Intracranial three-dimensional cerebral blood flow scanning, scanning path planning methods and systems

Abstract

This application relates to the field of constructing three-dimensional intracranial blood flow models, and particularly to a method and system for intracranial three-dimensional cerebral blood flow scanning and scanning path planning. The method includes determining a theoretical model and an actual model; mapping the cerebral vascular reference region in the theoretical model to the actual model to determine a projection model; wherein the projection model has a predicted distribution region corresponding to this cerebral vascular region; determining a scanning path based on the predicted distribution region in the projection model; and sending the current scanning path to the scanning execution module. Constructing the scanning path through the predicted distribution region increases the probability of scanning the actual cerebral blood vessels at the locations traversed by the scanning path, enabling the probe of the scanning execution module to quickly and accurately scan the location of the cerebral blood vessels, thereby improving scanning efficiency and enhancing the efficiency of subsequent algorithms in constructing the cerebral vascular model.

Description

Technical Field

[0001] This application relates to the field of constructing intracranial three-dimensional cerebral blood flow models, and in particular to a method and system for intracranial three-dimensional cerebral blood flow scanning and scanning path planning. Background Technology

[0002] When screening and diagnosing patients with cerebrovascular disease, it is necessary to use specialized scanning equipment to scan the intracranial cerebral blood vessels of the patient's skull in order to construct an intracranial cerebral blood flow model. The commonly used method is to use ultrasound equipment such as transcranial Doppler ultrasound equipment to determine three-dimensional blood flow data through ultrasound echo signals to confirm the distribution and functional status of cerebral blood vessels.

[0003] In related technologies, such as the three-dimensional imaging noise reduction method, device, terminal equipment and storage medium for intracranial cerebral blood flow disclosed in Chinese invention patent application with publication number CN113040821A, the method in this application uses an ultrasound probe to perform ultrasound scanning on the intracranial cavity of the object to be tested, and determines three-dimensional blood flow data based on the ultrasound echo signal obtained from the scan. Then, the three-dimensional blood flow data is divided into several data layers according to the depth direction. Next, the interpolation data of each layer is determined according to the motion trajectory of the ultrasound probe. After inserting the interpolation data into each data layer, the three-dimensional reconstruction is completed.

[0004] Regarding the above technical solution, the inventors believe that although the above method can reflect the functional state of cerebral blood vessels more sensitively, during the ultrasound scanning process, the scanning probe of the ultrasound equipment needs to change different ultrasound incident angles to scan different areas on the patient's skull and find the cerebral blood vessels distributed in each area. Since the distribution of cerebral blood vessels in the skull may vary for different patients, it is difficult to ensure that the scanning probe can quickly and accurately scan the location of cerebral blood vessels, thus affecting the scanning efficiency. Summary of the Invention

[0005] The purpose of this application is to provide a scanning path planning method for intracranial three-dimensional cerebral blood flow scanning, which has the characteristic of improving scanning efficiency.

[0006] The above-mentioned objective of this application is achieved through the following technical solution:

[0007] A scanning path planning method for intracranial three-dimensional cerebral blood flow scanning includes: determining a theoretical model and an actual model; mapping the cerebral vascular reference region in the theoretical model to the actual model to determine a projection model; wherein the projection model has a predicted distribution region corresponding to this cerebral vascular region; determining a scanning path based on the predicted distribution region in the projection model; and sending the current scanning path to the scanning execution module.

[0008] By adopting the above technical solution and combining theoretical and practical models, the regions with a high probability of cerebral blood vessels appearing in the prediction are represented in the projection model through the predicted distribution region. By constructing the scanning path through the predicted distribution region, the probability of scanning the real cerebral blood vessels at the locations traversed by the scanning path can be higher. This enables the probe of the scanning execution module to quickly and accurately scan the location of cerebral blood vessels, thereby improving scanning efficiency and enhancing the efficiency of subsequent algorithms in constructing cerebral blood vessel models.

[0009] Optionally, the method for constructing the theoretical model includes: acquiring vascular reference data and constructing a theoretical cerebral vascular model based on the vascular reference data; acquiring skeletal reference data and constructing a theoretical skull model based on the skeletal reference data; wherein both the theoretical skull model and the actual model can reflect the shape of the human skull; and determining the theoretical model based on the theoretical cerebral vascular model and the theoretical skull model.

[0010] By employing the aforementioned technical solutions, a theoretical cerebral vascular model constructed using vascular reference data can reflect the high-probability distribution of cerebral blood vessels in different individuals. Similarly, a theoretical skull model constructed using skeletal reference data can reflect the high-probability shape of the skull in different individuals. Combining these two models can demonstrate the skull shape and corresponding cerebral vascular distribution in most individuals, providing valuable reference information. Based on the discrepancies between the theoretical skull model and the actual skull model, adjustments can be made to the cerebral vascular distribution area in the theoretical model to compensate for these discrepancies, ensuring that the predicted distribution area is more accurately mapped to the actual model.

[0011] Optionally, the specific method for determining the scanning path based on the predicted distribution area in the projection model includes: model partitioning: partitioning the model based on the projection model to determine multiple regions to be scanned; probability calculation: assessing the probability of cerebral blood vessels appearing in the region to be scanned based on the distribution of the predicted distribution area to determine the partition probability corresponding to the region to be scanned; and determining the scanning path based on the partition probability.

[0012] By adopting the above technical solution, the partition probability can reflect the probability of cerebral blood vessels appearing in the area to be scanned in the actual scan, and has a guiding and reference role for the scanning path. The scanning path constructed using the partition probability can pass through the locations where cerebral blood vessels appear with a high probability in the projection model, thereby improving the efficiency of the system scanning to obtain real cerebral blood vessels.

[0013] Optionally, the specific method for the model partitioning step includes: partitioning the projection model based on the partitioning level to determine multiple regions to be scanned; wherein the partitioning level affects the proportion of a single region to be scanned in the projection model; determining whether the current region to be scanned meets the partitioning iteration condition, updating the partitioning level according to the determination result, and returning the current partitioning level.

[0014] By adopting the above technical solution and utilizing the partition iteration condition, it is possible to determine whether the current range of the area to be scanned meets the requirements of subsequent scanning steps, and to update the range of the area to be scanned in a timely manner by adjusting the partition level.

[0015] Optionally, the specific method for determining whether the range of the area to be scanned meets the partitioning iteration conditions, updating the partitioning level based on the determination result, and returning the determined partitioning level includes: comparing the range of the current area to be scanned with the maximum range threshold and the minimum range threshold respectively, updating the partitioning level based on the comparison result, and returning the determined partitioning level.

[0016] By adopting the above technical solution, the maximum range threshold is used to limit the maximum range of the area to be scanned, thereby reducing the risk of the area to be scanned being too large, so that the probe of the scanning execution module can complete the scanning task more effectively; while the minimum range threshold is used to limit the minimum range of the area to be scanned, thereby reducing the risk of the area to be scanned being too small, and reducing the amount of repeated data collection caused by the probe scanning too much area to be scanned in a single scan.

[0017] Optionally, the specific method for the probability calculation step includes: calculating the probability calculation index associated with the region to be scanned; determining the partition probability corresponding to the region to be scanned based on the probability calculation index; wherein the probability calculation index includes the theoretical existence probability; the specific method for calculating the theoretical existence probability associated with the region to be scanned includes: calculating the theoretical existence probability associated with the region to be scanned based on the ratio between the portion of the predicted distribution region overlapping the region to be scanned and the whole region to be scanned.

[0018] By employing the above technical solution, the probability of a region corresponding to the area to be scanned is calculated using probability calculation indicators. These indicators include the theoretical existence probability, which is calculated using the proportion of the predicted distribution area within the area to be scanned. By calculating the region probability using the theoretical existence probability, the degree of occupancy of the predicted cerebral vascular morphology within the area to be scanned can be converted into the probability of cerebral blood vessels appearing in the area to be scanned.

[0019] Optionally, the specific method for calculating the theoretical existence probability associated with the region to be scanned further includes: classifying all regions to be scanned based on a comparison of the theoretical existence probability and the probability classification coefficient, and determining low-probability partitions and high-probability partitions; updating the theoretical existence probability of all low-probability partitions that satisfy the vascular connectivity condition based on the probability amplification coefficient; wherein the vascular connectivity condition includes the low-probability partition being located between two high-probability partitions.

[0020] By employing the above technical solution, the greater the proportion of the predicted cerebral blood vessels occupying the scanned area, the higher the probability of cerebral blood vessels appearing in that area. However, there is a possibility that the cerebral blood vessels themselves are relatively narrow, causing the predicted proportion of cerebral blood vessels occupying the scanned area to not effectively reflect the partition probability. Therefore, it is necessary to introduce a vascular connectivity condition. Using the vascular connectivity condition, when one low-probability partition is located between two high-probability partitions, it is assumed that, based on the assumption that both high-probability partitions contain cerebral blood vessels, the connected area between them must also contain cerebral blood vessels. Therefore, the probability of the low-probability partition located between two high-probability partitions is adjusted before proceeding to the next step to reduce the risk of missed cerebral blood flow sampling.

[0021] Optionally, before calculating the probability calculation index associated with the region to be scanned, the method further includes: determining at least one probability calculation index associated with the region to be scanned and index weights corresponding to the probability calculation indexes; wherein, the index weights include theoretical existence weights corresponding to theoretical existence probabilities; in the specific method of determining the partition probability corresponding to the region to be scanned based on the probability calculation indexes, the method includes: determining the partition probability corresponding to the region to be scanned based on all probability calculation indexes associated with the region to be scanned and index weights corresponding to the probability calculation indexes.

[0022] And / or, the probability calculation index further includes model matching degree, and the index weight further includes model matching weight corresponding to the model matching degree; the specific method for calculating the model matching degree associated with the scanned region includes: based on the scan results of the scanned region that has been scanned, calculating the deviation value between the cerebral vascular morphology of the scan results and the cerebral vascular morphology of the projection model, and determining the model matching degree based on the deviation value.

[0023] And / or, the probability calculation index further includes scan completion rate, and the index weight further includes a completion rate weight corresponding to the scan completion rate; the specific method for calculating the scan completion rate associated with the scanned region includes: determining the sum of adjacent blood vessels corresponding one-to-one with the scanned region based on the scan results of the scanned region that has been scanned; wherein, the sum of adjacent blood vessels is used to indicate the sum of the lengths of all cerebral blood vessels connected to this scanned region in the scan results, and the predicted lengths of cerebral blood vessels in this scanned region; comparing the values ​​of all adjacent blood vessel sums, and determining the maximum blood vessel length based on the adjacent blood vessel sum with the largest value; determining the scan completion rate associated with the scanned region based on the ratio of the adjacent blood vessel sum to the maximum blood vessel length;

[0024] And / or, the probability calculation index further includes image matching degree, and the index weight further includes image matching weight corresponding to the image matching degree; the specific method for calculating the image matching degree associated with the region to be scanned includes: determining the actual scanned image based on the scanning results of the region to be scanned that has been scanned; determining the theoretical scanned image, and calculating the difference between the cerebral blood vessel morphology in the actual scanned image and the cerebral blood vessel morphology in the theoretical scanned image, and determining the image matching degree associated with the region to be scanned based on this difference;

[0025] And / or, the probability calculation index further includes vascular correlation degree, and the index weight further includes vascular correlation weight corresponding to the vascular correlation degree; the specific method for calculating the vascular correlation degree associated with the scanned area includes: determining the current vascular position based on the scanning results of the scanned area that has been scanned; determining the vascular correlation degree corresponding to each scanned area based on the current vascular position and the predicted distribution area; wherein, according to the gradient direction in which the distance between the scanned area and the current vascular position gradually increases, and the distribution trend direction of the predicted distribution area, the vascular correlation degree corresponding to each scanned area gradually decreases.

[0026] By employing the above technical solutions, model matching degree can reflect the matching degree between the currently predicted projection model and the actual cerebral vascular model of the current patient. The higher the matching degree, the higher the reliability of the projection model and the higher the corresponding partition probability. Scan completion degree can predict the continuous cerebral blood vessels that can be obtained after scanning the area to be scanned. The greater the length of the cerebral blood vessels that can be obtained, the higher the scan completion degree and the higher the corresponding partition probability. Image matching degree can reflect the matching degree between the actual ultrasound image of the current patient and the predicted ultrasound image. The higher the matching degree, the higher the corresponding partition probability. Vascular correlation degree can be used to radiate the influence of each area to be scanned based on the actual location of cerebral blood vessels. This is mainly used to reduce the large deviation between the projection model and the actual human body due to insufficient statistical data or small sample size, which may lead to the difficulty in finding cerebral blood vessels based on the projection model.

[0027] Optionally, the specific method for determining the partition probability corresponding to the region to be scanned based on all probability calculation indicators associated with the region to be scanned and the indicator weights corresponding to the probability calculation indicators one-to-one includes: determining the final weight based on the sum of the indicator weights corresponding to the region to be scanned; calculating the proportion of the indicator weights corresponding to the probability calculation indicators in the final weights, and updating the indicator weights corresponding to the probability calculation indicators based on the calculation results; determining the indicator calculation results based on the product of the probability calculation indicators and the corresponding indicator weights, and determining the partition probability based on the sum of the calculation results of all indicators associated with the partition to be scanned.

[0028] By adopting the above technical solution and utilizing multiple probability calculation indicators and their corresponding weights, the partition probability can be comprehensively calculated from multiple dimensions, such as the predicted area occupied by blood vessels, the degree of model matching, and the distribution of blood vessel location associations. This allows the partition probability to reflect the values ​​corresponding to each probability calculation indicator, making the calculation of the partition probability more comprehensive and representative, so that subsequent algorithms can complete scanning more quickly, accurately, and stably.

[0029] The second objective of this application is to provide a three-dimensional intracranial cerebral blood flow scanning method, which has the characteristic of improving scanning efficiency.

[0030] The second objective of this invention is achieved through the following technical solution:

[0031] The intracranial three-dimensional cerebral blood flow scanning method includes any of the scanning path planning methods applied to intracranial three-dimensional cerebral blood flow scanning as described above. After sending the current scanning path to the scanning execution module, the intracranial three-dimensional cerebral blood flow scanning method further includes: based on the scanning results corresponding to the area to be scanned, sequentially determining whether there are cerebral blood vessels in the area to be scanned, returning the predicted distribution area based on the projection model according to the determination results, determining the scanning path, and updating the scanning path.

[0032] The third objective of this application is to provide a scanning path planning system for intracranial three-dimensional cerebral blood flow scanning, which has the characteristic of improving scanning efficiency.

[0033] The third objective of this invention is achieved through the following technical solution:

[0034] The data acquisition module is used to determine the theoretical model and the actual model;

[0035] The model mapping module is used to map the cerebral vascular reference region in the theoretical model to the actual model to determine the projection model; wherein, the projection model has a predicted distribution region corresponding to this cerebral vascular region;

[0036] The algorithm scheduling module is used to determine the scanning path based on the predicted distribution area in the projection model;

[0037] The path sending module is used to send the current scan path to the scan execution module.

[0038] The fourth objective of this application is to provide a computer storage medium capable of storing corresponding programs, which has the characteristic of improving scanning efficiency.

[0039] The fourth objective of this invention is achieved through the following technical solution:

[0040] A computer-readable storage medium storing a computer program that can be loaded by a processor and executed by any of the above-described scanning path planning methods for intracranial three-dimensional cerebral blood flow scanning. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the intracranial three-dimensional cerebral blood flow scanning method of this application.

[0042] Figure 2 It is a schematic diagram of the theoretical model, the actual model, the projection model, and the cerebral blood vessel model.

[0043] Figure 3 This is a schematic diagram of a sub-process of the intracranial three-dimensional cerebral blood flow scanning method of this application.

[0044] Figure 4 This is a schematic diagram of generating a scanning path by combining a partition probability matrix and a robotic arm scheduling algorithm, where the number of regions to be scanned in the partition probability matrix is ​​9.

[0045] Figure 5 This is a schematic diagram of generating a scanning path by combining a partition probability matrix and a robotic arm scheduling algorithm, where the number of regions to be scanned in the partition probability matrix is ​​16.

[0046] Figure 6This is a schematic diagram of the sub-process of step S3 in the intracranial three-dimensional cerebral blood flow scanning method.

[0047] Figure 7 This is a diagram illustrating the updating of partition levels in the area to be scanned.

[0048] Figure 8 This is a schematic diagram of step S32 in the intracranial three-dimensional cerebral blood flow scanning method.

[0049] Figure 9 This is a schematic diagram of the overall process of intracranial three-dimensional cerebral blood flow scanning.

[0050] Figure 10 This is a schematic diagram illustrating the theoretical existence probability of the region to be scanned. The number of regions to be scanned in the diagram is 9.

[0051] Figure 11 This is a schematic diagram illustrating the theoretical existence probability of the region to be scanned. The number of regions to be scanned in the diagram is 16.

[0052] Figure 12 This is a schematic diagram illustrating the calculation of the scanning completion rate for the area to be scanned.

[0053] Figure 13 This is a schematic diagram illustrating the calculation of vascular correlation in the area to be scanned.

[0054] Figure 14 This is a schematic diagram of the scanning path planning system for intracranial three-dimensional cerebral blood flow scanning according to this application.

[0055] Figure 15 This is a schematic diagram of the modules of the intracranial three-dimensional cerebral blood flow scanning device of this application.

[0056] In the diagram, 1 is the data acquisition module; 2 is the model mapping module; 3 is the algorithm scheduling module; 4 is the path sending module; and 5 is the analysis and processing module. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0058] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the related objects before and after it have an "or" relationship.

[0059] Furthermore, the labels for each step in this embodiment are for illustrative purposes only and do not represent a limitation on the execution order of each step. In practical applications, the execution order of each step can be adjusted or performed simultaneously as needed, and such adjustments or substitutions are all within the protection scope of this invention.

[0060] The following is in conjunction with the instruction manual appendix. Figures 1-15 The embodiments of this application will be described in further detail.

[0061] This application provides a method for intracranial three-dimensional cerebral blood flow scanning. By combining preset historical data and real-time data from the site, progressively detailed scanning results are obtained during the scanning process, and the scanning path is continuously optimized. This enables the location of cerebral blood vessels to be found more accurately and quickly throughout the entire intracranial three-dimensional cerebral blood flow scanning process, reducing the overall workload of the scanning operation and improving the overall scanning rate, so that subsequent three-dimensional imaging of cranial cerebral blood vessels can be performed quickly.

[0062] Reference Figure 1 The main process of intracranial three-dimensional cerebral blood flow scanning is described as follows:

[0063] S1. Obtain input data and input model.

[0064] The input data includes environmental information, scanning data, and historical scanning data.

[0065] Environmental information refers to relevant parameters of the scanning environment, including the location to be scanned (e.g., Willis ring), probe type (e.g., single-element / ring array probe), current probe position, and current temporal window position. Specifically, the location to be scanned indicates the area to be scanned for the current person, which can be used in subsequent algorithms to locate the area to be scanned in the 3D model. The probe type and current probe position are needed in subsequent algorithms to calculate the coordinate data of the probe motion trajectory; different probe types may have different corresponding motion trajectories. Scanning data includes the real-time scan results obtained during this intracranial 3D cerebral blood flow scan. The current window position refers to the location of the temporal window and / or occipital window of the person. The current window position can be used as a reference for the probe's scanning position because the probe needs to emit ultrasound waves into the intracranial cavity from the temporal window and / or occipital window positions, otherwise it is difficult to obtain a clear ultrasound echo signal. Historical scan data refers to the results of the current person's previous scans. If this person has not undergone a three-dimensional intracranial blood flow scan for the first time, the system has already recorded the results and data of the person's previous three-dimensional intracranial blood flow scan, and the corresponding historical scan data can be obtained. If the system has not recorded any results and data of the person's previous three-dimensional intracranial blood flow scan, the corresponding historical scan data cannot be obtained.

[0066] The input model includes a theoretical model and a practical model. The theoretical model is a three-dimensional cerebral blood flow model that reflects the skeletal shape of the human skull and the distribution of cerebral blood vessels. It provides a theoretical reference for scanning path planning. Specifically, if historical scan data is available for the current person, the three-dimensional cerebral blood flow model built based on that data is directly used as the theoretical model. If no historical scan data is available, a three-dimensional cerebral blood flow model built based on a large amount of reference data is used as the theoretical model. The practical model is a three-dimensional model of the current person's skull obtained through detection.

[0067] Specifically, methods for constructing theoretical models based on a large amount of reference data include:

[0068] 1) Obtain vascular reference data and construct a theoretical cerebral vascular model based on the vascular reference data.

[0069] Vascular reference data is data obtained by scanning different people using intracranial three-dimensional cerebral blood flow scanning methods with relevant technologies. Through vascular reference data, the distribution of cerebral blood vessels in the skull of different people can be statistically analyzed, thereby constructing a theoretical cerebral vascular model that has reference value and can reflect the distribution of cerebral blood vessels.

[0070] 2) Obtain skeletal reference data and construct a theoretical skull model based on the skeletal reference data.

[0071] Skeletal reference data consists of different human cranial structure data corresponding to vascular reference data. Through skeletal reference data, different human cranial structure shapes can be statistically analyzed, thereby constructing a theoretical skull model with reference value that can reflect the shape of the cranial structure.

[0072] 3) Based on the theoretical cerebral vascular model and the theoretical skull model, the theoretical model is determined.

[0073] By combining theoretical cerebral vascular models and theoretical skull models, a theoretical model can be constructed that can simultaneously reflect the shape of the skull (i.e., the shape of the external skull bones) and the distribution of cerebral blood vessels.

[0074] S2. Map the cerebral vascular reference region in the theoretical model to the actual model to determine the projection model.

[0075] The cerebral vascular reference region refers to the area in the theoretical model where cerebral blood vessels are distributed, used to indicate the location of cerebral blood vessels. In this embodiment, the cerebral blood vessels are composed of a series of three-dimensional coordinate points, and the location of the cerebral blood vessels can be represented by three-dimensional coordinate point cloud data.

[0076] A projection model is a model derived by combining the actual skull shape with a cerebral vascular reference area. Constructing a projection model is essentially the process of mapping the cerebral vascular distribution in the theoretical model to the actual skull shape. However, skull shapes vary among different patients; therefore, the deviation between the theoretical and actual skull shapes must be considered during the construction of the projection model.

[0077] Reference Figure 2 and Figure 3 In step S2, the following is included:

[0078] S21. Determine the cranial deviation value based on the difference between the cranial shape corresponding to the actual model and the cranial shape corresponding to the theoretical model.

[0079] The methods for determining cranial deviation values ​​include, but are not limited to: calculating the coordinate difference between the temporal window position of the actual model and the temporal window position of the theoretical model, calculating the coordinate difference between the occipital window position of the actual model and the occipital window position of the theoretical model, calculating the difference between the cranial length of the actual model and the cranial length of the theoretical model, and calculating the difference between the cranial width of the actual model and the cranial width of the theoretical model.

[0080] When the system cannot construct a theoretical model based on historical scan data, the theoretical model can represent the skull shape of a normal person, and the skull deviation value can represent the skull deviation between the current person and a normal person. For example, if the height difference between the temporal window position in the actual model and the temporal window position in the theoretical model is 0.5 cm, then it can be considered that the temporal window position of the current person is shifted upward by 0.5 cm compared to the temporal window position of a normal person. Based on the above deviation, the theoretical model can be adjusted to make the theoretical model closer to the actual skull shape of the current person.

[0081] S22. Adjust the cerebral vascular reference region based on the cranial deviation value, and map the adjusted cerebral vascular reference region onto the actual model to determine the projection model.

[0082] Based on the cranial deviation value, three-dimensional image transformations (such as scaling up, scaling down, stretching, shrinking, translation, rotation, etc.) are performed on the theoretical model and the cerebral vascular reference region to ensure that the shape and positional differences between the cranial shape corresponding to the theoretical model and the cranial shape corresponding to the actual model meet the error requirements. It can be considered that the cranial shape corresponding to the theoretical model and the cranial shape corresponding to the actual model have a very high similarity. Therefore, the adjusted cerebral vascular reference region of the theoretical model can be used as the reference region for the distribution of cerebral blood vessels in the actual model.

[0083] If the system cannot construct a theoretical model based on historical scan data, for example, if Model A is a theoretical model constructed based on the skull shape of person a, whose face is oval and whose cerebral blood vessels are located at position P1; and Model B is an actual model constructed based on the skull shape of patient b, whose face is also oval, but whose face is larger, meaning that the skull shape of Model B deviates from that of Model A, then:

[0084] Based on the cranial deviation values ​​of models A and B, the position P1(r1, h1, v1) is mapped to model B to obtain the position P2(r2, h2, v2). The resulting model C is the projection model corresponding to patient b. Here, r, h, and v are all coordinate values.

[0085] It is worth noting that positions P1 and P2 mentioned above are illustrative, and the values ​​of positions P1 and P2 are both the center point of the blood vessel. In some feasible embodiments, data reflecting the morphology of the blood vessels, such as the curvature, radius of curvature, and coordinates of each point on the surface of the blood vessels, can be added to provide a more accurate description of the location of the blood vessels.

[0086] S3. Determine the scanning path based on the predicted distribution area in the projection model.

[0087] The predicted distribution area refers to the cerebral blood vessel distribution presented in the projection model through mapping in step S2. It reflects areas with a high probability of cerebral blood vessels appearing in subsequent actual scans, and a predicted cerebral blood vessel path can be formed based on the predicted distribution area. The scanning path guides the probe of the scanning execution module and indicates the probe's movement trajectory. In this embodiment, the scanning execution module includes a probe and a robotic arm that drives the probe's movement. The scanning path should be planned based on the predicted distribution area so that the scanning execution module can quickly scan the patient's actual cerebral blood vessel locations using the probe.

[0088] Reference Figure 3 and Figure 4 In step S3, the following is included:

[0089] S31. Model Partitioning: Based on the projection model, the model is partitioned to determine multiple areas to be scanned.

[0090] The scanning area of ​​the projection model is divided into multiple regions of the same size but located at different positions. Since the area that can be scanned within a single time interval of the probe is small, it is necessary to partition the entire scanning range. In subsequent steps, the probability of cerebral blood vessels being present in each scanning region can be calculated independently, and the scanning region with the highest probability of cerebral blood vessels is prioritized to construct a scanning path with a higher scanning rate and more accurate blood vessel localization.

[0091] In this embodiment, the model partitioning step involves first projecting the projection model from a 3D model onto a 2D plane using a projection method. Then, the 2D plane is used to divide the areas to be scanned. Projection methods include, but are not limited to, central projection, parallel projection, and Mercator projection. In other embodiments, 3D partitioning can be performed directly in 3D without using a projection method.

[0092] S32. Probability Calculation: Based on the predicted distribution area and each area to be scanned, determine the partition probability corresponding to each area to be scanned.

[0093] The partition probability corresponds one-to-one with the area to be scanned. The partition probability refers to the estimated probability of cerebral blood vessels appearing in the area to be scanned. Since the predicted distribution area is the area of ​​the estimated distribution of cerebral blood vessels, the partition probability corresponding to each area to be scanned can be calculated based on the correlation between the predicted distribution area and each area to be scanned.

[0094] As shown in the figure, the projection module divides the area into 9 regions to be scanned, namely region z_0, region z_1, ..., region z_8.

[0095] S33. Determine the scanning path based on the partition probability.

[0096] By using partition probabilities, the priority of each region to be scanned during scanning can be determined; the higher the partition probability, the higher the priority of the region to be scanned. The scanning path refers to the path traversed by the probe during scanning; the scanning path is actually a data model of the probe's motion trajectory.

[0097] Step S33 includes:

[0098] S331. Generate a partition probability matrix based on the partition probability and corresponding position of each region to be scanned.

[0099] S332. Determine the scanning path based on the partition probability matrix and the robotic arm scheduling algorithm.

[0100] In this embodiment, the probe moves via a robotic arm. The robotic arm scheduling algorithm can fully consider factors such as the lifespan of the robotic arm and its effective movement efficiency. The robotic arm scheduling algorithm includes, but is not limited to, the shortest addressing algorithm and the elevator scheduling algorithm.

[0101] By combining the partition probability matrix with the robotic arm scheduling algorithm, a more efficient and executable probe motion trajectory data model can be quickly generated. For example, if the predicted cerebrovascular path passes through regions z_0, z_3, z_4, z_7, and z_8, the scanning path should also pass through these regions. However, due to the limited scanning range of the probe in a single scan, and the area to be scanned not yet meeting the conditions for shrinkage, the probe needs to perform two scans to complete the full scan. The final scanning path will first pass through region z_0 → region z_3 → region z_4, then move to region z_7 → region z_8 via region z_6. During the probe's movement from region z_6 to region z_7, half of region z_6 is scanned incidentally, expanding the scan data, improving the effective movement of the robotic arm, and providing reference data for the next iteration planning.

[0102] Reference Figure 5 On the other hand, the robotic arm scheduling algorithm can plan the scanning path based on the current position of the probe. For example, if the probe is located in region z_2, and the predicted cerebrovascular path in the distribution area passes through regions z_0, z_4, z_8, z_9, z_13, z_14, and z_15, an elevator scheduling algorithm can be used to make the robotic arm move in one direction as a whole, reducing repetitive back-and-forth movements. Then, the shortest addressing algorithm can be combined to make the probe move to the target scanning area as quickly as possible. The final scanning path is: region z_2 → region z_1 → region z_0 → region z_4 → region z_8 → region z_9 → region z_13 → region z_14 → region z_15.

[0103] Reference Figure 3 S4. Send the current scan path to the scan execution module.

[0104] The scanning execution module drives the probe to move according to the scanning path and completes the scan. During this process, the scanning execution module continuously uploads the scan results, so the system can obtain the scan results of each area to be scanned in sequence.

[0105] In summary, steps S1-S4 constitute the scanning path planning method in the intracranial three-dimensional cerebral blood flow scanning method. The purpose is to plan the scanning path and output it to the scanning execution module.

[0106] S5. Output a scan report based on the scan results from the scan path.

[0107] In this embodiment, the scanning execution module continuously and in real time uploads the scanning results during the scanning process, and the system constructs a cerebral vascular model in real time based on the scanned data.

[0108] Once all regions along the scanning path have been scanned, the intracranial three-dimensional cerebral blood flow scan is complete and a scan report is generated, which can produce and display a complete three-dimensional cerebral blood flow data model.

[0109] In one specific embodiment, step S5 includes:

[0110] S51. Based on the scanning results of the scanned area, determine whether there are cerebral blood vessels in the scanned area. If so, proceed to S52; otherwise, proceed to S53.

[0111] During the scanning process based on the scanning path, the scanning execution module generates and uploads the corresponding scanning result for each area to be scanned after passing through it. Therefore, the system can receive the scanning results of each area to be scanned successively and determine whether there are cerebral blood vessels in each area to be scanned.

[0112] S52. Send a pause command to the scan execution module, store the scan results, and return to step S3.

[0113] Upon receiving a pause command, the scanning execution module temporarily halts the scanning of the current scan path. The presence of cerebral blood vessels in the scanned area can affect the probability distribution of other scanned regions, necessitating a replanning of the scan path. Therefore, scanning of the current path must be paused while awaiting an updated scan path.

[0114] S53. Save the results of this scan.

[0115] After the scanning execution module completes the scanning of all locations along the scanning path, the system outputs a scanning report based on all stored scanning results.

[0116] Reference Figure 6 and Figure 7 In one specific embodiment, step S31 includes:

[0117] S311. Based on the current partitioning level, partition the projection model to determine multiple areas to be scanned.

[0118] The partition level determines the proportion of a single area to be scanned in the projection model. The partition level is dynamically variable, so that the extent of the area to be scanned also adjusts dynamically. The size of the area to be scanned corresponding to the partition level is a preset value.

[0119] If the partition levels include level 1, level 2, level 3... level n, then as the partition level increases, the total number of areas to be scanned will increase, while the size of a single area to be scanned will decrease. For example, at level 1, the projection model divides the area into 4 areas to be scanned; at level 2, the projection model divides the area into 9 areas to be scanned; and at level 3, the projection model divides the area into 16 areas to be scanned.

[0120] S312. Determine whether the current area to be scanned meets the partitioning iteration conditions, update the partitioning level according to the determination result, and return to step S311.

[0121] The size of the area to be scanned is dynamically adjusted. It changes when the area meets the partitioning iteration conditions, and the change in the size of the area to be scanned is based on the change in partitioning level. In this embodiment, the size of the area to be scanned decreases each time the partitioning iteration condition is triggered. During the intracranial three-dimensional brain blood flow scanning process, the size of the area to be scanned becomes smaller and smaller with multiple iterations of the scanning partitions.

[0122] The partitioning iteration conditions specifically include range triggering conditions and blood vessel triggering conditions. Satisfying either one of these conditions satisfies the partitioning iteration conditions.

[0123] The range trigger condition is: the area of ​​the current single area to be scanned is greater than the preset maximum range threshold. The value of the maximum range threshold is related to the single scan range of the probe in the scanning execution module. When the area to be scanned is greater than the preset maximum range threshold, it indicates that the current partition level is too small, resulting in an excessively large area to be scanned, making it difficult for the probe to complete the scan in time. Therefore, it is necessary to increase the partition level to reduce the area to be scanned.

[0124] The vascular trigger condition is as follows: cerebral blood vessels are detected in the scan results of the previous scan area, and the area of ​​the current scan area is greater than a preset minimum range threshold. The minimum range threshold is less than the maximum range threshold, and its value is related to the single scan range of the probe in the scanning execution module. Once cerebral blood vessels are detected in the previous scan area, it indicates that the probability of encountering cerebral blood vessels in subsequent scans increases as the scan path progresses. Therefore, a more refined division of the scan area is needed to collect more scan data and reduce the risk of missed scans.

[0125] If the area of ​​the current area to be scanned is greater than the preset minimum range threshold, it means that the area of ​​the current area to be scanned is moderate and will not affect the single scan of the probe. If the area of ​​the current area to be scanned is less than or equal to the minimum range threshold, it means that the area of ​​the current area to be scanned is too small. The probe may scan multiple areas to be scanned in a single scan. Therefore, the next iteration of the area to be scanned will only occur when the area of ​​the current area to be scanned is greater than the minimum range threshold.

[0126] During intracranial three-dimensional cerebral blood flow scanning, the area to be scanned undergoes multiple iterations, forming a smaller scanning region. Comparing the changes in the scanning region before and after iterations, before iterations (or in the early stages of intracranial three-dimensional cerebral blood flow scanning), the amount of observable data is smaller, and the area to be scanned is larger, which is beneficial for the probe to quickly locate the actual cerebral blood vessels. After iterations (or in the later stages of intracranial three-dimensional cerebral blood flow scanning), the amount of observable data is larger, and the area to be scanned is smaller, which is beneficial for the probe to more comprehensively locate the actual cerebral blood vessels.

[0127] On the other hand, the area of ​​the region to be scanned is limited to between the minimum range threshold and the maximum range threshold to limit the iteration of the region to be scanned, so that the area of ​​the region to be scanned can always adapt to the single scan range of the probe.

[0128] Specifically, the method for updating the partition level is to increase the level of the current partition level. As shown in Figure (7a), the entire projection model is divided into 4 regions to be scanned. If it is determined that the area of ​​the region to be scanned is too large and meets the range triggering condition, the partition level is updated and the entire projection model is re-divided into 9 regions to be scanned. The division result is shown in Figure (7b).

[0129] As shown in Figure (7b), the entire projection model is divided into 9 regions to be scanned. If cerebral blood vessels are found in the scanning results of the regions to be scanned in the previous round, and it is determined that the area of ​​the region to be scanned is greater than the minimum range threshold, thus meeting the blood vessel triggering condition, the partition level is updated, and the entire projection model is re-divided into 16 regions to be scanned. The division result is shown in Figure (7c).

[0130] As shown in Figure (7c), the entire projection model is divided into 16 regions to be scanned. If cerebral blood vessels are still found in the scanning results of the regions to be scanned in the previous round, but it is determined that the area of ​​the region to be scanned is less than or equal to the minimum range threshold and does not meet the blood vessel triggering condition, then the partition level will not be updated and the entire projection model will still have 16 regions to be scanned.

[0131] Reference Figure 8 and Figure 9 In one specific embodiment, step S32 includes:

[0132] S321. Determine at least one probability calculation index associated with the region to be scanned and the index weights corresponding to the probability calculation indexes.

[0133] The partition probability of each region to be scanned is calculated using probability calculation metrics and metric weights. There are various probability calculation metrics, each with a different calculation method; the metric weights correspond one-to-one with the probability calculation metrics. The correlation between the region to be scanned and the probability calculation metrics lies in whether the region to be scanned meets the conditions for calculating the probability calculation metrics.

[0134] In this embodiment, the probability calculation indicators include two Level 1 indicators, each of which includes multiple Level 2 indicators. Correspondingly, the indicator weights include two Level 1 indicators, each of which includes multiple Level 2 weights. See Table 1 – Partition Probability Calculation Weight Allocation Table for details. The Level 1 indicators include model trend indicators and image information indicators. The model trend indicators include three Level 2 indicators: theoretical existence probability, model matching degree, and scan completion degree. The image information indicators include two Level 2 indicators: image matching degree and vascular correlation degree. The indicator weights include: theoretical existence weight (initial value 25%) corresponding to the theoretical existence probability; model matching weight (initial value 20%) corresponding to the model matching degree; scan completion weight (initial value 10%) corresponding to the scan completion degree; image matching weight (initial value 25%) corresponding to the image matching degree; and vascular correlation weight (initial value 20%) corresponding to the vascular correlation degree.

[0135]

[0136] Table 1 - Weight Allocation Table for Partition Probability Calculation

[0137] Among the various probability calculation indicators mentioned above, the theoretical existence probability is the most basic probability calculation indicator. The calculation of the theoretical existence probability only requires predicting the location of the distribution area and the location of the area to be scanned. Therefore, the area to be scanned must meet the calculation conditions of the theoretical existence probability, and the area to be scanned must be associated with at least the theoretical existence probability.

[0138] Model matching degree requires calculating the matching degree between the cerebral blood vessel morphology in the predicted distribution area and the actual cerebral blood vessel morphology of the current patient. Therefore, model matching degree needs to be calculated based on the cerebral blood vessels that have already been scanned. Before the cerebral blood vessels are scanned, the area to be scanned can meet the calculation conditions of model matching degree, and the area to be scanned can be associated with model matching degree.

[0139] The scan completion rate needs to calculate the actual length of the patient's brain blood vessels. Therefore, the scan completion rate needs to be calculated based on the brain blood vessels that have already been scanned. Before the brain blood vessels are scanned, the scan area can meet the calculation conditions for the scan completion rate, and the scan area can be associated with the scan completion rate.

[0140] Image matching degree requires calculating the matching degree between the cerebral blood vessel morphology in the theoretical scan image and the actual cerebral blood vessel morphology of the current patient. Therefore, image matching degree needs to be calculated based on the cerebral blood vessels that have already been scanned. Before the cerebral blood vessels are scanned, the scanned partitions can meet the calculation conditions for image matching degree, and only then can image matching degree be associated with the scan completion degree.

[0141] Vascular correlation needs to be based on the actual location of the patient's brain blood vessels. Therefore, vascular correlation needs to be calculated based on the brain blood vessels that have already been scanned. Before the brain blood vessels are scanned, the area to be scanned can meet the calculation conditions for vascular correlation, and only then can the area to be scanned be associated with vascular correlation.

[0142] On the other hand, the probability calculation index settings for the area to be scanned can also be dynamically selected and switched by the user to limit the way the partition probability of the area to be scanned is calculated.

[0143] S322. Calculate all probability metrics associated with the region to be scanned.

[0144] Among them, the theoretical existence probability, model matching degree, scan completion degree, image matching degree, and blood vessel correlation degree all have different calculation methods. The following are specific examples.

[0145] The purpose of the theoretical probability is to combine the predicted distribution area and the area to be scanned, and to select the area with a higher probability of cerebral blood vessels from among the areas to be scanned, and to scan it first.

[0146] Reference Figure 10 Methods for calculating the theoretical existence probability associated with the region to be scanned include:

[0147] 1) Calculate the theoretical existence probability associated with the region to be scanned based on the ratio between the portion of the predicted distribution region that overlaps with the region to be scanned and the whole region to be scanned.

[0148] The greater the overlap between the area to be scanned and the predicted distribution area, i.e., the larger the area of ​​the predicted distribution area occupied by the area to be scanned, the higher the theoretical existence probability of the area to be scanned. In this embodiment, the area to be scanned is a two-dimensional planar area, and the calculation of the theoretical existence probability is based on the area ratio; in other embodiments, if the area to be scanned is a three-dimensional solid area, the calculation of the theoretical existence probability requires a reference volume occupancy.

[0149] 2) Based on the comparison of theoretical existence probability and probability classification coefficient, all regions to be scanned are classified to determine low-probability partitions and high-probability partitions.

[0150] In this embodiment, the scanned regions with a theoretical existence probability greater than or equal to the probability grading coefficient are classified as high-probability partitions, and the scanned regions with a theoretical existence probability lower than the probability grading coefficient are also classified as high-probability partitions. In this embodiment, the probability grading coefficient is a preset value set at 70%; in other embodiments, the probability grading coefficient can also be obtained by equally dividing the maximum and minimum theoretical existence probabilities.

[0151] 3) Based on the probability increase coefficient, update the theoretical existence probability of all low-probability partitions that satisfy the blood vessel connectivity condition.

[0152] The blood vessel connectivity condition is that a low-probability partition is located between two high-probability partitions that meet the requirements. The probability amplification coefficient is a preset value; in this embodiment, the probability amplification coefficient is 0.2. The theoretical existence probability is updated by increasing the theoretical existence probability by the proportion corresponding to the probability amplification coefficient. For example, if the probability amplification coefficient is 0.2, the theoretical existence probability is updated by increasing it by 20%.

[0153] In actual scanning, some cerebral blood vessels may be relatively narrow, resulting in a smaller predicted distribution area corresponding to these vessels. However, theoretically, there must be a blood vessel between two adjacent cerebral blood vessel locations; therefore, the probability of the area located between them needs to be increased accordingly.

[0154] If the predicted distribution area in region z_4 occupies a small area, based solely on the area occupied, the original theoretical probability of region z_4 is 65%. Since 65% is less than 70%, region z_4 is a low-probability partition. However, regions z_3 and z_5, located on either side of region z_4, are both high-probability partitions. Therefore, region z_4 satisfies the vascular connectivity condition, and its theoretical probability is updated from the original 65% to 65%*(100%+20%) = 78%.

[0155] It is worth noting that the update of the theoretical existence probability is not recursive. That is, the high-probability partition in the vascular connectivity condition does not include the scanned region whose theoretical existence probability has been adjusted due to the probability amplification factor, in order to avoid causing a chain reaction.

[0156] Reference Figure 11 Suppose that region z_9, due to satisfying the vascular connectivity condition, has its theoretical existence probability increased by 20%, meaning the adjusted theoretical existence probability is 65%*(1+20%) = 78%. At this point, although the theoretical existence probability of the scanned regions on either side of region z_5 (regions z_1 and z_9) is greater than 70%, region z_5 does not satisfy the vascular connectivity condition. Conversely, if the probability is recursively adjusted to satisfy the vascular connectivity condition for region z_5, making its theoretical existence probability 60%*(1+20%) = 72%, then region z_5 will also become a high-probability region, causing a location with a low probability of cerebral blood vessels to mistakenly become a high-probability region.

[0157] The model matching degree reflects the degree of matching between the current predicted projection model and the current patient's actual cerebrovascular model. The higher the matching degree, the higher the reliability of the projection model.

[0158] Reference Figure 2 Methods for calculating the model matching degree associated with the region to be scanned include:

[0159] 1) Construct a cerebral vascular model based on the cerebral vascular data obtained from the scanning results of the scanned area that has already been scanned.

[0160] 2) Based on the cerebral vascular model and the projection model, calculate the deviation value between the cerebral vascular model and the projection model, and determine the model matching degree based on this deviation value.

[0161] By comparing the cerebral vascular model D and the projection model C, the deviation values ​​of the corresponding vascular spatial morphology (such as the difference in vascular curvature radius, vascular volume, and vascular position) can be calculated to obtain the model matching degree. The smaller the deviation value, the higher the model matching degree.

[0162] Reference Figure 12 The purpose of scan completion is to predict the continuous cerebral blood vessels that can be obtained after the scan is completed in the area to be scanned. The longer the length of the cerebral blood vessels that can be obtained, the higher the scan completion.

[0163] Assuming that regions z_3, z_5, and z_8 have already been scanned, it is predicted that cerebral blood vessels exist in regions z_0 and z_4. This is because scanning region z_4 can connect the cerebral blood vessels in regions z_3, z_5, and z_8, compared to scanning region z_0, which only connects regions z_0 and z_3. Therefore, scanning region z_4 is considered to have a higher priority.

[0164] Methods for calculating the scan completion rate associated with the area to be scanned include:

[0165] 1) Based on the scanning results of the scanned area, determine the total number of adjacent blood vessels corresponding to the scanned area.

[0166] The sum of adjacent vessels is the sum of all cerebral vessels connected to the scanned region in the scan results, plus the predicted lengths of cerebral vessels in this scanned region. The sum of adjacent vessels is used to predict the continuous length of cerebral vessels that can be obtained after scanning this scanned region.

[0167] Based on the scan results, there are two real cerebral blood vessels, namely real blood vessel ② and real blood vessel ④. Real blood vessel ② is 1 mm long and is located in region z_3. Real blood vessel ④ is 1.8 mm long and is located in regions z_5 and z_8.

[0168] Based on the predicted distribution area, there are currently two predicted cerebral blood vessels, namely predicted vessel ① and predicted vessel ③. The length of predicted vessel ① is 1 mm and the actual vessel ① is distributed in region z_0. The length of predicted vessel ③ is 1.5 mm and the actual vessel ③ is distributed in region z_4.

[0169] The sum of the neighboring vessels in region z_0 of the area to be scanned = the length of the predicted vessel ① + the length of the actual vessel ② = 2mm. This can be understood as follows: if region z_0 is scanned first, the maximum length of the connected vessel that can be obtained is expected to be 2mm.

[0170] The total length of neighboring vessels in region z_4 within the area to be scanned = the length of the actual vessel ② + the predicted vessel ③ + the length of the actual vessel ④ = 4.3 mm. This can be understood as follows: if region z_4 is scanned first, the maximum length of the connected vessel that can be obtained is expected to be 4.3 mm. Since 4.3 mm > 2 mm, the scanning priority of region z_4 is higher than that of region z_0.

[0171] 2) Compare the sum of all neighboring vessels and determine the maximum vessel length based on the sum of the neighboring vessels with the largest sum.

[0172] In the example above, since the sum of the neighboring vessels in region z_4 is the maximum value, the maximum vessel length is 4.3 mm.

[0173] 3) Determine the scan completion rate associated with the scanned area based on the ratio of the sum of adjacent blood vessels to the length of the largest blood vessel.

[0174] The maximum blood vessel length is used as the 100% weight, and the weight of each scanned region is the ratio of the sum of the lengths of the neighboring blood vessels to the maximum blood vessel length.

[0175] As in the example above, the maximum blood vessel length is 4.3 mm, so the scan completion rate of region z_4 is 4.3 mm / 4.3 mm = 100%; the scan completion rate of region z_0 is 2 mm / 4.3 mm = 46.5%.

[0176] Image matching accuracy reflects the degree of matching between the actual ultrasound image of the current patient and the predicted ultrasound image.

[0177] Methods for calculating the image matching degree associated with the region to be scanned include:

[0178] 1) Determine the actual scanned image based on the scan results of the area to be scanned that has already been scanned.

[0179] The actual scanned image is an ultrasound image generated in real time based on the scan data uploaded by the scan execution module.

[0180] 2) Determine the theoretical scan image and calculate the difference between the cerebral blood vessel morphology in the actual scan image and the cerebral blood vessel morphology in the theoretical scan image. Based on this difference, determine the image matching degree associated with the area to be scanned.

[0181] Theoretical scan images can be obtained in two ways: first, when historical scan data of the current personnel is unavailable, they can be obtained by statistical analysis of a large amount of historical data, or by storing a large number of actual patient scan data images of different scan sites in a model library and using these as theoretical scan images; second, when historical scan data of the current personnel is available, theoretical scan images can be obtained based on historical scan data.

[0182] The ultrasound image is the ultrasound image obtained by the system in the projection model based on the current scanning position, i.e., the ultrasound image calculated from historical learning data.

[0183] If image img1 is the actual scanned image and image img2 is the theoretical scanned image, by identifying image img1, we find that the center of the cerebral blood vessels is located at p_cur(20, 60), while the theoretical image img2 requires the center of the cerebral blood vessels to be located at p_target(30, 60), a difference of (10, 0). Therefore, the image matching degree is: (1-10 / 30)*50% + (1-0 / 60)*50% = 83%. This calculation is for illustrative purposes only. Actual calculations can also refer to other vascular parameters, including but not limited to the radius of curvature of the blood vessels and blood flow velocity.

[0184] Reference Figure 13The role of vascular correlation is to radiate the influence of the actual location of cerebral blood vessels on each area to be scanned. It is mainly used to reduce the large deviation between the projection model and the actual human body due to insufficient statistical data or small sample size, which makes it difficult to find cerebral blood vessels based on the projection model.

[0185] Methods for calculating the correlation degree of blood vessels associated with the area to be scanned include:

[0186] 1) Determine the current blood vessel location based on the scanning results of the area to be scanned that has already been scanned.

[0187] 2) Based on the current vascular location and predicted distribution area, determine the vascular correlation degree corresponding to all areas to be scanned.

[0188] Specifically, the correlation between the scanned regions and the current blood vessel locations gradually decreases according to the gradient direction where the distance between the scanned region and the current blood vessel location gradually increases, and the distribution trend direction of the predicted distribution area. Specifically, the scanned region immediately adjacent to the location of the discovered cerebral blood vessel in the predicted blood vessel trend (the distribution trend of the predicted distribution area) receives the highest score, the scanned region on the non-predicted blood vessel trend (the distribution trend of the non-predicted distribution area) receives the second highest score, and the weights of other scanned regions decrease sequentially according to the distribution trend of the non-predicted distribution area.

[0189] If we assume that the current scanning region is z_0 and a strong vascular feature signal (such as a blood flow Doppler signal greater than a preset threshold) is found in this region, then we assume that there are cerebral blood vessels in region z_0.

[0190] Region z_0 has two adjacent regions to be scanned, namely regions z_1 and z_3. Since region z_3 is the adjacent region of the predicted vascular trend (the distribution trend of the predicted distribution area), 100 (the highest score) is taken as the vascular correlation score. Since region z_1 is the adjacent region of the non-predicted vascular trend (the distribution trend of the non-predicted distribution area), 90 (the second highest score) is taken as the vascular correlation score. The remaining regions to be scanned decrease in score by 10. When the gradients of one of the regions to be scanned are inconsistent from two directions, the higher score is taken. For example, the value of region z_4 calculated based on the direction of region z_1 is 80, and the value calculated based on the direction of region z_3 is 90. Since 90 > 80, 90 is taken as the vascular correlation score.

[0191] Reference Figure 8 S323. Based on all probability calculation indicators associated with the region to be scanned and the indicator weights corresponding to the probability calculation indicators, determine the partition probability corresponding to the region to be scanned.

[0192] The number of probability calculation indicators associated with the region to be scanned is at least 1, and the number of indicator weights is consistent with the number of probability calculation indicators.

[0193] Step S323 includes:

[0194] S3231. Determine the final weight based on the sum of the indicator weights corresponding to the area to be scanned.

[0195] In case A, if only the theoretical existence probability needs to be calculated for the area to be scanned, then there is only one probability calculation index, which is the theoretical existence probability. Therefore, the final weight is the theoretical existence weight = 25%.

[0196] In case B, if the theoretical existence probability and model matching degree of the region to be scanned need to be calculated, the final weight is the sum of the theoretical existence weight and the model matching weight = 25% + 20% = 45%.

[0197] In case C, if the area to be scanned meets the calculation conditions of all probability calculation indicators, then there are 5 probability calculation indicators, and the final weight is the sum of the probability calculation indicators = 25% + 20% + 10% + 25% + 20% = 100%.

[0198] S3232. Calculate the proportion of the indicator weight corresponding to the probability calculation indicator in the final weight, and update the indicator weight corresponding to the probability calculation indicator based on the calculation result.

[0199] In case A, the probability calculation index only includes the theoretical probability of existence. In this case, the calculated proportion is 25% / 25% = 100%, so the weight of theoretical existence is 100%.

[0200] In case B, the probability calculation indicators include: theoretical existence probability and model matching degree. The theoretical existence weight is 25% / 45% = 55.6%, and the model matching weight is 20% / 45% = 44.4%.

[0201] In case C, the probability calculation indicators include: theoretical existence probability, model matching degree, scan completion degree, image matching degree, and vascular association degree. In this case, the weights of theoretical existence, model matching, completion degree, image matching, and vascular association all remain at their original weight values.

[0202] S3233. Based on the product of the probability calculation index and the corresponding index weight, determine the index calculation result, and based on the sum of the calculation results of all the indexes associated with the partition to be scanned, determine the partition probability.

[0203] In case A, there is only one result for the index calculation of the area to be scanned, and the partition probability = probability calculation index * 100%;

[0204] In case B, there are two results for calculating the index of the area to be scanned: partition probability = probability calculation index * 55.6% + model matching degree * 44.4%.

[0205] In case C, there are five possible results for calculating the indicators of the area to be scanned: partition probability = theoretical existence probability * 25% + model matching degree * 20% + scan completion degree * 10% + image matching degree * 25% + blood vessel correlation degree * 20%.

[0206] It is worth noting that the above-mentioned probability calculation indicators (theoretical existence probability, model matching degree, scan completion degree, image matching degree, and vascular correlation degree) are not a limitation on all possible types of probability calculation indicators. The embodiments of this application provide a preferred solution, which divides the calculation into theoretical existence probability, model matching degree, scan completion degree, image matching degree, and vascular correlation degree based on model trend indicators and image information indicators. As data accumulates and computing power improves, users can also increase the types of probability calculation indicators and corresponding calculation methods according to actual needs.

[0207] Correspondingly, the specific values ​​of the weights of the above-mentioned indicators are all the preferred values ​​in the embodiments of this application. In practical applications, users can adjust the weights of each indicator according to the actual needs of the current scanning project and the actual situation of the personnel. Furthermore, during the scanning process, as data accumulates, the weights of each indicator can also change dynamically. For example, as the scanning progresses, if it is found that the data obtained from the scanning is highly consistent with the data of the theoretical model in the calculation of a certain probability calculation indicator, then the weight of this probability calculation indicator will be increased accordingly, while the weights of other indicators will be reduced.

[0208] The implementation principle of the intracranial three-dimensional cerebral blood flow scanning method disclosed in this application is as follows: by combining theoretical and actual models, the regions with a high probability of cerebral blood vessels appearing in the prediction are represented in the projection model through the predicted distribution region. By constructing the scanning path through the predicted distribution region, the probability of scanning the real cerebral blood vessels at the locations traversed by the scanning path is higher, so that the probe of the scanning execution module can quickly and accurately scan the location of cerebral blood vessels, thereby improving scanning efficiency and enhancing the efficiency of subsequent algorithms in constructing cerebral blood vessel models.

[0209] During the scanning path planning process, each region to be scanned participates in the planning through its own partition probability. The probability calculation indicators for partition probability are varied, including theoretical existence probability, model matching degree, scan completion degree, image matching degree, and vascular correlation degree. Each probability calculation indicator has different calculation methods and different indicator weights. Partition probability can be comprehensively calculated from multiple dimensions such as predicted vascular area, model matching degree, and vascular location correlation distribution, so that the partition probability can reflect the values ​​corresponding to each probability calculation indicator. It has the characteristics of: reflecting the matching degree between the currently predicted projection model and the current patient's actual cerebral vascular model; predicting the continuous cerebral vascular length that can be obtained after scanning the region to be scanned; reflecting the matching degree between the current patient's actual ultrasound image and the predicted ultrasound image; and radiating influence on each region to be scanned based on the actual location of cerebral blood vessels. This makes the calculation of partition probability more comprehensive and representative, enabling subsequent algorithms to complete scans more quickly, accurately, and stably.

[0210] This application also provides a scanning path planning system for intracranial three-dimensional cerebral blood flow scanning, which corresponds one-to-one with the scanning path planning method in the above embodiments. (Refer to...) Figure 14 The scanning path planning system includes:

[0211] The data acquisition module 1 includes an information acquisition submodule for acquiring input data and a model imaging submodule for acquiring the input model. The data acquisition module 1 sends the theoretical model and the actual model to the model mapping module 2 based on the input model.

[0212] Model mapping module 2 is used to map the cerebral vascular reference region in the theoretical model to the actual model, determine and send the projection model to algorithm scheduling module 3;

[0213] Algorithm scheduling module 3 is used to determine and send the scanning path to path sending module 4 based on the predicted distribution area in the projection model;

[0214] Path sending module 4 is used to send the current scan path to the scan execution module.

[0215] The scanning path planning system for intracranial three-dimensional cerebral blood flow scanning provided in this embodiment can realize each step of the scanning path planning method in the aforementioned embodiment due to the functions of each module and the logical connection between them. Therefore, it can achieve the same technical effect as the aforementioned embodiment. The principle analysis can be found in the relevant description of the aforementioned method steps, which will not be repeated here.

[0216] This application also provides an intracranial three-dimensional cerebral blood flow scanning device, which corresponds one-to-one with the intracranial three-dimensional cerebral blood flow scanning method in the above embodiments, and includes each module of the above-described scanning path planning system. (Refer to...) Figure 15 The intracranial three-dimensional cerebral blood flow scanning device also includes:

[0217] The analysis and processing module 5 is used to receive the scanning results from the scanning execution module, and based on the scanning results corresponding to the area to be scanned, to determine whether there are cerebral blood vessels in the area to be scanned. According to the determination results, the scanning results are sent to the algorithm scheduling module 3 to update the scanning path.

[0218] The intracranial three-dimensional cerebral blood flow scanning device provided in this embodiment can realize each step of the intracranial three-dimensional cerebral blood flow scanning method in the aforementioned embodiment due to the functions of each module and the logical connection between them. Therefore, it can achieve the same technical effect as the aforementioned embodiment. The principle analysis can be found in the relevant description of the aforementioned method steps, which will not be repeated here.

[0219] This application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed by the above-described scanning path planning method for intracranial three-dimensional cerebral blood flow scanning. When the computer program is executed by the processor, it implements the steps of the above-described intracranial three-dimensional cerebral blood flow scanning method.

[0220] The readable storage medium provided in this embodiment can achieve the same technical effect as the aforementioned embodiments because the computer program therein, after being loaded and run on the processor, will implement the various steps of the aforementioned embodiments. For the principle analysis, please refer to the relevant description of the aforementioned method steps, which will not be repeated here.

[0221] The computer-readable storage medium includes, for example, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0222] The embodiments described in the specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the methods and principles of this application should be covered within the scope of protection of this application.

Claims

1. A scan path planning method for intracranial three-dimensional cerebral blood flow scanning, characterized by, The method comprises the following steps: determining a theoretical model and an actual model; mapping a cerebral vascular reference region in the theoretical model to the actual model to determine a projection model; wherein the projection model has a predicted distribution region corresponding to the cerebral vascular region; determining a scanning path based on the predicted distribution region in the projection model; sending the current scanning path to a scanning execution module; wherein the step of mapping the cerebral vascular reference region in the theoretical model to the actual model to determine the projection model comprises: determining a skull deviation value based on the difference between the skull shape corresponding to the actual model and the skull shape corresponding to the theoretical model; adjusting the cerebral vascular reference region based on the skull deviation value, and mapping the adjusted cerebral vascular reference region to the actual model to determine the projection model; the method for constructing the theoretical model comprises: obtaining vascular reference data and constructing a theoretical cerebral vascular model based on the vascular reference data; obtaining skeletal reference data and constructing a theoretical skull model based on the skeletal reference data; wherein both the theoretical skull model and the actual model can reflect the shape of the human skull; determining the theoretical model based on the theoretical cerebral vascular model and the theoretical skull model; in the specific method for determining the scanning path based on the predicted distribution region in the projection model, the method comprises: model partitioning: partitioning the projection model to determine a plurality of scanning regions; probability calculation: evaluating the probability of the cerebral vascular appearing in the scanning region based on the distribution of the predicted distribution region to determine the partition probability corresponding to the scanning region; determining the scanning path based on the partition probability; in the specific method for model partitioning, the method comprises: partitioning the projection model based on the partition level to determine a plurality of scanning regions; wherein the partition level affects the proportion of a single scanning region in the projection model; determining the partition level based on the judgment result, and returning the determined partition level; the partition iteration condition specifically comprises a range trigger condition and a vascular trigger condition; the range trigger condition is that the area of the current single scanning region is greater than a preset maximum range threshold; the vascular trigger condition is that the cerebral vascular is found in the scanning result of the previous scanning region, and the area of the current scanning region is greater than a preset minimum range threshold.

2. The scan path planning method for intracranial three-dimensional cerebral blood flow scan according to claim 1, wherein, in the specific method for determining the partition level based on the judgment result, the method comprises: comparing the range of the current scanning region with the maximum range threshold and the minimum range threshold, updating the partition level based on the comparison result, and returning the determined partition level.

3. The scan path planning method for intracranial three-dimensional cerebral blood flow scan according to claim 1, wherein, in the specific method for probability calculation, the method comprises: calculating the probability calculation index associated with the scanning region; determining the partition probability corresponding to the scanning region based on the probability calculation index; wherein the probability calculation index comprises a theoretical existence probability; in the specific method for calculating the theoretical existence probability associated with the scanning region, the method comprises: The theoretical existence probability associated with the to-be-scanned region is calculated based on a proportion value between a part of the predicted distribution region overlapping the to-be-scanned region and the whole to-be-scanned region.

4. The scan path planning method for intracranial three-dimensional cerebral blood flow scan according to claim 3, wherein, In the specific method of calculating the theoretical existence probability associated with the to-be-scanned region, further comprising: comparing the theoretical existence probability and the probability grading coefficient to classify all to-be-scanned regions and determine low-probability partitions and high-probability partitions; updating the theoretical existence probability of all low-probability partitions that meet the blood vessel connectivity condition based on the probability increasing coefficient; wherein the blood vessel connectivity condition includes that the low-probability partition is located between two high-probability partitions.

5. The scan path planning method for intracranial three-dimensional cerebral blood flow scan according to claim 3, wherein, Before calculating the probability calculation index associated with the to-be-scanned region, further comprising: determining at least one probability calculation index associated with the to-be-scanned region and an index weight corresponding to the probability calculation index; wherein the index weight includes a theoretical existence weight corresponding to the theoretical existence probability; In the specific method of determining the partition probability corresponding to the to-be-scanned region based on the probability calculation index, comprising: determining the partition probability corresponding to the to-be-scanned region based on all probability calculation indexes associated with the to-be-scanned region and the index weight corresponding to the probability calculation index; And / or, the probability calculation index further includes a model matching degree, and the index weight further includes a model matching weight corresponding to the model matching degree; In the specific method of calculating the model matching degree associated with the to-be-scanned region, comprising: calculating a deviation value between the brain blood vessel morphology of the scan result and the brain blood vessel morphology of the projection model based on the scan result of the to-be-scanned region that has been scanned, and determining the model matching degree based on the deviation value; And / or, the probability calculation index further includes a scan completion degree, and the index weight further includes a completion degree weight corresponding to the scan completion degree; In the specific method of calculating the scan completion degree associated with the to-be-scanned region, comprising: determining a total sum of adjacent blood vessels corresponding to the to-be-scanned region based on the scan result of the to-be-scanned region that has been scanned; wherein the total sum of adjacent blood vessels is used to indicate the sum of the lengths of all brain blood vessels connected to the to-be-scanned region in the scan result and the brain blood vessels predicted to exist in the to-be-scanned region; comparing the numerical values of all total sums of adjacent blood vessels, and determining a maximum blood vessel length based on the total sum of adjacent blood vessels with the largest numerical value; determining the scan completion degree associated with the to-be-scanned region based on the ratio of the total sum of adjacent blood vessels to the maximum blood vessel length; And / or, the probability calculation index further includes an image matching degree, and the index weight further includes an image matching weight corresponding to the image matching degree; In the specific method of calculating the image matching degree associated with the to-be-scanned region, comprising: determining an actual scan image based on the scan result of the to-be-scanned region that has been scanned; determining a theoretical scan image, calculating a gap between the brain blood vessel morphology in the actual scan image and the brain blood vessel morphology in the theoretical scan image, and determining the image matching degree associated with the to-be-scanned region based on the gap; And / or, the probability calculation index further includes a blood vessel correlation degree, and the index weight further includes a blood vessel correlation weight corresponding to the blood vessel correlation degree; In the specific method of calculating the blood vessel correlation degree associated with the to-be-scanned region, comprising: determining a current blood vessel position based on a scanning result of a to-be-scanned region which has completed scanning; determining a blood vessel correlation degree corresponding to each to-be-scanned region based on the current blood vessel position and a predicted distribution region; wherein the blood vessel correlation degrees corresponding to the to-be-scanned regions gradually decrease in a gradient direction according to a distance between the to-be-scanned regions and the current blood vessel position, and a distribution trend direction of the predicted distribution region.

6. The scan path planning method for intracranial three-dimensional cerebral blood flow scan according to claim 5, wherein, In a specific method of determining a sub-region probability corresponding to a to-be-scanned region based on all probability calculation indexes associated with the to-be-scanned region and index weights corresponding to the probability calculation indexes, the method comprises: determining a maximum weight based on a sum of the index weights corresponding to the to-be-scanned region; calculating a proportion of the index weight corresponding to the probability calculation index in the maximum weight, and updating the index weight corresponding to the probability calculation index based on a calculation result; determining an index calculation result based on a product of the probability calculation index and the corresponding index weight, and determining the sub-region probability based on a sum of all index calculation results associated with the to-be-scanned region.

7. A method of intracranial three-dimensional cerebral blood flow scanning, characterized by, The scanning path planning method of any one of claims 1 to 6 applied to intracranial three-dimensional cerebral blood flow scanning further comprises, after the current scanning path is sent to the scanning execution module: sequentially judging whether a cerebral blood vessel exists in each to-be-scanned region based on a scanning result corresponding to the to-be-scanned region, and returning to determining the scanning path based on the predicted distribution region in the projection model according to a judgment result to update the scanning path.

8. A scan path planning system for intracranial three-dimensional cerebral blood flow scanning, characterized by, comprises: a data acquisition module (1) configured to determine a theoretical model and an actual model; a model mapping module (2) configured to map a cerebral blood vessel reference region in the theoretical model to the actual model to determine a projection model; wherein the projection model has a predicted distribution region corresponding to the cerebral blood vessel region; an algorithm scheduling module (3) configured to determine a scanning path based on the predicted distribution region in the projection model; a path sending module (4) configured to send the current scanning path to a scanning execution module; wherein the step of mapping the cerebral blood vessel reference region in the theoretical model to the actual model to determine the projection model comprises: determining a skull deviation value based on a difference between a skull shape corresponding to the actual model and a skull shape corresponding to the theoretical model; adjusting the cerebral blood vessel reference region based on the skull deviation value, and mapping the adjusted cerebral blood vessel reference region to the actual model to determine the projection model; the construction method of the theoretical model comprises: acquiring blood vessel reference data, and constructing a theoretical cerebral blood vessel model based on the blood vessel reference data; acquiring bone reference data, and constructing a theoretical skull model based on the bone reference data; wherein the theoretical skull model and the actual model can reflect the shape of a human skull; determining the theoretical model based on the theoretical cerebral blood vessel model and the theoretical skull model; in a specific method of determining the scanning path based on the predicted distribution region in the projection model, the method comprises: model partitioning: performing model partitioning based on the projection model to determine a plurality of to-be-scanned regions; Probability calculation: based on the distribution of the prediction distribution area, the probability of the brain blood vessels appearing in the to-be-scanned area is evaluated, and the partition probability corresponding to the to-be-scanned area is determined; Based on the partition probability, the scanning path is determined; In the specific method of the model partitioning step, the following steps are included: The projection model is partitioned based on the partition level to determine a plurality of to-be-scanned areas; wherein the partition level affects the proportion of a single to-be-scanned area in the projection model; It is judged whether the current to-be-scanned area meets the partition iteration condition, and the partition level is updated according to the judgment result, and the current partition level is returned; The partition iteration condition specifically includes a range trigger condition and a blood vessel trigger condition; The range trigger condition is that the area of the current single to-be-scanned area is greater than a preset maximum range threshold; The blood vessel trigger condition is that the brain blood vessels are found in the scanning result of the previous to-be-scanned area, and the area of the current to-be-scanned area is greater than a preset minimum range threshold.

9. A computer readable storage medium, characterized in that, A computer program capable of being loaded and executed by the processor to perform the method of any one of claims 1 to 6 or the method of claim 7 is stored.

Images