, but also may introduce additional errors. Secondly, these systems can often only measure the coronary artery diameter at a certain moment, and cannot reflect the changes in the entire cardiac cycle. Summary of the invention

[0005] The present invention provides a method for calculating coronary artery deformation during a cardiac cycle, aiming to solve the complex problem of vascular deformation preprocessing and improve the simplicity and accuracy of detecting coronary artery diameter changes.

[0006] To achieve the above object, the present invention provides a method for calculating coronary artery deformation during a cardiac cycle, comprising:

[0007] Acquire image data, and select a frame of image as a key frame, wherein the image data is coronary artery image data including the entire cardiac cycle;

[0008] Positioning a blood vessel segment for deformation analysis in the key frame as a reference analysis segment;

[0009] Selecting a plurality of analysis frames from the image data, and locating a blood vessel segment corresponding to the reference analysis segment in each of the analysis frames as a comparison analysis segment;

[0010] The reference analysis segment and each comparison analysis segment are subjected to contour segmentation and registration to calculate the deformation of each position in the blood vessel segment to be analyzed.

[0011] In one embodiment, the key frame selection conditions include at least one or more of: the cardiac cycle is in the end diastole, the lesion is clearly exposed, and there is little blood vessel overlap.

[0012] In one embodiment, the analysis frame selection condition includes: the plurality of analysis frames and the key frame are in the same complete cardiac cycle, and the plurality of analysis frames respectively cover at least the mid-diastole, early-systole and end-systole of a complete cardiac cycle.

[0013] In one embodiment, the registration includes registering corresponding positions of the reference analysis segment and the comparison analysis segment to obtain registered blood vessel parameters, where the blood vessel parameters include a maximum diameter and a minimum diameter of the blood vessel.

[0014] In one embodiment, in step S104, the deformation calculation includes radial tube wall deformation calculation, and the radial tube wall deformation calculation formula is:

[0015] Max.RWS=maximum diameter-minimum diameter

[0016] Max.RWS%=(maximum diameter-minimum diameter) / maximum diameter

[0017] When Max.RWS% is greater than a preset threshold, unstable plaque exists in the blood vessel segment.

[0018] In one embodiment, when the contour segmentation result is not good, the segmentation result is manually modified; and / or, when the registration result is not good, forced registration points are added to the key frame and the analysis frame.

[0019] In one embodiment, when the corresponding points in the analysis frame and the key frame are not in good position, registration point marks are added to the corresponding point positions of the analysis frame and the key frame so that the positions corresponding to the registration point marks on the analysis frame and the key frame are forced to be matched.

[0020] A device for calculating coronary artery deformation during a cardiac cycle comprises a memory, a processor and a computer program stored in the memory. When the processor executes the computer program, the method for calculating coronary artery deformation during a cardiac cycle is implemented.

[0021] A computer-readable medium stores a computer program, and when the computer program is executed by a processor, the method for calculating coronary vessel deformation during a cardiac cycle is implemented.

[0022] A device for calculating coronary artery deformation during a cardiac cycle comprises an input device, an output device, a memory, a processor and a computer program. When the computer program is executed by the processor, the method for calculating coronary artery deformation during a cardiac cycle is implemented.

[0023] The present invention has the following beneficial effects:

[0024] 1. Simplicity: This technology directly matches key frames and analysis frames, performs contour segmentation and registration, and does not require complex image preprocessing and segmentation. In addition, this technical solution realizes the automatic matching of key frames and analysis frames, automatic contour segmentation and registration process, which significantly reduces the burden on staff and greatly improves work efficiency and convenience.

[0025] 2. Accuracy: This scheme analyzes the coronary artery diameter throughout the entire cardiac cycle, reflecting the changes in blood vessel diameter throughout the cardiac cycle, and further improving the accuracy of coronary artery diameter measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flowchart of a method for calculating coronary artery deformation during a cardiac cycle according to an embodiment of the present invention;

[0027] Figure 2 A schematic diagram of key frame selection in a method for calculating coronary artery deformation during a cardiac cycle according to an embodiment of the present invention;

[0028] Figure 3 A schematic diagram of deformation analysis of a key frame and an analysis frame according to an embodiment of the present invention;

[0029] Figure 4 A schematic diagram of the deformation difference at each position of the reference analysis section and the comparison analysis section according to an embodiment of the present invention;

[0030] Figure 5 A schematic diagram of adding forced registration points in a reference analysis segment and a comparison analysis segment according to an embodiment of the present invention;

[0031] Figure 6 This is a structural block diagram of a device for calculating coronary artery deformation during a cardiac cycle according to an embodiment of the present invention;

[0032] Figure 7 The structure block diagram of a device for calculating coronary artery deformation during a cardiac cycle according to an embodiment of the present invention.

[0033] Among them, 100 is a processor; 200 is a memory; 300 is an input device; and 400 is an output device. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the implementation of this application clearer, the technical scheme in the embodiment of this application will be described in more detail below in conjunction with the drawings in the embodiment of this application. In the drawings, the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of this application, not all of them. The embodiments described below with reference to the drawings are exemplary and are intended to be used to explain this application, and should not be construed as limitations on this application. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0035] Figure 1 The figure is a flow chart of a method for calculating coronary artery deformation during a cardiac cycle according to an embodiment of the present invention. The method for calculating coronary artery deformation during a cardiac cycle includes:

[0036] Acquire image data, and select a frame of image as a key frame, wherein the image data is coronary artery image data including the entire cardiac cycle;

[0037] Positioning a blood vessel segment for deformation analysis in the key frame as a reference analysis segment;

[0038] Selecting a plurality of analysis frames from the image data, and locating a blood vessel segment corresponding to the reference analysis segment in each of the analysis frames as a comparison analysis segment;

[0039] The reference analysis segment and each comparison analysis segment are subjected to contour segmentation and registration to calculate the deformation of each position in the blood vessel segment to be analyzed.

[0040] Specifically, acquiring image data of all stages of the entire cardiac cycle specifically includes:

[0041] When the image data includes electrocardiogram data, by extracting cardiac cycle data, the blood vessel images of all stages of the entire cardiac cycle are directly obtained;

[0042] When the image data does not include ECG data, a neural network trained through deep learning is used to obtain vascular images of all stages of the entire cardiac cycle.

[0043] Furthermore, the vascular image data within a certain cardiac cycle is directly selected as the key frame, and the blood vessels with clear exposure of the lesions to be analyzed and less blood vessel overlap are determined as the reference analysis segment, and then the blood vessels corresponding to other stages in the same cardiac cycle are automatically selected according to the algorithm. This method does not require preprocessing and segmentation of the key frames, and does not require segmentation of the vascular images of the vascular stage of a certain cardiac cycle. It only needs to automatically select the same blood vessel in the same cardiac cycle, which greatly reduces the amount of data that needs to be processed.

[0044] Further, when automatically selecting the analysis frame, illustratively, a target detection algorithm based on deep learning, such as the MobileNetV2 network, can be used to extract features from the input image. YOLO, SSD, Faster R-CNN, etc. can also be used to confirm the comparison analysis segment within the analysis frame. The target detection algorithm based on deep learning plays an important role in medical image processing. It can automatically identify and locate specific targets in the image, such as blood vessels, organs, tumors, etc. In the diagnosis of cardiovascular diseases, the target detection algorithm can quickly and accurately identify the vascular structure, thereby assisting in subsequent vascular tracking, registration and analysis. However, the present invention is not limited to this. The present application can also obtain a series of continuous cardiac images for dynamic imaging technologies such as echocardiography, MRI movie imaging, etc. through image sequence analysis. By analyzing these image sequences, different stages of the cardiac cycle can be identified, and vascular images of the same stage can be selected. Vessel tracking and matching can also be used. In the images of multiple cardiac cycles, the trajectory of the same blood vessel is identified using a vascular tracking algorithm, and the vascular image of the corresponding cardiac cycle stage is found through a matching algorithm.

[0045] Furthermore, when automatically performing contour segmentation, contour segmentation can be performed in a variety of ways. The following are some common contour segmentation methods:

[0046] Threshold-based segmentation algorithm: This method divides the pixels in the image into several regions according to their grayscale values, usually using single threshold, multiple thresholds, and adaptive thresholds. Global thresholding, such as Otsu's method, is a typical example. It automatically selects a threshold so that the sum of the intra-class variances of the two classes (foreground and background) is minimized.

[0047] Edge-based segmentation algorithm: segmentation is performed by detecting edges or contours in the image. Common algorithms include Canny algorithm, Sobel algorithm, Prewitt operator, etc. These methods use edge detection operators, such as first-order or second-order derivative operators, to highlight the edge information in the image and perform segmentation based on this information.

[0048] Region-based segmentation algorithm: divide the pixels in the image into several regions and perform segmentation based on the similarity between regions, such as grayscale, color and other features. Commonly used algorithms include K-means algorithm, watershed algorithm, etc.

[0049] Energy-based segmentation algorithm: Image segmentation is performed by defining an energy function. Common algorithms include GrabCut algorithm, GraphCut algorithm, etc. This type of algorithm usually transforms the image segmentation problem into a problem of solving the minimum value of the energy function, and finds the optimal segmentation boundary through an optimization algorithm.

[0050] Segmentation algorithm based on contour model: Use continuous curves to express the edge of the target, and define an energy functional whose independent variable includes the edge curve. The segmentation process is transformed into a process of solving the minimum value of the energy functional, which can generally be achieved by solving the Euler equation corresponding to the function. The most representative is the active contour model (such as the Snake model), which can automatically adjust the contour shape to match the target boundary according to the image features.

[0051] Segmentation algorithm based on deep learning: For example, deep learning models such as convolutional neural networks (CNN) are used to automatically learn image features and perform segmentation. By using the proposed edge and connected domain enhancement modules and adding them to the neural network model layer by layer, the vascular edge information of different feature scales is fused, so that the segmentation result can better retain the edge information of the vascular contour and the connectivity between the contour areas, which is conducive to achieving accurate and complete vascular contour segmentation results, and helps to improve the accuracy and time efficiency of assessing the degree of vascular stenosis.

[0052] Preferably, this embodiment uses a deep learning-based segmentation algorithm to perform contour segmentation on the reference analysis segment and the comparison analysis segment.

[0053] Furthermore, the registration includes: obtaining the contour graph of the image, registering the blood vessels at the same position at different times to obtain the deformation difference of the blood vessels. After the registration, the image has multiple blood vessels, and the parameters of the blood vessels can be obtained, including the maximum diameter and the minimum diameter of the blood vessels.

[0054] Specifically, feature-based registration methods include:

[0055] Feature point matching: 1) SIFT (Scale Invariant Feature Transform): Use DOG (Differential Gaussian) to simplify LOG (Laplacian of Gaussian), improve the search speed of feature points, and the registration accuracy is very high. 2) SURF (Speeded Up Robust Features): Drawing on the idea of ​​SIFT, the Gaussian second-order differential template in DOG is approximately simplified, so that the template only needs a few simple addition and subtraction operations to complete the image filtering. The amount of calculation is independent of the size of the filter template. It also has the invariance to image rotation, translation and scale changes, and is faster than the SIFT algorithm without reducing performance. 3) Contour feature matching: Extract contour feature points in the image, such as edge points, and match them. Directly use the results of contour segmentation to find corresponding points by comparing the shape, size, direction and other features of the contour, thereby achieving image registration.

[0056] Registration method based on transform domain: Fourier-Mellin transform is one of the commonly used methods, which transforms the image into the frequency domain, performs log-polar transform on the image amplitude spectrum, and solves the relative rotation angle and translation offset in the log-polar coordinate space. This method can handle the rotation and translation transformation of the image, but may not work well when dealing with scale changes.

[0057] Hybrid method based on contour detection and feature matching: This method combines the advantages of contour detection and feature matching. First, the target contour in the image is extracted through contour detection, and then the corresponding points or regions are found using the feature matching algorithm. Finally, the image registration is achieved through geometric transformations such as affine transformation and transmission transformation. For example, the image can be contour segmented first, and then the contour feature points are extracted. Then, the corresponding relationship between the reference image and the image to be registered is found through the feature matching algorithm. Finally, the image to be registered is transformed to the same coordinate system as the reference image through affine transformation or transmission transformation to achieve image registration.

[0058] Registration methods based on deep learning: With the development of deep learning technology, image registration methods based on deep learning are also gradually emerging. For example, using a U-net-like network structure, using stride convolution instead of pooling layer for downsampling, retaining more spatial position features required for registration. Deep learning models can learn complex correspondences between images through a large amount of training data, thereby achieving high-precision image registration.

[0059] Preferably, this embodiment adopts a deep learning-based registration method, using deep learning models such as convolutional neural networks to automatically learn image features and perform registration.

[0060] The principle of selecting key frames is to select image data with high image clarity and high vascular boundary clarity. In one embodiment, the key frame selection conditions include at least one or more of: the cardiac cycle is in the end diastole, the lesion is clearly exposed, and the blood vessels overlap less. Blood vessels in the end diastole are generally more dilated and less shortened. The exposure of vascular lesions in the end diastole is generally clearer, so in one embodiment, it is recommended to select the end diastole image as the key frame. The image data is image data including vascular morphology, and preferably, DSA angiography image data is selected.

[0061] Furthermore, DSA technology can generate high-resolution vascular images, clearly showing the course, distribution, stenosis, occlusion, aneurysm, vascular malformation and other pathological conditions of the blood vessels. Such high-resolution images provide doctors with precise anatomical and pathological information, which helps to accurately diagnose vascular diseases. Moreover, DSA removes unnecessary background images such as soft tissue and bones through digital subtraction technology, retaining only vascular images. This makes the contrast of vascular images higher and easier to observe and analyze. However, the present invention is not limited to this, and the present application can also use other image data containing cardiac cycle data, such as echocardiography, CT, and MRI, which are the main image types that contain cardiac cycle and can show vascular morphology.

[0062] In one embodiment, the analysis frame selection condition includes: the plurality of analysis frames and the key frame are in the same complete cardiac cycle, and the plurality of analysis frames respectively cover at least the mid-diastole, early-systole and end-systole of a complete cardiac cycle.

[0063] A complete cardiac cycle includes multiple stages such as early diastole, mid-diastole, end-diastole, early systole and end-systole. In one embodiment, considering the amount of calculation and calculation efficiency, four frames of images in four stages are selected as the basis for calculation. Controlling the number of frames involved in the calculation can effectively improve the operating efficiency and reduce the requirements for software and hardware equipment, so that the present invention has a wider applicability. In one embodiment, an image in the end of diastole is selected as a key frame, and three frames of images in the mid-diastole, early systole and end-systole are selected as analysis frames, a total of four frames of images, and the images of mid-diastole, end-diastole, early systole and end-systole can better reflect the changes in vascular morphology in the entire cardiac cycle. The present invention mainly studies the diameter changes in the cardiac cycle. In a complete cardiac cycle, the period with the smallest blood vessel diameter is generally at the end of systole, and the period with the largest blood vessel diameter is generally at the end of diastole. Selecting mid-diastole and early systole between the two can already meet the diameter changes in the complete cardiac cycle, so these four phases will be recommended by default during specific operations. In other embodiments, according to actual usage requirements, when implementing in the software, the user can also select images of other phases for analysis.

[0064] Specifically, the method for determining the cycle phase of the cardiac cycle specifically includes: dividing a complete cardiac cycle of the heart into four equal or unequal length phases based on a predetermined time period, and the diastole and the systole are each subdivided into two sub-phases. As an example, this division method is not absolutely fixed, and can also be flexibly adjusted to divide the cardiac cycle into three main parts, but no matter how it is divided, it must be ensured that at least one complete diastole and one complete systole are included. Because the heart muscles relax during diastole, allowing blood to flow back to the heart, this process is usually twice as long as the duration of the heart muscle contraction during systole to push blood out of the heart. Therefore, when the cardiac cycle is divided into multiple parts, the diastole will naturally be divided into two or more sub-phases to reflect its longer duration.

[0065] In one embodiment, the registration includes registering corresponding positions of the reference analysis segment and the comparison analysis segment to obtain registered blood vessel parameters, where the blood vessel parameters include a maximum diameter and a minimum diameter of the blood vessel.

[0066] In one embodiment, the reference analysis segment and the comparison analysis segment may also be registered by contour registration, and the registration may include:

[0067] 1) Feature extraction: Extract feature points or feature descriptors from the segmented vascular contours. These features should be able to reflect the morphological and structural information of the blood vessels.

[0068] 2) Matching algorithm: Using matching algorithms, such as feature-based matching, grayscale-based matching, etc., the blood vessel contours in the key frame and the analysis frame are matched to find the corresponding relationship between them.

[0069] 3) Transformation model: Based on the matching results, a transformation model is established, such as affine transformation, rigid transformation, non-rigid transformation, etc., to align the blood vessel contours in different images.

[0070] 4) Optimization and adjustment: The transformation model is adjusted through the optimization algorithm to minimize the registration error and improve the registration accuracy.

[0071] Furthermore, the vascular parameters include any one or more of the vascular contour, lumen diameter, value of lumen diameter changing over time, location of vascular lumen diameter change, vascular wall thickness, and hemodynamic parameters.

[0072] In one embodiment, in step S104, the deformation calculation includes radial tube wall deformation calculation, and the radial tube wall deformation calculation formula is:

[0073] Max.RWS=maximum diameter-minimum diameter

[0074] Max.RWS%=(maximum diameter-minimum diameter) / maximum diameter

[0075] When Max.RWS% is greater than a preset threshold, unstable plaque exists in the blood vessel segment.

[0076] Preferably, when the normal threshold value of Max.RWS% is set to 13%, when the result of calculating Max.RWS% is greater than 13%, it is considered that unstable plaque exists in the portion.

[0077] Specifically, Max.RWS is the absolute radial wall deformation of the blood vessel, and Max.RWS% is the relative radial wall deformation of the blood vessel.

[0078] During the cardiac cycle, the diameter of the blood vessels changes as the heart contracts and relaxes. This formula calculates the percentage change of the blood vessel from the minimum diameter to the maximum diameter relative to its minimum diameter. When the result of Max.RWS% is large, it means that the blood vessel wall has undergone a large deformation during the cardiac cycle, which may be related to a variety of factors, one of which is the possible presence of plaques in the blood vessel wall.

[0079] Furthermore, plaques can cause increased rigidity of the blood vessel walls: plaques are usually composed of fat, cholesterol, calcium and other substances that are deposited in the blood vessel walls, making the blood vessel walls locally harder and more rigid. Therefore, when the heart contracts, healthy blood vessel walls can more easily expand to accommodate more blood, while blood vessel walls containing plaques may limit their ability to expand due to the rigidity of the plaques.

[0080] The presence of plaques can also cause local narrowing of the blood vessel lumen: further restricting the flow of blood in the narrow area. During systole, the blood needs to pass through the narrow area quickly, which may cause the blood vessel wall in this area to be subjected to higher pressure, thereby increasing its deformation.

[0081] In one embodiment, when the contour segmentation result is not good, the segmentation result is manually modified; and / or, when the registration result is not good, forced registration points are added to the key frame and the analysis frame.

[0082] Specifically, when the effect of automatic segmentation of contours is not good, it needs to be manually modified, and the modification includes local correction, edge refinement and region merging or segmentation.

[0083] Furthermore, local correction includes: manually correcting the regions with obvious errors in the automatic segmentation results;

[0084] Edge refinement includes: manually adjusting the contour edge to make it more precise;

[0085] Region merging or segmenting includes: merging adjacent similar regions or further segmenting them.

[0086] When the registration effect is not good, you can add forced registration points, including:

[0087] Select salient features: Select points with salient features in the image as mandatory registration points. These points should be clearly identifiable in the images of both the key frame and the analysis frame.

[0088] Precise Marking: The positions of the mandatory registration points are precisely marked to ensure their correspondence in the keyframes and analysis frames.

[0089] Optimize registration parameters: Adjust the parameters of the registration algorithm according to the positions of the mandatory registration points to improve the registration accuracy.

[0090] In one embodiment, when the corresponding points in the analysis frame and the key frame are not in good position, registration point marks are added to the corresponding point positions of the analysis frame and the key frame so that the positions corresponding to the registration point marks on the analysis frame and the key frame are forced to be matched.

[0091] Specifically, registration points, also called feature points or control points, are points selected in the frame that have significant features, are easy to identify, and can be stably tracked between different frames. These points can be corner points, edge intersections, centers of high-contrast areas, etc. Generally, computer vision algorithms, such as SIFT, SURF, ORB and other feature detection algorithms, can be used to automatically detect potential registration points in the frame. However, after automatic detection, manual adjustments may be required to ensure that the selected registration points accurately correspond to specific positions in the keyframes, especially when the automatic detection algorithm cannot accurately identify all necessary features. At the detected or adjusted registration point positions, use visual markers, such as small dots, crosshairs, digital labels, etc. to mark them so that these points can be visually distinguished and tracked. The color, size, and shape of the markers can be adjusted as needed to improve readability and recognition.

[0092] After marking the registration points, the key step is to establish the correspondence between these points in different frames. Specifically, it is to determine the matching correspondence between each registration point one by one or in batches between the analysis frame and the key frame. Generally, a matching algorithm, such as SIFT, SURF, ORB and other feature detection algorithms, will find the most similar feature points between the two frames as corresponding points. For the marked registration points, this correspondence can be established manually or automatically to ensure their correct matching between different frames.

[0093] A device for calculating coronary artery deformation during a cardiac cycle, such as Figure 6 As shown, it includes a memory 200, a processor 100 and a computer program stored in the memory 200. When the processor 100 executes the computer program, the method for calculating the coronary vessel deformation during the cardiac cycle is implemented.

[0094] A computer-readable medium having a computer program stored thereon, wherein when the computer program is executed by the processor 100, the method for calculating coronary vessel deformation during the cardiac cycle is implemented.

[0095] A device for calculating coronary artery deformation during a cardiac cycle comprises an input device 300, an output device 400, a memory 200, a processor 100 and a computer program. When the computer program is executed by the processor, the method for calculating coronary artery deformation during a cardiac cycle is implemented.

[0096] An exemplary device for calculating coronary vessel deformation during a cardiac cycle is as follows Figure 7 As shown, the device for calculating coronary artery deformation during the cardiac cycle may be a device for calculating coronary artery deformation during the cardiac cycle integrated with a first imaging device, or a stand-alone device independent of the first imaging device, and the stand-alone device may communicate with the first imaging device to receive the collected input signal from the first imaging device. Figure 7 As shown, the coronary vessel deformation calculation device during the cardiac cycle includes one or more processors 100 and a memory 200. The processor 100 may be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and may control other components in the coronary vessel deformation calculation device during the cardiac cycle to perform desired functions. The memory 200 may include one or more coronary vessel deformation calculation devices during the cardiac cycle, and the coronary vessel deformation calculation device during the cardiac cycle may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 100 may run the program instructions to implement the coronary vessel deformation calculation method during the cardiac cycle of the present application described above. Various contents such as the current frame image, the result image of the previous frame image, etc. may also be stored in the computer-readable storage medium.

[0097] In one example, the device for calculating coronary vessel deformation during a cardiac cycle may further include: an input device 300 and an output device 400, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0098] For example, when the coronary artery deformation calculation device is integrated with the first imaging device during the cardiac cycle, the input device 300 may be the first imaging device, such as a camera, for capturing each frame of an image of a moving object. When the electronic device is a stand-alone device, the input device 300 may be a communication network connector for receiving the collected input signal from the first imaging device.

[0099] In addition, the input device may also include, for example, a keyboard, a mouse, and the like.

[0100] The output device 400 can output various information to the outside, including the determined deflection angle information, etc. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0101] In one embodiment, a DSA angiography image is first imported, a frame at the end of diastole is selected as a key frame, and a blood vessel segment to be subjected to deformation analysis is selected on the key frame as a reference analysis segment, and then the automatic calculation module is entered. Figure 2 As shown in the figure, the image in the image window is the selected key frame, and the selection logic is: the lesion is clearly exposed and the blood vessels overlap less. Click the head and tail of the blood vessel segment to be analyzed for radial wall deformation on the software interactive interface to determine the reference analysis segment on the key frame. The white mask area in the figure is the blood vessel segment to be analyzed for radial wall deformation.

[0102] Then the following automatic calculations are completed:

[0103] First, the analysis frames are automatically selected, and the selection logic is: at least including the early diastole, early systole and late systole in the same cardiac cycle, and the analysis segments are clear with little overlap.

[0104] Then the analysis segment is automatically selected and segmented, and the logic is selected: the vascular segment that remains the same as the reference analysis segment on the key frame is segmented, and the contour is segmented.

[0105] Finally, the benchmark analysis segments and comparison analysis segments on the analysis frames and key frames are automatically aligned.

[0106] The results are as follows Figure 3As shown, the right side is the analysis phase, where K is marked as the key frame and the rest are analysis frames. The frames of the analysis phase are linked, and the image identification point will display the blood vessel diameter at the current position of each frame. The leftmost side of the image is the image window, where the blood vessel contours and diameters can be edited. Below it is a list of blood vessel diameters obtained after editing the blood vessel contours and diameters in the image window. The horizontal axis corresponds to the points on the diameter in the blood vessel analysis segment, and the vertical axis is the maximum and minimum diameters of the blood vessels at the corresponding points on the diameter. When the image identification point is at a point corresponding to the diameter in the blood vessel analysis segment, the interactive interface will display the deformation size and deformation rate of the current position. The chart can also be switched to difference mode, and the results are shown below. Figure 4 As shown, the vertical axis is the deformation difference value at each position in the analysis segment.

[0107] When the automatically selected comparison analysis segment has poor contour segmentation or poor registration effect, manual modification or adding of forced registration points is supported.

[0108] When the analysis segment of the analysis frame is consistent with the key frame, the contour segmentation effects of each frame are consistent, and the corresponding points of each analysis frame and the key frame are still incorrect, you can add forced registration points. Figure 5 As shown, the same anatomical structure points are marked on each frame, such as branch openings, stenosis locations, etc. The background will establish a strong correlation based on the marked positions and the algorithm, and update the registration correspondence table of the analysis segment to recalculate the radial wall deformation.

[0109] The present invention has the following beneficial effects:

[0110] 1. Simplicity: This technology directly matches key frames and analysis frames, performs contour segmentation and registration, and does not require complex image preprocessing and segmentation. In addition, this technical solution realizes the automatic matching of key frames and analysis frames, automatic contour segmentation and registration process, which significantly reduces the burden on staff and greatly improves work efficiency and convenience.

[0111] 2. Accuracy: This scheme analyzes the coronary artery diameter throughout the entire cardiac cycle, reflecting the changes in blood vessel diameter throughout the cardiac cycle, and further improving the accuracy of coronary artery diameter measurement.

[0112] In the description of the present application, it should be noted that the terms used herein are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present application. For ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The technology, methods and equipment known to ordinary technicians in the relevant field may not be discussed in detail, but in appropriate cases, the technology, methods and equipment should be regarded as part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as merely exemplary, rather than as a limitation. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0113] It should be noted that, in the present application, the terms "comprises", "includes" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. It should also be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0114] The above embodiments are provided for persons familiar with the art to implement or use the present application. Personnel familiar with the art can make various modifications or changes to the above embodiments without departing from the application concept of the present application. Therefore, the protection scope of the present application is not limited to the above embodiments, but should be the maximum scope of the innovative features mentioned in the claims.

Claims

1. A method for calculating coronary artery deformation during a cardiac cycle, characterized in that: The method for calculating coronary artery deformation during the cardiac cycle comprises: Acquire image data, and select a frame of image as a key frame, wherein the image data is coronary artery image data including the entire cardiac cycle; Positioning a blood vessel segment for deformation analysis in the key frame as a reference analysis segment; Selecting a plurality of analysis frames from the image data, and locating a blood vessel segment corresponding to the reference analysis segment in each of the analysis frames as a comparison analysis segment; The reference analysis segment and each comparison analysis segment are subjected to contour segmentation and registration to calculate the deformation of each position in the blood vessel segment to be analyzed.

2. The method for calculating coronary artery deformation during the cardiac cycle according to claim 1, characterized in that: The key frame selection conditions include at least one or more of: the cardiac cycle is in the end diastole, the lesion is clearly exposed, and there is little blood vessel overlap.

3. The method for calculating coronary artery deformation during the cardiac cycle according to claim 1, characterized in that: The analysis frame selection conditions include: the plurality of analysis frames and the key frame are in the same complete cardiac cycle, and the plurality of analysis frames respectively cover at least the mid-diastole, early-systole and end-systole of a complete cardiac cycle.

4. The method for calculating coronary artery deformation during the cardiac cycle according to claim 1, characterized in that: The registration includes registering corresponding positions of the reference analysis segment and the comparison analysis segment to obtain registered blood vessel parameters, wherein the blood vessel parameters include the maximum diameter and the minimum diameter of the blood vessel.

5. The method for calculating coronary artery deformation during the cardiac cycle according to claim 4, characterized in that: In step S104, the deformation calculation includes the calculation of the radial tube wall deformation amount, and the calculation formula of the radial tube wall deformation amount is: Max.RWS=maximum diameter-minimum diameter Max.RWS%=(maximum diameter-minimum diameter) / maximum diameter When Max.RWS% is greater than a preset threshold, unstable plaque exists in the blood vessel segment.

6. The method for calculating coronary artery deformation during the cardiac cycle according to claim 4, characterized in that: When the contour segmentation result is not good, the segmentation result is manually modified; and / or, when the registration result is not good, mandatory registration points are added in the key frame and the analysis frame.

7. The method for calculating coronary artery deformation during the cardiac cycle according to claim 6, characterized in that: When the corresponding points in the analysis frame and the key frame are not in good position, by adding registration point marks to the corresponding point positions of the analysis frame and the key frame, the positions corresponding to the registration point marks on the analysis frame and the key frame are forced to match.

8. A device for calculating coronary artery deformation during a cardiac cycle, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: When executing the computer program, the processor implements the method for calculating coronary vessel deformation during the cardiac cycle according to any one of claims 1 to 7.

9. A computer-readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for calculating coronary vessel deformation during a cardiac cycle according to any one of claims 1 to 7 is implemented.

10. A device for calculating coronary artery deformation during a cardiac cycle, comprising an input device, an output device, a memory, a processor and a computer program, characterized in that: When the computer program is executed by a processor, the method for calculating coronary vessel deformation during a cardiac cycle according to any one of claims 1 to 7 is implemented.

Images