Scanning and positioning method, device, computer equipment and computer-readable storage medium

Abstract

Embodiments of the present invention disclose a scanning and positioning method, apparatus, computer device, and computer-readable storage medium. The method includes: obtaining three-dimensional image data of a part to be positioned; inputting the three-dimensional image data into a pre-trained position point feature model to obtain respective position point feature data output by the position point feature model; determining plane fitting points according to the respective position point feature data, and determining a plane equation to be positioned according to the plane fitting points. The method provided by the embodiments of the present invention can realize the positioning of a scanning plane only through the three-dimensional image data of the part to be positioned, and can be applicable to the situation of scanning and positioning different scanning parts through different scanning modalities, making the positioning of the scanning plane faster and more accurate, reducing the time required for the scanning process, improving the accuracy of the scanned image, and overcoming the inaccuracy of positioning according to anatomical landmark points. Even when the pre-scanning range is incomplete, the positioning of the scanning plane can still be well realized.

Description

[0001] This is a divisional application. The application number of the original application is 2018108353175, the application date is July 26, 2018, and the invention title is "Scanning and Positioning Method, Device, Computer Equipment and Computer Readable Storage Medium". Technical Field

[0002] Embodiments of the present invention relate to the technical field of medical image scanning, and in particular, to a scanning and positioning method, device, computer equipment and computer readable storage medium. Background Art

[0003] Medical imaging examinations play a very important auxiliary role in the clinical diagnosis of diseases. During medical scanning, it is often necessary to scan a fixed position / plane of the target area of the subject to assist in the diagnosis, treatment, etc. of diseases.

[0004] Taking magnetic resonance imaging as an example, for a conventional magnetic resonance scanning process, usually, after a doctor identifies the anatomical position through a pre-scan image, a reference positioning line is manually calibrated, and a scanning sequence is applied to the anatomical position according to the reference positioning line for accurate scanning.

[0005] Taking cardiac magnetic resonance scanning as an example, cardiac magnetic resonance positioning scanning requires the patient to hold their breath at least 5 times, namely three cross-sections, multi-layer axial, pseudo-two-chamber, pseudo-four-chamber, and multi-layer short axis. The breath-holding scanning time is at least 4 - 5 minutes, and the doctor also needs to manually position six times in the middle. Each positioning result will affect the subsequent scanning result. If the technician has insufficient experience, repeated scanning and positioning may be required. The positioning time is long and the final positioning result may be difficult to meet the needs of clinical diagnosis. At the same time, when performing cardiac positioning scanning through manual positioning, the results are not uniform. For the same subject, there may be significant differences in the positioning results between different operators and at different times by the same operator. It can be seen that the existing scanning and positioning methods are inefficient, prolong the time required for the entire scanning process, and due to manual and experience differences, the recognition results and calibration results may be inconsistent, and the accuracy of scanning parameters cannot be guaranteed.

[0006] Currently, the positioning of the scanning plane can also be performed according to the anatomical position points of the part to be positioned. Specifically, by detecting each anatomical position in the 3D detection image of the part to be positioned, the scanning plane is determined through the fixed anatomical position points. For example, through the 3D detection image, position points such as the mitral valve point, the apex point, the left ventricle to aortic outflow point, and the right ventricle maximum angle point in the heart region are determined, and the plane formed by the specific position points is determined as the scanning plane. For example, the short-axis plane is perpendicular to the line connecting the mitral valve point and the apex, passes through the center of the left ventricle, and the four-chamber plane passes through the line connecting the mitral valve and the apex and passes through the point of the maximum diameter of the right ventricle. However, these points are approximate position points summarized according to the doctor's experience. During actual scanning, the plane determined according to the anatomical position points may not be the optimal plane, and some pre-scanning images may have insufficient range, resulting in the absence of some anatomical position points, making the positioning of the scanning plane inaccurate. Summary of the Invention

[0007] The embodiments of the present invention provide a scanning positioning method, device, computer device, and computer-readable storage medium to achieve fast and accurate positioning of the scanning plane.

[0008] In a first aspect, an embodiment of the present invention provides a scanning positioning method, including:

[0009] Obtain three-dimensional image data of the part to be positioned;

[0010] Input the three-dimensional image data into a pre-trained position point feature model to obtain each position point feature data output by the position point feature model;

[0011] Extract plane fitting points from each of the position points according to each of the position point feature data, and determine the plane equation to be positioned according to the plane fitting points.

[0012] In a second aspect, an embodiment of the present invention further provides a scanning positioning device, including:

[0013] A data acquisition module for acquiring three-dimensional image data of the part to be positioned;

[0014] A feature data module for inputting the three-dimensional image data into a pre-trained position point feature model to obtain each position point feature data output by the position point feature model;

[0015] A plane determination module for determining plane fitting points according to each position point feature data and determining the plane equation to be positioned according to the plane fitting points.

[0016] In a third aspect, an embodiment of the present invention further provides a computer device, and the device includes:

[0017] One or more processors;

[0018] A storage device for storing one or more programs;

[0019] When the one or more programs are executed by the one or more processors, the one or more processors implement the scanning and positioning method provided in any embodiment of the present invention.

[0020] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the scanning and positioning method provided in any embodiment of the present invention.

[0021] In the embodiment of the present invention, three-dimensional image data of a part to be positioned is obtained; the three-dimensional image data is input into a pre-trained position point feature model to obtain respective position point feature data output by the position point feature model; plane fitting points are determined according to the respective position point feature data, and a plane equation to be positioned is determined according to the plane fitting points. The scanning and positioning method provided in the embodiment of the present invention can realize the positioning of the scanning plane of the part to be positioned only through the three-dimensional image data of the part to be positioned, and can be applicable to the situation of scanning and positioning different scanning parts through different scanning modalities, making the positioning of the scanning plane faster and more accurate, thereby reducing the time required for the scanning process, improving the accuracy of the scanned image, and overcoming the inaccuracy of positioning according to anatomical landmark points. Even when the pre-scanning range is incomplete, the positioning of the scanning plane can still be well realized. Description of the Drawings

[0022] Figure 1 is a flowchart of the scanning and positioning method provided in Embodiment 1 of the present invention;

[0023] Figure 2a is a flowchart of the scanning and positioning method provided in Embodiment 2 of the present invention;

[0024] Figure 2b is a schematic diagram of using a distance calculation model for scanning and positioning in the scanning and positioning method provided in the embodiment of the present invention;

[0025] Figure 3a is a flowchart of the scanning and positioning method provided in Embodiment 3 of the present invention;

[0026] Figure 3b is a schematic diagram of the training process of the distance field calculation model in the scanning and positioning method provided in Embodiment 3 of the present invention;

[0027] Figure 3c is a schematic diagram of training the distance calculation model in the scanning and positioning method provided in the embodiment of the present invention;

[0028] Figure 4aIt is the flowchart of the scanning and positioning method provided in the fourth embodiment of the present invention;

[0029] Figure 4b It is the schematic diagram of using the surface segmentation model for scanning and positioning in the scanning and positioning method provided in the embodiment of the present invention;

[0030] Figure 5a It is the flowchart of the scanning and positioning method provided in the fifth embodiment of the present invention;

[0031] Figure 5b It is the schematic diagram of training the surface segmentation model in the scanning and positioning method provided in the embodiment of the present invention;

[0032] Figure 6 It is the schematic structural diagram of the scanning and positioning device provided in the sixth embodiment of the present invention;

[0033] Figure 7 It is the schematic structural diagram of the computer device provided in the seventh embodiment of the present invention. Specific Embodiments

[0034] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. Additionally, it should be noted that for the convenience of description, only parts related to the present invention are shown in the accompanying drawings rather than all structures.

[0035] Embodiment 1

[0036] Figure 1 It is the flowchart of the scanning and positioning method provided in the first embodiment of the present invention. This embodiment is applicable to the situation of positioning and scanning various medical imaging parts through various scanning modalities, especially applicable to the situation of using a cardiac magnetic resonance imaging system to scan and position the heart. This method can be executed by a scanning and positioning device, and the scanning and positioning device can be implemented in software and / or hardware. For example, the scanning and positioning device can be configured in a computer device. As Figure 1 shown, this method specifically includes:

[0037] S110. Obtain three-dimensional image data of the part to be positioned.

[0038] In this embodiment, the scanning plane is positioned through the three-dimensional image data of the part to be positioned. Optionally, the three-dimensional image data of the part to be positioned is 3D reconnaissance data formed by pre-scanning the part to be positioned. Exemplarily, when the part to be positioned is the cardiac region, the three-dimensional image data is the 3D reconnaissance data of the cardiac region.

[0039] Optionally, the method for obtaining the three-dimensional image data of the part to be located includes: obtaining multi-layer image data in the same direction within the scanning area, performing image reconstruction on the multi-layer image data to form the three-dimensional image data of the scanning area, and using a preset image extraction algorithm to extract the three-dimensional image data of the part to be located from the three-dimensional image data of the scanning area. Optionally, an automatic threshold algorithm can be used to segment the three-dimensional image data of the scanning area, segment the three-dimensional image data of the scanning area into foreground data and background data, and then use a morphological method to extract the three-dimensional image data of the part to be located.

[0040] Exemplarily, multi-layer scanning can be performed along the z-axis direction to obtain multi-layer image data in the z-axis direction. After performing image reconstruction on the multi-layer image data in the z-axis direction using an image reconstruction algorithm, the three-dimensional image data of the scanning area is obtained, and then the three-dimensional image data of the part to be located is extracted from the three-dimensional image data of the scanning area. When the part to be located is the heart area, an automatic threshold segmentation algorithm can be used to segment the foreground and background of the three-dimensional image data of the scanning area, and then a morphological method is used to extract the three-dimensional image data of the heart area.

[0041] S120: Input the three-dimensional image data into a pre-trained position point feature model to obtain the feature data of each position point output by the position point feature model.

[0042] In this embodiment, after obtaining the three-dimensional image data of the part to be located, the three-dimensional image data of the part to be located is input into a pre-trained position point feature model to obtain the feature data of each position point in the three-dimensional image data of the part to be located.

[0043] Optionally, the feature data of each position point includes the coordinates of each position point and the distance parameter between each position point and the plane to be located. Among them, the distance parameter between each position point and the plane to be located can be the distance value between each position point and the plane to be located, or the distance-related value between each position point and the plane to be located.

[0044] Optionally, before inputting the three-dimensional image data into the pre-trained position point feature model, it further includes:

[0045] Normalize the three-dimensional image data of the part to be located through a preset image data normalization algorithm to obtain image normalization data.

[0046] In this embodiment, before inputting the three-dimensional image data of the part to be located into the position point feature model, the three-dimensional image data is normalized. Normalizing the three-dimensional image data means scaling the three-dimensional image data according to a set rule so that it falls within a small specific range, facilitating the data processing during subsequent data processing. Among them, the algorithm for normalizing the three-dimensional image data is not limited here. For example, the normalization algorithm can be the min-max normalization algorithm, the zero-mean normalization algorithm, or the decimal scaling normalization algorithm. Among them, the min-max normalization algorithm normalizes the data through the maximum and minimum values of the data, the zero-mean normalization algorithm normalizes the data through the mean and standard deviation of the data, and the decimal scaling normalization algorithm normalizes the data by moving the decimal point position of the data.

[0047] Optionally, the image data normalization algorithm is the min-max normalization algorithm, that is, the three-dimensional image data of the part to be located can be normalized through the min-max normalization algorithm. If the maximum pixel value of each position point of the part to be located is I max , the minimum value is I min , and the pixel value of a certain position point of the part to be located is I c , and the pixel value of this position point after normalization is I, then the normalization calculation formula for the pixel values of each position point of the part to be located is: I = (I c - I min ) / (I max - I min ).

[0048] Optionally, the image data normalization algorithm is the zero-mean normalization algorithm, that is, the three-dimensional image data of the part to be located can be normalized through the zero-mean normalization algorithm. If the mean pixel value of each position point of the part to be located is and the standard deviation is σ, and the pixel value of a certain position point of the part to be located is I c , and the pixel value of this position point after normalization is I, then the normalization calculation formula for the pixel values of each position point of the part to be located is:

[0049] S130. Determine the plane fitting points according to the feature data of each position point, and determine the plane equation to be located according to the plane fitting points.

[0050] In this embodiment, some position points can be selected from the feature data of each position point as plane fitting points, and the plane equation to be located is determined according to the selected plane fitting points; or the position points can not be screened, and all position points or randomly selected part of the position points can be directly used as plane fitting points, and the plane equation to be located is determined according to the plane fitting points.

[0051] Optionally, determining the plane equation to be located based on the plane-fitting points includes: screening out the plane-fitting points from each position point according to the characteristic parameters of each position point, and using a preset fitting algorithm to fit the coordinate values of each plane-fitting point to obtain the plane equation to be located.

[0052] Optionally, screening each position point according to the preset distance parameter range and the distance parameter between each position point of the three-dimensional image data of the part to be located and the plane to be located, using the position points corresponding to the distance parameters that meet the preset distance parameter range conditions as the plane-fitting points, and then fitting the plane equation to be located according to the coordinate values of each plane-fitting point. Exemplarily, the least squares fitting algorithm can be used to fit the coordinate values of each plane-fitting point to calculate the plane equation to be located.

[0053] Optionally, determining the plane equation to be located based on the plane-fitting points includes: determining the plane equation to be located according to the coordinate values of each plane-fitting point through an optimization algorithm. Optionally, when determining the plane equation to be located through the optimization algorithm, the plane-fitting points can be all the position points in the three-dimensional image data of the part to be located, or can be part of the position points in the three-dimensional image data.

[0054] Exemplarily, the parameters of the plane equation to be located can be calculated using an optimization algorithm. In this embodiment, there is no limitation on the optimization algorithm. For example, the optimization algorithm can be algorithms such as the gradient descent method, Newton's method, conjugate direction method, or conjugate gradient method.

[0055] Exemplarily, through determine the parameters of the plane equation to be located. Where n is the number of all plane-fitting points in the part to be located, x i is the x-axis coordinate value of the i-th plane-fitting point of the plane to be located, y i is the y-axis coordinate value of the i-th plane-fitting point of the plane to be located, z i is the z-axis coordinate value of the i-th plane-fitting point of the plane to be located, D i is the distance value between the i-th plane-fitting point of the part to be located and the plane to be located when the plane equation to be located is ax + by + cz + d = 0. Making the sum of the differences between the distance values between all plane-fitting points in the part to be located and the plane ax + by + cz + d = 0 and the distance values obtained through the distance field calculation model the smallest, and satisfying a 2 + b 2 + c 2 = 1, take the values of a, b, c, and d as the parameters of the plane to be located.

[0056] Optionally, determining the plane equation to be located based on the plane-fitting points includes: determining the plane equation to be located according to the coordinate values of each plane-fitting point by means of voting. Optionally, when determining the plane equation to be located by means of voting, the plane-fitting points can be all the position points in the three-dimensional image data of the part to be located, or can be some position points in the three-dimensional image data. That is, all or some plane-fitting points can be selected for voting, and the parameter combination with the highest probability is determined as the parameter of the plane equation to be located.

[0057] In the embodiment of the present invention, three-dimensional image data of a part to be located is obtained; the three-dimensional image data is input into a pre-trained position point feature model, and each position point feature data output by the position point feature model is obtained; plane-fitting points are determined according to each position point feature data, and the plane equation to be located is determined according to the plane-fitting points. The scanning and positioning method provided by the embodiment of the present invention can realize the positioning of the scanning plane only through the three-dimensional image data of the part to be located, and can be applied to the situation when different scanning parts are scanned and positioned by different scanning modalities, making the positioning of the scanning plane faster and more accurate, thereby reducing the time required for the scanning process, improving the accuracy of the scanned image, and overcoming the inaccuracy of positioning according to anatomical landmark points. When the pre-scanning range is incomplete, the positioning of the scanning plane can still be well realized.

[0058] Embodiment 2

[0059] Figure 2a is a flowchart of the scanning and positioning method provided by the second embodiment of the present invention. This embodiment is further optimized on the basis of the above embodiment. As Figure 2a shown, the method includes:

[0060] S210. Obtain three-dimensional image data of a part to be located.

[0061] S220. Input the three-dimensional image data into a pre-trained distance field calculation model, and obtain a distance matrix output by the distance field calculation model.

[0062] In this embodiment, the pre-trained position point feature model is specifically a distance field calculation model, and each position point feature data is specifically the distance value between each position point and the plane to be located. The distance value between each position point and the plane to be located in the three-dimensional image data of the part to be located is calculated by the distance field calculation model, and the distance values between each position point and the plane to be located form a distance matrix.

[0063] It should be noted that if the three-dimensional image data of the part to be located is normalized before being input into the distance field calculation model, the distance values included in the distance matrix output by the distance field calculation model can be the distance values after normalization or the actual distance values between each position point and the plane to be located. Preferably, the distance value after normalization is used as the distance value in the distance matrix output by the distance field calculation model to make the positioning of the plane to be located more accurate.

[0064] S230. Screen each position point according to the preset distance range and each distance value, use the position points that meet the fitting conditions as plane fitting points, and determine the plane equation to be located according to the plane fitting points.

[0065] Optionally, use the position points corresponding to the distance values within the preset distance range as plane fitting points, and use a preset fitting algorithm to fit each plane fitting to calculate the plane equation to be located. Optionally, the preset distance range can be determined according to the specific situation of the part to be located. Exemplarily, when the part to be located is the heart, the preset distance range can be (0, 0.1). Optionally, the preset fitting algorithm can be the least squares fitting algorithm.

[0066] In another embodiment of the present invention, it is also possible not to screen each position point, directly use all position points or randomly select some position points as plane fitting points, and determine the plane equation to be located according to the plane fitting points through an optimization algorithm or a voting method.

[0067] Optionally, for more detailed content on determining the plane equation to be located according to each plane fitting point through a preset fitting algorithm, an optimization algorithm or a voting method, reference can be made to the above embodiments, and details are not described herein again.

[0068] Figure 2b It is a schematic diagram of using a distance calculation model for scanning and positioning in the scanning and positioning method provided by an embodiment of the present invention. The figure schematically shows the process of positioning a scanning plane through a pre-trained distance field calculation model. As Figure 2b shown, input the three-dimensional image data of the part to be located extracted from the pre-scan image into the pre-trained distance field calculation model to obtain the distance field output by the distance field calculation model, and determine the plane parameters of the plane to be located according to the distance field.

[0069] It should be noted that 57 subjects were tested for cardiac magnetic resonance imaging using the scanning and positioning method provided by the embodiments of the present invention. The test results are shown in Table 1. Table 1 shows the average error of the normal vector of the scanning and positioning plane and the average error of the distance field determined by using the scanning and positioning method provided by the embodiments of the present invention at different scanning planes. The average error values of the normal vector and the distance field corresponding to each scanning and positioning plane can be obtained from Table 1. Moreover, cardiac magnetic resonance imaging was performed on the cardiac scanning planes of different subjects using the scanning and positioning method provided by the embodiments of the present invention. The imaging results were evaluated by doctors. Only one case required minor adjustment by doctors, and the others were clinically acceptable.

[0070] Table 1

[0071] Short-axis plane Two-chamber plane Three-chamber plane Four-chamber plane Average error of normal vector (°) 5.7 5.4 7.2 5.4 Average error of distance field (mm) 5.4 3.7 4.7 6.1

[0072] Among them, the calculation method of the normal vector angle error is:

[0073]

[0074] Among them, is the normal vector of the manually marked plane, is the normal vector of the scanning plane determined by the scanning and positioning method provided by the embodiments of the present invention. In this embodiment, each test data corresponds to a short-axis plane, a two-chamber plane, a three-chamber plane, and a four-chamber plane. The average error of the normal vector of each scanning plane in Table 1 refers to the average value of the angle errors of the normal vectors of the scanning and positioning planes corresponding to all test data.

[0075] The calculation method of the distance field error is: Among them, D0 is the distance between each position point in the cardiac region and the manually marked plane, and D1 is the actual distance between each position point in the cardiac region and the scanning plane determined by the scanning and positioning method provided by the embodiments of the present invention. Optionally, the actual distance between each position point and the scanning plane can be determined according to the distance matrix output by the distance field calculation model.

[0076] Optionally, if the distance values in the distance matrix output by the distance field calculation model are unnormalized distance values, the distance values in the distance matrix output by the distance field calculation model are directly used as the actual distances between the corresponding position points and the scanning plane. If the distance values in the distance matrix output by the distance field calculation model are normalized distance values, the distances between the position points output by the distance field calculation model and the scanning plane can be calculated in reverse to obtain the actual distances between the position points of the heart part and the scanning plane. Specifically, the product of each distance value output by the distance field calculation model and the preset threshold used for normalizing the distance matrix is used as the true distance between each position point and the scanning plane. The average distance field error in Table 1 refers to the average value of the distance value errors of the position points corresponding to the distance values within the preset threshold.

[0077] Based on the technical solution of the embodiment of the present invention, on the basis of the above embodiment, the process of inputting the three-dimensional image data into the pre-trained position point feature model to obtain the position point feature data output by the position point feature model, and determining the plane fitting points according to the position point feature data is specified. By calculating the distance values between the position points of the part to be located and the plane to be located through the pre-trained distance field calculation model, and determining the plane fitting points according to the distance values, the determination of the plane fitting points is made more accurate, and thus the positioning of the plane to be located is made more accurate.

[0078] Embodiment III

[0079] Figure 3a is a flowchart of the scanning and positioning method provided by Embodiment III of the present invention. This embodiment is further optimized on the basis of the above embodiment. As Figure 3a shown, the method includes:

[0080] S310. Obtain historical three-dimensional image data and plane parameters of the plane to be located corresponding to the historical three-dimensional image data.

[0081] In this embodiment, the pre-established distance field calculation model is trained based on the historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data. Among them, both the historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data are data for training. Optionally, the historical three-dimensional image data can be three-dimensional image data extracted from the historical scan data of the part to be located, and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data can be the normal vector of the plane to be located or the plane equation of the plane to be located.

[0082] Optionally, the acquisition method of the plane parameters of the plane to be located corresponding to the historical three-dimensional image data includes:

[0083] Using image processing software, each plane to be located is located by using a positioning line from historical three-dimensional image data, and plane parameters of each plane to be located are obtained. Optionally, a person can use the image processing software to locate the plane to be located with a positioning line to obtain the plane parameters of the corresponding plane to be located.

[0084] Optionally, the method for obtaining the plane parameters of the plane to be located corresponding to the historical three-dimensional image data includes:

[0085] Calculating the plane parameters of the plane to be located according to the image information of each sub-part image of the part to be located in the scanning image information corresponding to the historical three-dimensional image data.

[0086] Optionally, the scanning image information is high-resolution scanning image information obtained when the part to be located is formally scanned. Optionally, the plane parameters of the plane to be located corresponding to the historical three-dimensional image data are obtained from the real high-resolution scanning image information during scanning. Generally, the information volume of the formal scanning data during the scanning of the part to be located is more abundant than that of the general pre-scanning data, and the formal scanning data contains the annotation information of the doctor on the data of the formal scanning area of the scanning area (for example, the positioning processing of various cross-sections, chambers, etc. of the part to be located). The formal scanning data and the doctor's annotation information are used as the "gold standard" in the process of training the model, and the pre-established distance field calculation model is trained.

[0087] Optionally, the pre-scanning image is an image in DICOM format, and each DICOM image contains the image information of the current layer. For example, the upper left corner coordinate position of the current layer and the unit vector vx of the x-axis and the unit vector vy of the y-axis of the current layer. The unit vector perpendicular to vx and vy is used as the unit normal vector of the plane to be located, and the determined unit normal vector (a, b, c) is used as the plane parameter of the plane to be located. In addition, combining the upper left corner coordinate information (x0, y0, z0) of the current layer, the parameter d = -ax0 - by0 - cz0 is determined, and finally the plane parameter of the plane to be located is (a, b, c, d), and the plane equation of the plane to be located is ax + by + cz + d = 0.

[0088] Preferably, in the way of manual positioning, the image processing software is used to locate the plane to be located with a positioning line to obtain the plane parameters of the corresponding plane to be located. Using the image processing software to obtain the plane parameters of the plane to be located is more convenient and accurate.

[0089] S320. Calculating a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters.

[0090] Optionally, historical distance values between each position point in the historical three-dimensional image data and the plane to be located are calculated through a preset distance calculation algorithm, and the historical distance values between each position point and the plane to be located form a historical distance matrix corresponding to the historical three-dimensional image data. Optionally, the distance calculation algorithm can be the Euclidean distance algorithm. Exemplarily, if the equation of the plane to be located is ax + by + cz + d = 0, then the distance between any position point (x i , y i , z i ) in the historical three-dimensional image data and the plane to be located is:

[0091]

[0092] S330. Generate a training sample set based on the historical three-dimensional image data and the historical distance matrix, and use the training sample set to train a pre-established distance field calculation model to obtain a trained distance field calculation model.

[0093] In this embodiment, the historical three-dimensional image data and the historical distance matrix corresponding to the historical three-dimensional image data are used as sample pairs to train a pre-established distance field calculation model. Optionally, the pre-established distance field calculation model is a convolutional neural network model.

[0094] Figure 3b FIG. is a schematic diagram of the training process of the distance field calculation model in the scanning and positioning method provided in Embodiment 3 of the present invention. Exemplarily, the convolutional neural network adopted in the embodiments of the present invention is as Figure 3b shown. Among them, the solid (short) straight arrow indicates that the channel before the arrow undergoes operations of a convolutional layer, a block normalization layer, and a relu activation layer to obtain the channel after the arrow, the dashed curved arrow and the plus sign represent superposition and the relu activation layer, and the solid (long) curved arrow represents a concat layer, that is, the channel in front is arranged side by side with the channel behind.

[0095] Exemplarily, the operation process of obtaining 16-channel A from 1-channel A is: subjecting 1-channel A to operations of a convolutional layer, a block normalization layer, and a relu activation layer to obtain 16-channel A 。The operation process of obtaining 32-channel A from 16-channel A and 16-channel B is as follows: After superimposing 16-channel A and 16-channel B, it is operated through a relu activation layer. The operation result obtained through the relu activation layer is then operated through a convolutional layer, a block normalization layer, and a relu activation layer to obtain 32-channel A. The operation process of obtaining 32-channel C from 16-channel B and 64-channel C is as follows: After operating 64-channel C through a convolutional layer, a block normalization layer, and a relu activation layer, 16-channel D is obtained. After arranging 16-channel B side by side with 16-channel D, 32-channel C is obtained. The operation process of obtaining 64-channel C from 64-channel A, 64-channel B, and 32-channel B is as follows: After superimposing 64-channel A and 64-channel B, it is operated through a relu activation layer. The operation result obtained through the relu activation layer is then operated through a convolutional layer, a block normalization layer, and a relu activation layer to obtain 32-channel E. After arranging 32-channel B side by side with 32-channel E, 64-channel C is obtained. The operation process of obtaining 1-channel B from 32-channel C and 32-channel D is as follows: After superimposing 32-channel C and 32-channel D, it is operated through a relu activation layer. The operation result obtained through the relu activation layer is then operated through a convolutional layer, a block normalization layer, and a relu activation layer to obtain 1-channel B.

[0096] Optionally, the method for training the distance field calculation model is not limited herein. Exemplarily, the training method for the distance field calculation model can be the backpropagation algorithm, the stochastic gradient descent method, or the stochastic optimization method. In this embodiment, the stochastic optimization method (adam method) can be used as the training method for the distance field calculation model.

[0097] Optionally, the distance field calculation model output I out and the gold standard I label The 1-norm between them is used as the cost function during the training process of the distance field calculation model. Among them, the gold standard I label is the historical distance matrix corresponding to the historical three-dimensional image data, that is, when the historical three-dimensional image data is input into the distance field calculation model, the standard distance matrix that should be output. Exemplarily, the cost function loss = |I out - I label |. Optionally, the cost function during the training process of the distance field calculation model can also be the 2-norm, weighted 1-norm, or weighted 2-norm between the distance field calculation model output I out and the gold standard I label , which is not limited herein.

[0098] Optionally, to make the trained distance field calculation model applicable to various types of 3D image data, the training data can be augmented by deforming the historical 3D image data and the corresponding historical distance matrix, and using the deformed data as the training data for the distance field calculation model. Exemplarily, the historical 3D image data and the corresponding historical distance matrix can be translated (e.g., randomly translated along the x, y, or z directions, with a translation range of ±50 mm), rotated (e.g., rotated by a random angle around a random rotation axis, with an angle range of ±20°), scaled (e.g., randomly scaled by a factor of 0.7 - 1.3), etc. The processed data is used as training samples to train the distance field calculation model. Deforming the training data and using it as training data can augment the limited training data, increasing the training samples of the distance field calculation model, so that the trained distance field calculation model can still output an accurate distance matrix when the input 3D image data is inaccurate or offset.

[0099] Optionally, before generating a training sample set using the historical 3D image data and the historical distance matrix corresponding to the historical 3D image data, it further includes:

[0100] Chunk the historical 3D image data and the historical distance matrix corresponding to the historical 3D image data according to a preset chunking rule, and use the chunked historical 3D data and the historical distance matrix corresponding to the historical 3D image data to generate a training sample set.

[0101] Optionally, image patches of the same size are extracted by means of a sliding window or random selection. The physical range of the image patches can be 50 mm * 50 mm * 50 mm - 120 mm * 120 mm * 120 mm, and the spatial resolution range of the image patches can be 2 mm * 2 mm * 2 mm - 5 mm * 5 mm * 5 mm. Exemplarily, the size of the image patch is 100 mm * 100 mm * 100 mm, and the spatial resolution is 3 mm * 3 mm * 3 mm. Using the chunked historical 3D data and the historical distance matrix corresponding to the historical 3D image data to generate a training sample set reduces the size of each training sample pair, making the training process based on the training sample set faster.

[0102] Optionally, before generating training samples using the historical 3D image data and the historical distance matrix, it further includes:

[0103] Normalize the historical 3D image data and the historical distance matrix data respectively to form historical image normalized data and a historical distance normalized matrix.

[0104] Optionally, for more details on normalizing the historical 3D image data, refer to the above embodiments and will not be elaborated here.

[0105] Optionally, the historical distance matrix can be normalized using the same normalization algorithm as the historical three-dimensional image data, or a different normalization algorithm from the historical three-dimensional image data can be used to normalize the historical distance matrix. In this embodiment, since the data distribution laws of the historical three-dimensional image data and the historical distance matrix are different, a normalization algorithm different from the historical three-dimensional image data is used to normalize the historical distance matrix.

[0106] Exemplarily, a distance threshold T1 can be preset, and the historical distance matrix is truncated and normalized according to the distance matrix. Specifically, if any distance value in the historical distance matrix is D i , then the distance value where D i > T1 is set to T1, and then all the distance values in the historical distance matrix are divided by T1 to obtain a historical distance normalization matrix. It can be seen that the historical distance normalization values in the historical distance normalization matrix are between 0 and 1. Optionally, the value range of the distance threshold T1 is (30mm, 200mm). The size of the normalized historical distance normalization matrix is the same as the size of the historical three-dimensional image data, and each value in the historical distance normalization matrix is the normalized distance from the corresponding position point to the plane to be located.

[0107] Figure 3c FIG. is a schematic diagram for training a distance calculation model in the scanning and positioning method provided by an embodiment of the present invention. The figure schematically shows the process of training the distance calculation model. As Figure 3c shown, according to the plane parameters of the plane to be located, the distance field corresponding to the three-dimensional image data is calculated, and the distance field and the three-dimensional image data of the part to be located extracted from the pre-scanned image are used as training samples to train the distance calculation model, and a trained distance calculation model is obtained.

[0108] S340. Obtain the three-dimensional image data of the part to be located.

[0109] S350. Input the three-dimensional image data into the pre-trained distance field calculation model to obtain the distance matrix output by the distance field calculation model.

[0110] S360. Screen each position point according to the preset distance range and each distance value, use the position points that meet the fitting conditions as distance plane fitting points, and determine the plane equation to be located according to the distance plane fitting points.

[0111] It should be noted that the training method of the distance field calculation model provided by the embodiments of the present invention can be executed independently. That is to say, the operation steps in S310 - S330 provided by the embodiments of the present invention can be used alone to complete the training of the distance field calculation model, and the operations of determining the plane equation to be located through the distance calculation model based on the three-dimensional image data of the part to be located in the subsequent steps S340 - S360 are no longer executed.

[0112] The technical solution of the embodiments of the present invention adds an operation of training the distance field calculation model on the basis of the above embodiments. By obtaining historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data; calculating the historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters; generating a training sample set based on the historical three-dimensional image data and the historical distance matrix, and using the training sample set to train the pre-established distance field calculation model to obtain a trained distance field calculation model, making the obtained distance calculation model more accurate.

[0113] Embodiment 4

[0114] Figure 4a is a flowchart of the scanning and positioning method provided by Embodiment 4 of the present invention, and this embodiment is further optimized on the basis of the above embodiments. As Figure 4a shown, the method includes:

[0115] S410. Obtain three-dimensional image data of the part to be located.

[0116] S420. Input the three-dimensional image data into the pre-trained surface segmentation model to obtain the distance segmentation matrix output by the surface segmentation model.

[0117] In this embodiment, the pre-trained position point feature model is specifically a surface segmentation model, and each position point feature data is specifically the distance segmentation value between each position point and the plane to be located. The distance segmentation values between each position point and the plane to be located form a distance segmentation matrix by calculating through the surface segmentation model.

[0118] Optionally, the surface segmentation model is a classification model, and its output includes a foreground output channel and a background output channel. Obtain the data output by the foreground output channel and use it as the distance segmentation matrix corresponding to the three-dimensional image data of the part to be located.

[0119] S430. Screen each position point according to the preset first segmentation threshold and each distance segmentation value, and use the position points that meet the segmentation conditions as plane fitting points.

[0120] Optionally, after obtaining the distance segmentation matrix output by the surface segmentation model, each position point in the three-dimensional image data of the part to be located is screened according to a preset first segmentation threshold and the distance segmentation value between each position point in the distance segmentation matrix and the plane to be located, and the position point corresponding to the distance segmentation value greater than the preset first segmentation threshold is used as the plane fitting point. Optionally, the preset first segmentation threshold can be adjusted according to the specific part to be located. Illustratively, when the part to be located is the cardiac region, the preset first segmentation threshold can be 0.5.

[0121] S440. Fit each plane fitting point through a preset fitting algorithm to obtain the plane equation of the part to be located.

[0122] In this embodiment, a fitting algorithm is used to fit each plane fitting point to obtain the plane equation of the part to be located. Optionally, the method of forming the plane equation of the part to be located according to each plane fitting point is similar to the method of determining the plane equation of the part to be located according to the plane fitting point in the above embodiment, and the more detailed content can be found in the above embodiment and will not be elaborated here.

[0123] Figure 4b It is a schematic diagram of using a surface segmentation model for scanning and positioning in the scanning and positioning method provided by an embodiment of the present invention. The figure schematically shows the process of positioning the scanning plane through a pre-trained surface segmentation model. As Figure 4b shown, the three-dimensional image data of the part to be located extracted from the pre-scan image is input into a pre-trained surface segmentation model, and the distance segmentation matrix output by the surface segmentation model is obtained, and the plane parameters of the plane to be located are determined according to the distance segmentation matrix.

[0124] The technical solution of the embodiment of the present invention concretizes the process of inputting the three-dimensional image data into a pre-trained position point feature model to obtain the position point feature data output by the position point feature model and determining the plane fitting point according to the position point feature data on the basis of the above embodiment. By calculating the distance segmentation value between each position point of the part to be located and the plane to be located through a pre-trained surface segmentation model, and using a preset first segmentation threshold to use the position point corresponding to the distance segmentation value that meets the preset segmentation condition as the segmentation plane fitting point, the determination of the segmentation plane fitting point is more accurate, thereby making the positioning of the plane to be located more accurate, and screening out some position points from each position point as the segmentation plane fitting point makes the fitting speed of the plane to be located faster, reducing the time required for the scanning process.

[0125] Embodiment Five

[0126] Figure 5a It is a flowchart of the scanning and positioning method provided by Embodiment Five of the present invention, and this embodiment is further optimized on the basis of the above embodiment. As Figure 5aAs shown, the method includes:

[0127] S510. Obtain historical three-dimensional image data and plane parameters of a plane to be located corresponding to the historical three-dimensional image data.

[0128] S520. Calculate a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters.

[0129] In this embodiment, the manner of obtaining the historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data, and the manner of calculating the historical distance matrix are similar to those in the above embodiment. For specific details, refer to the above embodiment and will not be elaborated here.

[0130] S530. Segment the historical distance matrix according to a preset second segmentation threshold to obtain a historical distance segmentation matrix.

[0131] In this embodiment, after segmenting the historical distance matrix according to the preset second segmentation threshold, use the obtained historical distance segmentation matrix to train a surface segmentation model. Optionally, the historical distance segmentation matrix is composed of segmentation distance values between each position point in the historical three-dimensional image data and the plane to be located.

[0132] Optionally, segmenting the historical distance matrix according to a preset second segmentation threshold to obtain a historical distance segmentation matrix includes:

[0133] Adjust each historical distance value in the historical distance matrix according to the preset second segmentation threshold, and use the matrix composed of the adjusted historical distance values as the historical distance segmentation matrix.

[0134] Exemplarily, if the second segmentation threshold is T, then adjust each historical distance value in the historical distance matrix according to the second segmentation threshold T. Optionally, if any historical distance value in the historical distance matrix is D i , then adjust the historical distance value of D i < T to 1, and adjust other historical distance values to 0. Optionally, the value of the second segmentation threshold can be adjusted according to the part to be located. Exemplarily, when the part to be located is the heart region, the second segmentation threshold can be 5 mm.

[0135] S540. Generate a training sample set based on the historical three-dimensional image data and the historical segmentation distance matrix, and use the training sample set to train a pre-established surface segmentation model to obtain a trained surface segmentation model.

[0136] Optionally, the cost function in the training process of the surface segmentation model can be a cost function for classification such as focal loss function focalloss, dice, etc. In this embodiment, the training method of the surface segmentation model is similar to the training method of the distance calculation model in the above embodiment. For more detailed content, please refer to the above embodiment and will not be elaborated here.

[0137] Figure 5b It is a schematic diagram of training the surface segmentation model in the scanning and positioning method provided by the embodiment of the present invention. The figure schematically shows the process of training the surface segmentation model. As Figure 5b shown, calculate the distance segmentation matrix corresponding to the three-dimensional image data according to the plane parameters of the plane to be located, and use the distance segmentation matrix and the three-dimensional image data of the part to be located extracted from the pre-scanned image as training samples to train the surface segmentation model to obtain a trained surface segmentation model.

[0138] S550. Obtain the three-dimensional image data of the part to be located.

[0139] S560. Input the three-dimensional image data into the pre-trained surface segmentation model to obtain the distance segmentation matrix output by the surface segmentation model.

[0140] S570. Screen each position point according to the preset first segmentation threshold and each distance segmentation value, and use the position points that meet the segmentation conditions as plane fitting points.

[0141] S580. Fit each plane fitting point through a preset fitting algorithm to obtain the plane equation to be located.

[0142] It should be noted that the training method of the surface segmentation model provided by the embodiment of the present invention can be executed independently. That is to say, the operation steps in S510-S540 provided by the embodiment of the present invention can be used alone to complete the training of the surface segmentation model, and the subsequent steps S550-S580 of determining the plane equation to be located through the surface segmentation model based on the three-dimensional image data of the part to be located are no longer executed.

[0143] The technical solution of the embodiment of the present invention adds an operation of training the surface segmentation model on the basis of the above embodiment. By obtaining the historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data; calculating the historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters; segmenting the historical distance matrix according to the preset second segmentation threshold to obtain the historical distance segmentation matrix; generating a training sample set based on the historical three-dimensional image data and the historical segmentation distance matrix, and using the training sample set to train the pre-established surface segmentation model to obtain a trained surface segmentation model, making the trained surface segmentation model more accurate.

[0144] In another embodiment of the present invention, the pre-established plane determination model can also be directly trained according to the historical three-dimensional image data and the plane parameters corresponding to the historical three-dimensional image data; when scanning and positioning is required, the three-dimensional image data of the part to be positioned and the coordinate matrices of the x, y, and z axes of the three-dimensional image data of the part to be positioned are directly input into the trained plane determination model, and the plane parameters of the plane to be positioned output by the plane determination model are obtained. Optionally, the training method and data processing method of the plane model can be referred to the above embodiments and will not be elaborated here.

[0145] Embodiment Six

[0146] Figure 6 is a schematic structural diagram of the scanning and positioning device provided in Embodiment Six of the present invention. The scanning and positioning device can be implemented in software and / or hardware. For example, the scanning and positioning device can be configured in a computer device, such as Figure 6 shown, the device includes: a data acquisition module 610, a feature data module 620, and a plane determination module 630, where:

[0147] The data acquisition module 610 is used to acquire three-dimensional image data of the part to be positioned;

[0148] The feature data module 620 is used to input the three-dimensional image data into a pre-trained position point feature model to obtain each position point feature data output by the position point feature model;

[0149] The plane determination module 630 is used to determine plane fitting points according to each position point feature data and determine the plane equation of the plane to be positioned according to the plane fitting points.

[0150] In the embodiment of the present invention, the data acquisition module acquires three-dimensional image data of the part to be positioned; the feature data module inputs the three-dimensional image data into a pre-trained position point feature model to obtain each position point feature data output by the position point feature model; the plane determination module determines plane fitting points according to each position point feature data and determines the plane equation of the plane to be positioned according to the plane fitting points. The scanning and positioning method provided in the embodiment of the present invention can realize the positioning of the scanning plane only through the three-dimensional image data of the part to be positioned, and can be applicable to the situation of scanning and positioning different scanning parts through different scanning modalities, making the positioning of the scanning plane faster and more accurate, thereby reducing the time required for the scanning process, improving the accuracy of the scanned image, and overcoming the inaccuracy of positioning according to anatomical landmark points. Even when the pre-scanning range is incomplete, the positioning of the scanning plane can still be better realized.

[0151] On the basis of the above solution, the feature data module 620 is specifically used for:

[0152] Input the three-dimensional image data into a pre-trained distance field calculation model to obtain a distance matrix output by the distance field calculation model, where the distance matrix is composed of the distance values of each of the position points from the plane to be located.

[0153] Based on the above solution, the plane determination module 630 is specifically configured to:

[0154] Screen each of the position points according to a preset distance range and each of the distance values, use the position points that meet the fitting conditions as plane fitting points, and determine the plane equation to be located based on the plane fitting points.

[0155] Based on the above solution, the device further includes:

[0156] A historical data acquisition unit, configured to acquire historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data before acquiring the three-dimensional image data of the part to be located;

[0157] A distance matrix determination unit, configured to calculate a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters, where the distance matrix is composed of the distance values of each position point in the historical three-dimensional image data from the plane to be located;

[0158] A distance field model training unit, configured to generate a training sample set based on the historical three-dimensional image data and the historical distance matrix, and use the training sample set to train a pre-established distance field calculation model to obtain a trained distance field calculation model.

[0159] Based on the above solution, the feature data module 620 is specifically configured to:

[0160] Input the three-dimensional image data into a pre-trained surface segmentation model to obtain a distance segmentation matrix output by the surface segmentation model, where the distance segmentation matrix is composed of the distance segmentation values of each of the position points from the plane to be located.

[0161] Based on the above solution, the plane determination module 630 includes:

[0162] A fitting point determination unit, configured to screen each of the position points according to a preset first segmentation threshold and each of the distance segmentation values, and use the position points that meet the segmentation conditions as plane fitting points;

[0163] A plane fitting unit, configured to fit each of the plane fitting points through a preset fitting algorithm to obtain the plane equation to be located.

[0164] Based on the above solution, the device further includes:

[0165] A historical data acquisition unit, configured to acquire historical three-dimensional image data and plane parameters of a plane to be located corresponding to the historical three-dimensional image data before acquiring three-dimensional image data of a part to be located;

[0166] A distance matrix determination unit, configured to calculate a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters, where the historical distance matrix is composed of distance values between each position point in the historical three-dimensional image data and the plane to be located;

[0167] A segmentation matrix determination unit, configured to segment the historical distance matrix according to a preset second segmentation threshold to obtain a historical distance segmentation matrix, where the historical distance segmentation matrix is composed of segmentation distance values between each position point in the historical three-dimensional image data and the plane to be located;

[0168] A plane segmentation model training unit, configured to generate a training sample set based on the historical three-dimensional image data and the historical segmentation distance matrix, and use the training sample set to train a pre-established plane segmentation model to obtain a trained plane segmentation model.

[0169] The scanning and positioning device provided by the embodiment of the present invention can execute the scanning and positioning method provided by any embodiment, and has corresponding functional modules and beneficial effects for executing the method.

[0170] Embodiment Seven

[0171] Figure 7 It is a schematic structural diagram of a computer device provided by Embodiment Seven of the present invention. Figure 7 It shows a block diagram of an exemplary computer device 712 suitable for implementing the embodiment of the present invention. Figure 7 The shown computer device 712 is only an example and should not impose any limitation on the functions and usage scope of the embodiment of the present invention.

[0172] As Figure 7 shown, the computer device 712 is presented in the form of a general-purpose computing device. The components of the computer device 712 may include, but are not limited to: one or more processors 716, a system memory 728, and a bus 718 connecting different system components (including the system memory 728 and the processor 716).

[0173] The bus 718 represents one or more of several types of bus architectures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor 716, or a local bus using any of the multiple bus architectures. By way of example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0174] The computer device 712 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 712, including volatile and nonvolatile media, removable and non-removable media.

[0175] The system memory 728 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 730 and / or cache memory 732. The computer device 712 can further include other removable / non-removable, volatile / nonvolatile computer system storage media. By way of example only, a storage device 734 can be used for reading and writing on non-removable, nonvolatile magnetic media ( Figure 7 not shown, and typically referred to as a "hard disk drive"). Although Figure 7 not shown in the figure, a disk drive for reading and writing on a removable nonvolatile disk (such as a "floppy disk"), and an optical disk drive for reading and writing on a removable nonvolatile optical disk (such as a CD-ROM, DVD-ROM, or other optical media) can be provided. In these cases, each drive can be connected to the bus 718 through one or more data media interfaces. The memory 728 can include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the embodiments of the present invention.

[0176] A program / utility 740 having a set (at least one) of program modules 742 can be stored, for example, in the memory 728. Such program modules 742 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. The program modules 742 typically perform the functions and / or methods described in the embodiments of the present invention.

[0177] The computer device 712 can also communicate with one or more external devices 714 (such as a keyboard, a pointing device, a display 724, etc.), and can also communicate with one or more devices that enable a user to interact with the computer device 712, and / or communicate with any device that enables the computer device 712 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 722. Moreover, the computer device 712 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 720. As shown in the figure, the network adapter 720 communicates with other modules of the computer device 712 through the bus 718. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the computer device 712, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0178] The processor 716 executes various functional applications and data processing by running programs stored in the system memory 728, for example, implementing the scanning and positioning method provided by the embodiments of the present invention. The method includes:

[0179] Obtain three-dimensional image data of the part to be positioned;

[0180] Input the three-dimensional image data into a pre-trained position point feature model to obtain various position point feature data output by the position point feature model;

[0181] Determine plane fitting points according to the various position point feature data, and determine the plane equation to be positioned according to the plane fitting points.

[0182] Certainly, those skilled in the art can understand that the processor can also implement the technical solutions of the scanning and positioning method provided by any embodiment of the present invention.

[0183] Embodiment Eight

[0184] Embodiment Eight of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the scanning and positioning method provided by the embodiments of the present invention. The method includes:

[0185] Obtain three-dimensional image data of the part to be positioned;

[0186] Input the three-dimensional image data into a pre-trained position point feature model to obtain various position point feature data output by the position point feature model;

[0187] Determine the plane fitting points according to the feature data of each position point, and determine the plane equation to be located according to the plane fitting points.

[0188] Certainly, the computer-readable storage medium provided by the embodiments of the present invention, the computer program stored thereon is not limited to the method operations as described above, and can also execute the related operations in the scanning and positioning method provided by any embodiment of the present invention.

[0189] The computer storage medium of the embodiments of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device.

[0190] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, in which the computer-readable program code is carried. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0191] The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0192] Computer program code for performing the operations of the present invention may be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and also including conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0193] Note that the above is only the preferred embodiment of the present invention and the technical principles applied. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, more other equivalent embodiments may be included, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A scanning and positioning method, applicable to the situation of performing positioning scans on various medical imaging parts through various scanning modalities, characterized in that, Including: Obtaining three-dimensional image data of a part to be located; Inputting the three-dimensional image data into a pre-trained position point feature model to obtain respective position point feature data output by the position point feature model, where the feature data of each position point includes the coordinates of each position point and the distance parameter between each position point and the plane to be located; Selecting some position points as plane fitting points according to the respective position point feature data, and determining the plane equation to be located according to the plane fitting points.

2. The method according to claim 1, characterized in that, The step of inputting the three-dimensional image data into a pre-trained position point feature model to obtain respective position point feature data output by the position point feature model includes: Inputting the three-dimensional image data into a pre-trained distance field calculation model to obtain a distance matrix output by the distance field calculation model, where the distance matrix is composed of the distance values between each position point and the plane to be located.

3. The method according to claim 2, characterized in that, The step of selecting some position points as plane fitting points according to the respective position point feature data, and determining the plane equation to be located according to the plane fitting points includes: Screening each position point according to a preset distance range and each distance value, taking the position points that meet the fitting conditions as plane fitting points, and determining the plane equation to be located according to the plane fitting points.

4. The method according to claim 2, wherein Before obtaining the three-dimensional image data of the part to be located, it further includes: Obtaining historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data; Calculating a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters, where the distance matrix is composed of the distance values between each position point in the historical three-dimensional image data and the plane to be located; Generating a training sample set based on the historical three-dimensional image data and the historical distance matrix, and using the training sample set to train a pre-established distance field calculation model to obtain a trained distance field calculation model.

5. The method according to claim 1, wherein The step of inputting the three-dimensional image data into a pre-trained position point feature model to obtain respective position point feature data output by the position point feature model includes: Inputting the three-dimensional image data into a pre-trained surface segmentation model to obtain a distance segmentation matrix output by the surface segmentation model, where the distance segmentation matrix is composed of the distance segmentation values between each position point and the plane to be located.

6. The method according to claim 5, wherein The step of selecting some position points as plane fitting points according to the respective position point feature data, and determining the plane equation to be located according to the plane fitting points includes: Screening each position point according to a preset first segmentation threshold and each distance segmentation value, and taking the position points that meet the segmentation conditions as plane fitting points; Fitting each plane fitting point through a preset fitting algorithm to obtain the plane equation to be located.

7. The method according to claim 5, characterized in that Before obtaining the three-dimensional image data of the part to be located, it further includes: Obtaining historical three-dimensional image data and the plane parameters of the plane to be located corresponding to the historical three-dimensional image data; Calculating a historical distance matrix corresponding to the historical three-dimensional image data according to the historical three-dimensional image data and the plane parameters, where the historical distance matrix is composed of the distance values between each position point in the historical three-dimensional image data and the plane to be located; Segmenting the historical distance matrix according to a preset second segmentation threshold to obtain a historical distance segmentation matrix, where the historical distance segmentation matrix is composed of the segmentation distance values between each position point in the historical three-dimensional image data and the plane to be located; Generating a training sample set based on the historical three-dimensional image data and the historical segmentation distance matrix, and using the training sample set to train a pre-established surface segmentation model to obtain a trained surface segmentation model.

8. A scanning and positioning device is applicable to the situation of performing positioning scans on various medical imaging parts through various scanning modalities, and is characterized in that Including: A data acquisition module, configured to acquire three-dimensional image data of a part to be located; A feature data module, configured to input the three-dimensional image data into a pre-trained position point feature model to obtain the feature data of each position point output by the position point feature model, where the feature data of each position point includes the coordinates of each position point and the distance parameter between each position point and the plane to be located; A plane determination module, configured to screen out some position points as plane fitting points according to the feature data of each position point, and determine the plane equation of the plane to be located according to the plane fitting points.

9. A computer device, characterized in that, The device includes: One or more processors; A storage device, configured to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the scanning and positioning method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the scanning and positioning method according to any one of claims 1-7.

Images