Blood vessel registration method and apparatus, electronic device, and storage medium
By acquiring feature point sets from intravascular and external images and determining multidimensional feature sequences for registration, the problem of adding equipment and changing the diagnostic process in existing technologies is solved, achieving accurate registration between intravascular and external images and improving diagnostic efficiency.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- SUZHOU PULSE RONGYING MEDICAL TECH CO LTD
- Filing Date
- 2025-12-10
- Publication Date
- 2026-07-02
Smart Images
Figure CN2025141348_02072026_PF_FP_ABST
Abstract
Description
Vascular registration methods, devices, electronic equipment and storage media
[0001] This application claims priority to Chinese Patent Application No. 202411915640.5, filed with the Chinese Patent Office on December 24, 2024, the entire contents of which are incorporated herein by reference. Technical Field
[0002] This application relates to the field of medical technology, such as a vascular registration method, apparatus, electronic device, and storage medium. Background Technology
[0003] With the rapid development of medical imaging technology, evaluating the degree of coronary artery stenosis and plaque stability through medical image analysis and guiding percutaneous coronary intervention is a crucial part of clinical treatment.
[0004] Coronary angiography (CAG) is currently the most commonly used extravascular imaging modality. Using contrast agents and X-rays, it images the coronary arteries, providing an overall anatomical view and aiding in determining the precise location and extent of lesions. However, it only offers a two-dimensional assessment of lumen size and cannot provide information on plaque morphology or atherosclerotic burden, making it difficult to clearly visualize lesions within the vessel wall. Intravascular ultrasound (IVUS), as the current mainstream intravascular imaging technique, can observe the morphology of the lumen and plaque, and can also determine the nature of lesions based on echo characteristics. It facilitates the quantitative measurement of lumen and plaque size and the degree of luminal stenosis, and its penetration is far superior to optical coherence tomography (OCT). However, it faces the challenge of locating the ideal stent edge landing area defined by IVUS on CAG images. Therefore, combining multimodal image registration is of great significance for the development of clinical diagnostic and treatment techniques.
[0005] The registration methods for intracavitary and extracavitary images in related technologies involve methods such as detecting and tracking IVUS sensors. These requirements increase the need for additional equipment and alter the normal diagnostic procedures in the catheterization lab. Summary of the Invention
[0006] This application provides a vascular registration method, apparatus, electronic device, and storage medium to address the problem that related technologies require additional equipment and alter the normal diagnostic process in a catheterization lab.
[0007] According to one aspect of this application, a vascular registration method is provided, comprising:
[0008] Obtain the first set of lumen feature points in the intravascular image corresponding to the target vascular segment, and determine the first multidimensional feature sequence corresponding to the first set of lumen feature points.
[0009] Obtain the second lumen feature point set of the extravascular image corresponding to the target vascular segment, and determine the second multidimensional feature sequence corresponding to the second lumen feature point set;
[0010] Based on the first multidimensional feature sequence and the second multidimensional feature sequence, blood vessel registration is performed to obtain a registration point set.
[0011] Optionally, the method further includes: resampling at least one of the first feature points in the first lumen feature point set or the second feature points in the second lumen feature point set based on the same sampling rate.
[0012] Optionally, determining the first multidimensional feature sequence corresponding to the first lumen feature point set includes: obtaining a first initial feature sequence corresponding to the first lumen feature point set, performing feature extraction on the first initial feature sequence, and obtaining the first multidimensional feature sequence.
[0013] The first initial feature sequence includes initial feature information of a first feature point in the first lumen feature point set, the initial feature information including the identifier of the first feature point and the vessel diameter information; the first multidimensional feature sequence includes multidimensional feature data corresponding to the first feature point, the multidimensional feature data including at least one of the identifier of the first feature point, vessel diameter information, vessel diameter gradient features and vessel diameter event features.
[0014] Optionally, the feature extraction of the first initial feature sequence includes: for each first feature point, determining a neighborhood of a preset length for the first feature point; and determining the blood vessel diameter gradient feature of the first feature point based on the blood vessel diameter information of multiple neighborhood feature points within the neighborhood.
[0015] Optionally, the step of extracting features from the first initial feature sequence includes: obtaining a first branch feature point set of the intravascular image; the first branch feature point set includes branch feature points corresponding to each branch vessel in the intravascular image; and processing the vessel diameter information of the first feature point corresponding to the branch feature point in the first lumen feature point set based on preset processing parameters to obtain the processed vessel diameter information of the first feature point.
[0016] Optionally, the feature extraction of the first initial feature sequence further includes: obtaining a blood vessel diameter sequence formed by the blood vessel diameter information of the plurality of first feature points in the first lumen feature point set, wherein the blood vessel diameter information of the first feature point includes at least one of the processed blood vessel diameter information and the unprocessed blood vessel diameter information; for each first feature point, determining the event interval corresponding to the first feature point within the blood vessel diameter sequence, wherein the event type of each first feature point is consistent within the event interval; and determining the event feature of the first feature point based on the event interval corresponding to the first feature point.
[0017] Optionally, determining the second multidimensional feature sequence corresponding to the second lumen feature point set includes: sliding a preset sliding window on the second lumen feature point set to obtain multiple local feature point sets corresponding to the window region, wherein each local feature point set corresponding to the window region includes multiple second feature points within the window region; determining the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively; the second multidimensional feature sequence corresponding to the second lumen feature point set includes the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively.
[0018] Optionally, the step of registering blood vessels based on the first multidimensional feature sequence and the second multidimensional feature sequence to obtain a registration point set includes: matching the local multidimensional feature sequences corresponding to the plurality of local feature point sets with the first multidimensional feature sequence corresponding to the first lumen feature point set to determine the matching degree data corresponding to the plurality of local feature point sets; and determining the registration point set based on the matching degree data corresponding to the plurality of local feature point sets.
[0019] According to another aspect of this application, a blood vessel registration device is provided, comprising:
[0020] The first multidimensional feature sequence acquisition module is configured to acquire the first lumen feature point set of the intravascular image of the target vascular segment and determine the first multidimensional feature sequence corresponding to the first lumen feature point set.
[0021] The second multidimensional feature sequence acquisition module is configured to acquire the second lumen feature point set of the extravascular image of the target blood vessel segment and determine the second multidimensional feature sequence corresponding to the second lumen feature point set.
[0022] The registration module is configured to perform blood vessel registration based on the first multidimensional feature sequence and the second multidimensional feature sequence to obtain a registration point set.
[0023] According to another aspect of this application, an electronic device is provided, the electronic device comprising:
[0024] At least one processor; and
[0025] A memory communicatively connected to the at least one processor; wherein,
[0026] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the vascular registration method described in any embodiment of this application.
[0027] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the blood vessel registration method according to any embodiment of this application. Attached Figure Description
[0028] Figure 1 is a flowchart of a blood vessel registration method provided in Embodiment 1 of this application;
[0029] Figure 2 is a schematic diagram of a blood vessel registration provided in Embodiment 1 of this application;
[0030] Figure 3 is a flowchart of a blood vessel registration method provided in Embodiment 2 of this application;
[0031] Figure 4 is a schematic diagram of the branched blood vessel provided in Embodiment 2 of this application;
[0032] Figure 5 is a flowchart of a blood vessel registration method provided in Embodiment 3 of this application;
[0033] Figure 6 is a schematic diagram of a matching process provided in Embodiment 3 of this application;
[0034] Figure 7 is a schematic diagram of a blood vessel registration device provided in Embodiment 4 of this application;
[0035] Figure 8 is a schematic diagram of the structure of an electronic device provided in Embodiment 5 of this application. Detailed Implementation
[0036] It should be noted that the terms "first lumen feature point set," "second lumen feature point set," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0037] Example 1
[0038] Figure 1 is a flowchart of a vascular registration method provided in Embodiment 1 of this application. This embodiment is applicable to the vascular registration of intravascular ultrasound images and coronary angiography images. The method can be executed by a vascular registration device, which can be implemented in hardware and / or software and can be configured in a computer, server, etc. As shown in Figure 1, the method includes:
[0039] S110. Obtain the first set of lumen feature points in the intravascular image corresponding to the target vascular segment, and determine the first multidimensional feature sequence corresponding to the first set of lumen feature points.
[0040] In this context, the target vascular segment refers to the vascular segment of interest, and the intravascular image can be an intravascular ultrasound image of the target vascular segment, which is composed of an image sequence. In this embodiment, the intravascular image corresponding to the target vascular segment is acquired, and lumen segmentation is performed on the intravascular image corresponding to the target vascular segment to obtain a first lumen segmentation result. The implementation method of lumen segmentation for the intravascular image is not limited here. Each frame of the image sequence is used as a sampling point, and the first lumen segmentation result is sampled to obtain a first lumen feature point set. The first lumen feature point set includes multiple first feature points.
[0041] Initial feature information for each feature point in the first lumen feature point set is obtained from the intravascular image. This initial feature information includes the identifier of the first feature point and the vessel diameter information. The identifier of the first feature point can be its sequence number, which can be obtained by sequentially assigning multiple first feature points from proximal to distal. The vessel diameter information can be the diameter of the vessel at the location of the first feature point, which can be extracted from the intravascular image.
[0042] By extracting the initial feature information of each feature point in the first lumen feature point set, multidimensional feature data of each feature point is obtained, and the multidimensional feature data in the first lumen feature point set forms the first multidimensional feature sequence corresponding to the first lumen feature point set.
[0043] The first multidimensional feature sequence is composed of multidimensional feature data of the first feature point. The multidimensional feature data of the first feature point includes the identifier of the first feature point, the blood vessel diameter information, the blood vessel diameter gradient feature, and the blood vessel diameter event feature. Accordingly, the first multidimensional feature sequence includes the identifier sequence of the first feature point, the blood vessel diameter sequence, the blood vessel diameter gradient feature sequence, and the blood vessel diameter event feature; or the first multidimensional feature sequence can be understood as a sequence formed by multidimensional data groups of multiple first feature points.
[0044] S120. Obtain the second lumen feature point set of the extravascular image corresponding to the target vascular segment, and determine the second multidimensional feature sequence corresponding to the second lumen feature point set.
[0045] The extravascular image can be a coronary angiography image of the target vessel segment. In this embodiment, the extravascular image corresponding to the target vessel segment is acquired, and the lumen of the extravascular image corresponding to the target vessel segment is segmented to obtain a second lumen segmentation result. The vessel diameter of the extravascular image corresponding to the target vessel segment is sampled to obtain multiple second feature points, which form a second lumen feature point set. For example, the vessel diameter of the extravascular image can be sampled based on a preset sampling rate, or the vessel diameter of the extravascular image can be uniformly sampled based on a preset number of feature points.
[0046] Initial feature information for each feature point in the second lumen feature point set is obtained from the extravascular image. This initial feature information includes the identifier of the second feature point and the vessel diameter information. The identifier of the second feature point can be its sequence number, which can be obtained by sequentially assigning multiple second feature points from proximal to distal. The vessel diameter information can be the diameter of the vessel at the location of the second feature point, which can be extracted from the extravascular image.
[0047] By extracting the initial feature information of each feature point in the second lumen feature point set, multidimensional feature data of each feature point is obtained. The multidimensional feature data in the second lumen feature point set forms the second multidimensional feature sequence corresponding to the second lumen feature point set.
[0048] The second multidimensional feature sequence is composed of multidimensional feature data of the second feature point. This multidimensional feature data includes the identifier of the second feature point, vessel diameter information, vessel diameter gradient features, and vessel diameter event features. Accordingly, the second multidimensional feature sequence includes the identifier sequence of the second feature point, the vessel diameter sequence, the vessel diameter gradient feature sequence, and the vessel diameter event features; or, the second multidimensional feature sequence can be understood as a sequence formed by multidimensional data sets of multiple second feature points.
[0049] In some embodiments, the method may optionally further include: resampling a first feature point in the first lumen feature point set and / or a second feature point in the second lumen feature point set based on the same sampling rate.
[0050] Since the average physical length of adjacent sampling points in the first lumen feature point set may differ from that in the second lumen feature point set, it is necessary to resample the first feature point in the first lumen feature point set and / or the second feature point in the second lumen feature point set to ensure that the sampling rate between adjacent sampling points in the first lumen feature point set is consistent with that in the second lumen feature point set. Resampling methods include, but are not limited to, nearest neighbor interpolation, bilinear interpolation, and cubic spline interpolation, etc., and are not limited here.
[0051] In this embodiment, a target sampling rate (i.e., the same sampling rate for resampling) is selected from a first sampling rate corresponding to a first set of lumen feature points and a second sampling rate corresponding to a second set of lumen feature points. Resampling is then performed on the first feature points in the first set of lumen feature points or the second feature points in the second set of lumen feature points based on the target sampling rate. The target sampling rate can be the larger of the first and second sampling rates. For example, if the target sampling rate is the first sampling rate, the second feature points in the second set of lumen feature points are resampled based on the first sampling rate; if the target sampling rate is the second sampling rate, the first feature points in the first set of lumen feature points are resampled based on the second sampling rate. Optionally, a third sampling rate can be preset, and resampling is performed on the first feature points in the first set of lumen feature points and the second feature points in the second set of lumen feature points based on the preset third sampling rate. The third sampling rate can be greater than the first sampling rate, and the third sampling rate can be greater than the second sampling rate.
[0052] Optionally, after resampling, the identifiers of the first feature points in the first lumen feature point set are adaptively updated, and the identifiers of the first feature points are determined based on the position of the first feature points in the target vascular segment; and / or, the identifiers of the second feature points in the second lumen feature point set are adaptively updated, and the identifiers of the second feature points are determined based on the position of the second feature points in the target vascular segment.
[0053] S130. Based on the first multidimensional feature sequence and the second multidimensional feature sequence, perform blood vessel registration to obtain a registration point set.
[0054] In this embodiment, the first multidimensional feature sequence and the second multidimensional feature sequence are registered to obtain a registration point set. It is understood that the number of feature points in the longer point set of the first lumen feature point set or the second lumen feature point set may differ, and the registration point set may be a local point set from the feature point set with the larger number of feature points.
[0055] Optionally, the registration point set can be a local point set within a first lumen feature point set, including local first feature points within the first lumen feature point set, where the vessel segment corresponding to the registration point set matches the vessel segment corresponding to the extravascular image. Optionally, the registration point set can be a local point set within a second lumen feature point set, including local second feature points within the second lumen feature point set, where the vessel segment corresponding to the registration point set matches the vessel segment corresponding to the intravascular image. For example, see Figure 2, which is a schematic diagram of vessel registration provided in an embodiment of this application.
[0056] In some embodiments, the first multidimensional feature sequence and the second multidimensional feature sequence can be registered using a registration model. For example, the first multidimensional feature sequence and the second multidimensional feature sequence are input into the registration model, and a set of registration points is output.
[0057] By determining the first multidimensional feature sequence corresponding to the first lumen feature point set and the second multidimensional feature sequence corresponding to the second lumen feature point set, the feature comprehensiveness of the first and second multidimensional feature sequences is improved, providing a feature basis for vascular registration and further improving the accuracy of the vascular registration process.
[0058] The technical solution of this embodiment involves acquiring a first set of lumen feature points from the intravascular image corresponding to the target vascular segment, determining a first multidimensional feature sequence corresponding to the first set of lumen feature points, acquiring a second set of lumen feature points from the extravascular image corresponding to the target vascular segment, and determining a second multidimensional feature sequence corresponding to the second set of lumen feature points. Based on the first and second multidimensional feature sequences, vascular registration is performed to obtain a registration point set. This enables rapid registration of intravascular and extravascular images without requiring additional equipment, thus eliminating the need to alter the normal diagnostic procedures in the catheterization lab.
[0059] Example 2
[0060] Figure 3 is a flowchart of a blood vessel registration method provided in Embodiment 2 of this application. Based on the above embodiments, optionally, determining the first multidimensional feature sequence corresponding to the first lumen feature point set includes: obtaining a first initial feature sequence corresponding to the first lumen feature point set; performing feature extraction on the first initial feature sequence to obtain the first multidimensional feature sequence; the first initial feature sequence includes initial feature information of a first feature point in the first lumen feature point set, the initial feature information including the identifier of the first feature point and blood vessel diameter information; the first multidimensional feature sequence includes multidimensional feature data corresponding to the first feature point, the multidimensional feature data including one or more of the identifier of the first feature point, blood vessel diameter information, blood vessel diameter gradient features, and blood vessel diameter event features. As shown in Figure 3, the method includes:
[0061] S210. Obtain the first lumen feature point set of the intravascular image corresponding to the target blood vessel segment, obtain the first initial feature sequence corresponding to the first lumen feature point set, extract features from the first initial feature sequence, and obtain the first multidimensional feature sequence.
[0062] The first initial feature sequence includes initial feature information of the first feature point in the first lumen feature point set, the initial feature information including the identifier of the first feature point and the vessel diameter information; the first multidimensional feature sequence includes multidimensional feature data corresponding to the first feature point, the multidimensional feature data including one or more of the identifier of the first feature point, vessel diameter information, vessel diameter gradient features and vessel diameter event features.
[0063] The first initial feature sequence refers to the feature sequence formed by the identifier and vessel diameter information of each first feature point in the first lumen feature point set before feature extraction. In this embodiment, the first initial feature sequence is constructed based on the identifier and vessel diameter information of each first feature point in the first lumen feature point set. Feature extraction is performed on the first initial feature sequence to obtain the vessel diameter gradient feature and vessel diameter event feature of each first feature point. A first multidimensional feature sequence is constructed based on the identifier, vessel diameter information, vessel diameter gradient feature, and vessel diameter event feature of each first feature point. In some embodiments, optionally, before feature extraction of the first initial feature sequence, the method further includes: normalizing and smoothing the vessel diameter information in the first initial feature sequence. The normalization method includes, but is not limited to, Min-Max normalization, Z-score normalization, logarithmic normalization, standard deviation normalization, etc., and is not limited here. The smoothing method includes, but is not limited to, moving average method, exponential smoothing method, polynomial smoothing method, low-pass filter smoothing method, Kalman filter smoothing method, etc., and is not limited here.
[0064] Based on the above embodiments, optionally, the feature extraction of the first initial feature sequence includes: for each first feature point, determining a neighborhood of a preset length for the first feature point; and determining the blood vessel diameter gradient feature of the first feature point based on the blood vessel diameter information of multiple neighborhood feature points within the neighborhood.
[0065] In this embodiment, multiple first feature points in the first lumen feature point set are sorted based on identifiers. For each first feature point, a neighborhood of the first feature point is determined based on the first feature point and a preset length. The number of neighborhood feature points within the neighborhood of the first feature point is related to the preset length; for example, the preset length can be understood as the number of first feature points, such as a preset length of 50, correspondingly, the number of neighborhood feature points is 50. The first feature point is located at the center of the neighborhood, and the neighborhood feature points are located before or after the first feature point. For each feature point within the neighborhood, the vessel diameter gradient data of that feature point is calculated, and the vessel diameter gradient data corresponding to the multiple feature points within the neighborhood form the vessel diameter gradient feature of the first feature point. The multiple feature points within the neighborhood include the first feature point and its neighboring feature points.
[0066] Optionally, the blood vessel diameter gradient feature of the first feature point is determined based on the blood vessel diameter information of multiple neighboring feature points within the neighborhood; for example, the calculation formula for the blood vessel diameter gradient data of any feature point within the neighborhood is as follows:
[0067] ;
[0068] in, It is the gradient feature of the blood vessel diameter at the i-th neighboring feature point. This represents the blood vessel diameter information of the i-th neighboring feature point. This represents the blood vessel diameter information of the (i-1)th neighboring feature point. This represents the blood vessel diameter information of the (i+1)th neighboring feature point within the neighborhood.
[0069] Understandably, since the first and last feature points in the neighborhood have only one neighboring feature point, for a preset length l, the dimension of the blood vessel diameter gradient feature is... That is to say, each first feature point includes The dimension of the blood vessel diameter gradient feature. For example, if the preset length is 50, the dimension of the blood vessel diameter gradient feature is 48.
[0070] Based on the above embodiments, optionally, the feature extraction of the first initial feature sequence includes: obtaining a first branch feature point set of the intravascular image; the first branch feature point set includes branch feature points corresponding to each branch vessel in the intravascular image; for the first feature point corresponding to the branch feature point in the first lumen feature point set, the vessel diameter information of the first feature point is processed based on preset processing parameters to obtain the processed vessel diameter information of the first feature point.
[0071] The first branch feature point set is used to describe the branch location information of the intravascular image sequence. In this embodiment, the vascular branches in the intravascular image are segmented to obtain a branch segmentation result. The midpoint of multiple first feature points in the location region corresponding to the same branch vessel in the first lumen feature point set is taken as the branch feature point. It can be understood that the location region corresponding to the first lumen feature point set includes multiple first feature points, and the midpoint of the location region is the midpoint among the multiple first feature points. For example, a certain branch vessel corresponds to first feature points 100 to 110 in the first lumen feature point set, and correspondingly, the branch feature point corresponding to this branch vessel can be 105. The target vessel segment may include multiple branch vessels, and the branch feature points corresponding to the multiple branch vessels form the first branch feature point set.
[0072] It should be noted that when resampling the first lumen feature point set, the branch feature points in the first branch feature point set need to be updated according to the identifier of the first feature point in the resampled first lumen feature point set.
[0073] For the first feature point corresponding to the branch feature point in the first lumen feature point set, the vessel diameter information of the first feature point is processed based on preset processing parameters to obtain the processed vessel diameter information of the first feature point. It can be understood that the vessel diameter information of other first feature points besides the branch feature point remains unchanged. For example, the formula for processing the vessel diameter information is as follows: ;in, This represents the processed blood vessel diameter information of the first feature point. Information indicating the diameter of the blood vessel at the first feature point. This is a preset processing parameter, which can be 0.01.
[0074] Optionally, based on the above embodiments, the feature extraction of the first initial feature sequence further includes: obtaining a blood vessel diameter sequence formed by the blood vessel diameter information of the plurality of first feature points in the first lumen feature point set, wherein the blood vessel diameter information of the first feature point includes at least one of the processed blood vessel diameter information and the unprocessed blood vessel diameter information; for each first feature point, determining the event interval corresponding to the first feature point within the blood vessel diameter sequence, wherein the event type of each first feature point is consistent within the event interval; and determining the event feature of the first feature point based on the event interval corresponding to the first feature point.
[0075] The vessel diameter event can include uphill events and downhill events. An uphill event can be understood as an event in which the vessel diameter information continuously increases, and a downhill event can be understood as an event in which the vessel diameter information continuously decreases. An event interval can be understood as an interval with a consistent event type, including multiple first feature points of the same time type. The event interval includes an uphill interval where the vessel diameter information changes monotonically and a downhill interval where the vessel diameter information changes monotonically and decreases. In this embodiment, a vessel diameter sequence is formed by obtaining the vessel diameter information of multiple first feature points in the first lumen feature point set. The vessel diameter sequence includes the vessel diameter information of the multiple first feature points in the first lumen feature point set before processing; or, the vessel diameter sequence includes the vessel diameter information of the multiple first feature points in the first lumen feature point set after processing.
[0076] For each first feature point, the corresponding event interval within the vessel diameter sequence is determined. The vessel diameter information of the first feature point and its adjacent first feature points is obtained and compared. Based on the comparison results, the event type is determined. For example, for first feature point i, if the vessel diameter information of first feature point i+1 is greater than that of first feature point i, the event type is uphill; if the vessel diameter information of first feature point i+1 is less than that of first feature point i, the event type is downhill. If the vessel diameter information of first feature point i+1 is equal to that of first feature point i, the event type is neither downhill nor uphill. Similarly, the vessel diameter information of first feature point i-1 is compared with that of first feature point i to determine the event type. Taking the event type as uphill as an example, the first feature point i+1 is added to the event interval of the first feature point i. The blood vessel diameter information of the first feature point i+1 and the first feature point i+2 is compared to determine the event type. In the case of the event type as uphill, the first feature point i+2 is added to the event interval of the first feature point i, and so on, until the event type changes, the search stops, and the event interval corresponding to the first feature point i is obtained.
[0077] If the blood vessel diameter information within the event interval corresponding to the first feature point monotonically decreases, then the event feature of the first feature point is: If the blood vessel diameter information monotonically increases within the event interval corresponding to the first feature point, then the event feature of the first feature point is: If the blood vessel diameter information remains unchanged within the event interval corresponding to the first feature point, then the event feature of the first feature point is: Where event = {upslope, downslope}, upslope represents the event feature value when the event type is uphill, and downslope represents the event feature value when the event type is downhill. This represents the feature value of the i-th first feature point within the event interval. For example, the formula for calculating the feature value of the first feature point within the event interval is as follows:
[0078] ;
[0079] in, This represents the feature value of the i-th first feature point within the event interval. This represents the identifier of the i-th first feature point. and These represent the event intervals of the i-th first feature point. The lower limit feature point identifier and the upper limit feature point identifier.
[0080] It is understandable that each first feature point of the first multidimensional feature sequence includes The dimensional feature data specifically includes: the one-dimensional identifier of the first sampling point, one-dimensional vessel diameter information, and two-dimensional vessel diameter event features. Characteristics of the gradient of blood vessel diameter.
[0081] S220. Obtain the second lumen feature point set of the extravascular image corresponding to the target vascular segment, obtain the second initial feature sequence corresponding to the second lumen feature point set, perform feature extraction on the second initial feature sequence, and determine the second multidimensional feature sequence corresponding to the second lumen feature point set.
[0082] The second initial feature sequence includes the identifier of the second feature point and the vessel diameter information. Before feature extraction from the second initial feature sequence, the vessel diameter information in the second initial feature sequence is normalized and smoothed.
[0083] Optionally, feature extraction is performed on the second initial feature sequence, including: for each second feature point, determining a neighborhood of a preset length for the second feature point; and determining the blood vessel diameter gradient feature of the second feature point based on the blood vessel diameter information of multiple neighboring feature points within the neighborhood. Further details are omitted here.
[0084] Optionally, the feature extraction of the second initial feature sequence includes: obtaining a second branch feature point set of the extravascular image; the second branch feature point set includes branch feature points corresponding to each branch vessel in the extravascular image; and processing the vessel diameter information of the second feature point corresponding to the branch feature point in the second lumen feature point set based on preset processing parameters to obtain the processed vessel diameter information of the second feature point. The processed vessel diameter information is as follows: This will not be elaborated upon here.
[0085] For example, Figure 4 is a schematic diagram of the branch vessel structure provided in Embodiment 2 of this application. As shown in Figure 4, points A and B are the intersection points of the branch vessel and the main trunk, point C is the intersection point of the branch diameter and the main trunk, and 100, 105 and 110 are the first feature points corresponding to points A, B and C on the branch vessel in the first lumen feature point set, and 105 is the midpoint, which is used as the branch feature point.
[0086] Optionally, feature extraction of the second initial feature sequence further includes: obtaining a blood vessel diameter sequence formed by the blood vessel diameter information of the plurality of second feature points in the second lumen feature point set, wherein the blood vessel diameter information of the second feature point includes at least one of the processed blood vessel diameter information and the unprocessed blood vessel diameter information; for each second feature point, determining the event interval corresponding to the second feature point within the blood vessel diameter sequence, wherein the event type of each second feature point is consistent within the event interval; and determining the blood vessel diameter event feature of the second feature point based on the event interval corresponding to the second feature point. Further details are omitted here.
[0087] S230. Based on the first multidimensional feature sequence and the second multidimensional feature sequence, perform blood vessel registration to obtain a registration point set.
[0088] The technical solution of this embodiment extracts multi-dimensional data from the first lumen feature point set to provide rich feature information for registration. Based on the multi-dimensional data, a first multi-dimensional feature sequence is constructed, and then vascular registration is performed based on the first multi-dimensional feature sequence and the second multi-dimensional feature sequence to obtain a registration point set, thereby improving the accuracy of registration between intravascular and extravascular images.
[0089] Example 3
[0090] Figure 5 is a flowchart of a blood vessel registration method provided in Embodiment 3 of this application. Based on the above embodiments, optionally, determining the second multidimensional feature sequence corresponding to the second lumen feature point set includes: sliding a preset sliding window on the second lumen feature point set to obtain multiple local feature point sets corresponding to the window region, wherein any one of the local feature point sets corresponding to the window region includes multiple second feature points within the window region; determining the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively; the second multidimensional feature sequence corresponding to the second lumen feature point set includes the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively. As shown in Figure 5, the method includes:
[0091] S310. Obtain the first set of lumen feature points in the intravascular image corresponding to the target vascular segment, and determine the first multidimensional feature sequence corresponding to the first set of lumen feature points.
[0092] S320. Obtain the second lumen feature point set of the extravascular image corresponding to the target blood vessel segment. Based on a preset sliding window, slide on the second lumen feature point set to obtain multiple local feature point sets corresponding to the window region. The local feature point set corresponding to any window region includes multiple second feature points within the window region. Determine the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively. The second multidimensional feature sequence corresponding to the second lumen feature point set includes the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively.
[0093] Figure 6 is a schematic diagram of a matching process provided in Embodiment 3 of this application. In this embodiment, as shown in Figure 6, a preset sliding window is used to slide on the second lumen feature point set to obtain multiple local feature point sets corresponding to the window region. An initial local feature sequence corresponding to the local feature point set is obtained, and feature extraction is performed on the initial local feature sequence to obtain multiple local multidimensional feature sequences. The sliding window can slide based on a preset step size.
[0094] The local multidimensional feature sequence includes multidimensional feature data corresponding to the second feature point within the window region. This multidimensional feature data includes the identifier of the second feature point, vessel diameter information, vessel diameter gradient features, and vessel diameter event features. The feature extraction method for the multidimensional feature data is similar to the multidimensional feature data extraction method for the first multidimensional feature sequence in the above embodiment, and will not be repeated here.
[0095] S330. Match the local multidimensional feature sequences corresponding to the plurality of local feature point sets with the first multidimensional feature sequence corresponding to the first lumen feature point set to determine the matching degree data corresponding to the plurality of local feature point sets; determine the registration point set based on the matching degree data corresponding to the plurality of local feature point sets.
[0096] The matching degree data is used to characterize the similarity between the local multidimensional feature sequence and the first multidimensional feature sequence. The higher the matching degree data, the higher the degree of matching between the local feature point set and the first lumen feature point set.
[0097] In this embodiment, the matching degree data between the local multidimensional feature sequence and the first multidimensional feature sequence corresponding to each local feature point set can be determined based on a pre-set sequence matching algorithm. The pre-set sequence matching algorithm includes, but is not limited to, Euclidean distance algorithm, cosine similarity algorithm, mutual information algorithm, Hausdorff distance algorithm, Pearson similarity coefficient algorithm, Manhattan distance algorithm, and dynamic kernel correlation algorithm. Optionally, the pre-set sequence matching algorithm can be the Dynamic Time Warping (DTW) algorithm. The DTW algorithm performs matching calculations on the local multidimensional feature sequence and the first multidimensional feature sequence, finds the shortest path for each corresponding sequence, calculates the score, and finally, through weighted calculation, obtains the optimal matching path between the local multidimensional feature sequence and the first multidimensional feature sequence, outputting the matching degree data.
[0098] In this embodiment of the application, as shown in Figure 6, the local multidimensional feature sequences corresponding to multiple local feature point sets are matched with the first multidimensional feature sequence corresponding to the first lumen feature point set to obtain matching degree data corresponding to multiple local feature point sets. The matching degree data corresponding to multiple local feature point sets can be sorted to obtain the local feature point set with the highest matching degree. The registration point set is determined based on the local feature point set with the highest matching degree.
[0099] The technical solution of this embodiment constructs multiple local multidimensional feature sequences by sliding a sliding window, which can capture the features of the outer wall of the blood vessel more meticulously, thereby improving the accuracy of matching. In addition, the application of the sliding window allows for the efficient extraction of multiple local feature point sets on the extravascular image, avoiding the need to compare the entire image one by one, thus improving registration efficiency.
[0100] Example 4
[0101] Figure 7 is a schematic diagram of a blood vessel registration device provided in Embodiment 4 of this application. As shown in Figure 7, the device includes:
[0102] The first multidimensional feature sequence acquisition module 410 is configured to acquire the first lumen feature point set of the intravascular image corresponding to the target blood vessel segment and determine the first multidimensional feature sequence corresponding to the first lumen feature point set.
[0103] The second multidimensional feature sequence acquisition module 420 is configured to acquire the second lumen feature point set of the extravascular image corresponding to the target vascular segment and determine the second multidimensional feature sequence corresponding to the second lumen feature point set.
[0104] The registration module 430 is configured to perform blood vessel registration based on the first multidimensional feature sequence and the second multidimensional feature sequence to obtain a registration point set.
[0105] The technical solution of this embodiment involves acquiring a first set of lumen feature points from the intravascular image corresponding to the target vascular segment, determining a first multidimensional feature sequence corresponding to the first set of lumen feature points, acquiring a second set of lumen feature points from the extravascular image corresponding to the target vascular segment, and determining a second multidimensional feature sequence corresponding to the second set of lumen feature points. Based on the first and second multidimensional feature sequences, vascular registration is performed to obtain a registration point set. This enables rapid registration of intravascular and extravascular images without requiring additional equipment, thus eliminating the need to alter the normal diagnostic procedures in the catheterization lab.
[0106] Optionally, based on the above embodiments, the device further includes a resampling module, configured to: resample the first feature point in the first lumen feature point set and / or the second feature point in the second lumen feature point set based on the same sampling rate.
[0107] Based on the above embodiments, optionally, the first multidimensional feature sequence obtaining module 410 is configured to: obtain a first initial feature sequence corresponding to the first lumen feature point set, perform feature extraction on the first initial feature sequence to obtain the first multidimensional feature sequence; the first initial feature sequence includes initial feature information of the first feature point in the first lumen feature point set, the initial feature information including the identifier of the first feature point and the blood vessel diameter information; the first multidimensional feature sequence includes multidimensional feature data corresponding to the first feature point, the multidimensional feature data including one or more of the identifier of the first feature point, blood vessel diameter information, blood vessel diameter gradient features and blood vessel diameter event features.
[0108] Based on the above embodiments, optionally, the first multidimensional feature sequence obtaining module 410 includes a blood vessel diameter gradient feature determining unit, configured to determine a neighborhood of a preset length for each first feature point; and determine the blood vessel diameter gradient feature of the first feature point based on the blood vessel diameter information of multiple neighborhood feature points within the neighborhood.
[0109] Based on the above embodiments, optionally, the first multidimensional feature sequence obtaining module 410 includes a blood vessel diameter information processing unit, configured to obtain a first branch feature point set of the intravascular image; the first branch feature point set includes branch feature points corresponding to each branch blood vessel in the intravascular image; for the first feature point corresponding to the branch feature point in the first lumen feature point set, the blood vessel diameter information of the first feature point is processed based on preset processing parameters to obtain the processed blood vessel diameter information of the first feature point.
[0110] Based on the above embodiments, optionally, the first multidimensional feature sequence obtaining module 410 includes a blood vessel diameter event feature determining unit, configured to obtain a blood vessel diameter sequence formed by blood vessel diameter information of a plurality of first feature points in the first lumen feature point set, wherein the blood vessel diameter information of the first feature point includes at least one of the processed blood vessel diameter information and the unprocessed blood vessel diameter information; for each first feature point, determine the event interval corresponding to the first feature point in the blood vessel diameter sequence, wherein the event type of each first feature point is consistent within the event interval; and determine the event feature of the first feature point based on the event interval corresponding to the first feature point.
[0111] Based on the above embodiments, optionally, the second multidimensional feature sequence obtaining module 420 is configured to: slide on the second lumen feature point set based on a preset sliding window to obtain multiple local feature point sets corresponding to the window region, wherein any one of the local feature point sets corresponding to the window region includes multiple second feature points within the window region; determine the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively; the second multidimensional feature sequence corresponding to the second lumen feature point set includes the local multidimensional feature sequences corresponding to the multiple local feature point sets respectively.
[0112] Based on the above embodiments, optionally, the registration module 430 is configured to: match the local multidimensional feature sequences corresponding to the plurality of local feature point sets with the first multidimensional feature sequence corresponding to the first lumen feature point set to determine the matching degree data corresponding to the plurality of local feature point sets; and determine the registration point set based on the matching degree data corresponding to the plurality of local feature point sets.
[0113] The vascular registration device provided in this application embodiment can execute the vascular registration method provided in any embodiment of this application, and has the corresponding functional modules and effects of the execution method.
[0114] Example 5
[0115] Figure 8 is a schematic diagram of an electronic device provided in Embodiment 5 of this application. The electronic device 10 can be a variety of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or claimed herein.
[0116] As shown in Figure 8, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0117] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0118] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs several of the methods and processes described above, such as the vascular registration method.
[0119] In some embodiments, the vascular registration method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the vascular registration method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the vascular registration method by any other suitable means (e.g., by means of firmware).
[0120] The various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0121] Computer programs used to implement the vascular registration method of this application may be written in any combination of one or more programming languages. These computer programs may be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0122] Example 6
[0123] This application also provides a computer-readable storage medium storing computer instructions for causing a processor to execute a blood vessel registration method, the method comprising:
[0124] Obtain the first set of lumen feature points in the intravascular image corresponding to the target vascular segment, and determine the first multidimensional feature sequence corresponding to the first set of lumen feature points.
[0125] Obtain the second lumen feature point set of the extravascular image corresponding to the target vascular segment, and determine the second multidimensional feature sequence corresponding to the second lumen feature point set;
[0126] Based on the first and second multidimensional feature sequences, blood vessel registration is performed to obtain the registration point set.
[0127] In the context of this application, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A machine-readable storage medium may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fiber, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0128] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user (e.g., a cathode ray tube (CRT) or liquid crystal display (LCD) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0129] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0130] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. The client-server relationship is established by running computer programs on the respective computers. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system. It addresses the shortcomings of traditional physical hosts and Virtual Private Server (VPS) services, such as high management difficulty and weak business scalability.
[0131] It should be understood that the various processes shown above can be used to reorder, add, or delete steps. For example, the multiple steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.
Claims
1. A method for vascular registration, comprising: Obtain the first set of lumen feature points in the intravascular image corresponding to the target vascular segment, and determine the first multidimensional feature sequence corresponding to the first set of lumen feature points. Obtain the second lumen feature point set of the extravascular image corresponding to the target vascular segment, and determine the second multidimensional feature sequence corresponding to the second lumen feature point set; Based on the first multidimensional feature sequence and the second multidimensional feature sequence, blood vessel registration is performed to obtain a registration point set.
2. The method according to claim 1, further comprising: Based on the same sampling rate, at least one of the first feature points in the first lumen feature point set or the second feature point in the second lumen feature point set is resampled.
3. The method of claim 1, wherein, Determining the first multidimensional feature sequence corresponding to the first lumen feature point set includes: Obtain the first initial feature sequence corresponding to the first lumen feature point set, and perform feature extraction on the first initial feature sequence to obtain the first multidimensional feature sequence; The first initial feature sequence includes initial feature information of a first feature point in the first lumen feature point set, the initial feature information including the identifier of the first feature point and the vessel diameter information; the first multidimensional feature sequence includes multidimensional feature data corresponding to the first feature point, the multidimensional feature data including at least one of the identifier of the first feature point, vessel diameter information, vessel diameter gradient features and vessel diameter event features.
4. The method of claim 3, wherein, The feature extraction of the first initial feature sequence includes: For each of the first feature points, a neighborhood of a preset length is determined for the first feature point; The blood vessel diameter gradient feature of the first feature point is determined based on the blood vessel diameter information of multiple neighboring feature points within the neighborhood.
5. The method of claim 3, wherein, The feature extraction of the first initial feature sequence includes: Obtain a first branch feature point set of the intravascular image; the first branch feature point set includes branch feature points corresponding to each branch vessel in the intravascular image. For the first feature point corresponding to the branch feature point in the first lumen feature point set, the blood vessel diameter information of the first feature point is processed based on preset processing parameters to obtain the processed blood vessel diameter information of the first feature point.
6. The method according to claim 5, wherein the feature extraction of the first initial feature sequence further comprises: A blood vessel diameter sequence is formed by obtaining blood vessel diameter information of multiple first feature points in the first lumen feature point set, wherein the blood vessel diameter information of the first feature point includes at least one of the processed blood vessel diameter information and the unprocessed blood vessel diameter information. For each first feature point, an event interval corresponding to the first feature point within the blood vessel diameter sequence is determined, and the event type of each first feature point is consistent within the event interval. The event features of the first feature point are determined based on the event interval corresponding to the first feature point.
7. The method of claim 1, wherein, Determining the second multidimensional feature sequence corresponding to the second lumen feature point set includes: Based on a preset sliding window, the sliding is performed on the second lumen feature point set to obtain multiple local feature point sets corresponding to the window region. The local feature point set corresponding to any window region includes multiple second feature points within the window region. Determine the local multidimensional feature sequences corresponding to the plurality of local feature point sets respectively; the second multidimensional feature sequence corresponding to the second lumen feature point set includes the local multidimensional feature sequences corresponding to the plurality of local feature point sets respectively.
8. The method of claim 7, wherein, The process of registering blood vessels based on the first multidimensional feature sequence and the second multidimensional feature sequence to obtain a registration point set includes: Based on the matching of the local multidimensional feature sequences corresponding to the multiple local feature point sets with the first multidimensional feature sequence corresponding to the first lumen feature point set, the matching degree data corresponding to the multiple local feature point sets is determined. The registration point set is determined based on the matching degree data corresponding to the multiple local feature point sets.
9. A blood vessel registration device, comprising: The first multidimensional feature sequence acquisition module is configured to acquire the first lumen feature point set of the intravascular image of the target vascular segment and determine the first multidimensional feature sequence corresponding to the first lumen feature point set. The second multidimensional feature sequence acquisition module is configured to acquire the second lumen feature point set of the extravascular image of the target blood vessel segment and determine the second multidimensional feature sequence corresponding to the second lumen feature point set. The registration module is configured to perform blood vessel registration based on the first multidimensional feature sequence and the second multidimensional feature sequence to obtain a registration point set.
10. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the vascular registration method according to any one of claims 1-8.
11. A computer-readable storage medium storing computer instructions for causing a processor to execute the blood vessel registration method of any one of claims 1-8.