Portable virtual reality medical information display radiotherapy auxiliary system

Abstract

The invention discloses a portable virtual reality medical information display radiotherapy auxiliary system, which relates to the field of radiotherapy and comprises an image acquisition module, a modeling module, a feature point identification module, an analysis module, a focus area determination module and a VR display module. Determining feature points and key points of a patient target area through an image acquisition module, a modeling module and a feature point identification module, and analyzing an association relationship between the feature points and the key points through an analysis module; the VR display module obtains real-time 3D structured light data of a patient, key point coordinates of the body surface of the patient are determined through the modeling module and the feature point recognition module, the human body surface contour and the focus area contour are determined through the focus area determination module and the analysis module, and the VR display module displays the contour and assists a doctor in adjusting the position of the medical linear accelerator. According to the invention, more accurate positioning guidance can be provided for the radiotherapy process, and the radiotherapy efficiency and precision are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radiotherapy, and more particularly to a portable virtual reality medical information display radiotherapy auxiliary system. Background Art

[0002] Radiotherapy, as a key means of cancer treatment, is suitable for 50%-65% of cancer patients. The core equipment of modern radiotherapy is the medical linear accelerator, and the effect of radiotherapy is highly dependent on the positioning accuracy of the medical linear accelerator.

[0003] However, traditional positioning methods have problems such as large errors, long time consumption, and insufficient precision. For example, the initial positioning based on lasers and body surface markers is easily affected by the blurring of markers, resulting in amplified errors; although the use of electronic portal imaging (EPID) and cone beam CT (CBCT) for positioning correction can provide higher accuracy, it increases the additional ionizing radiation dose, bringing potential normal tissue damage and secondary tumor risks; and CBCT can only reflect the positioning status at the time of scanning, lacking real-time performance. Inaccurate positioning may lead to delays or repeated adjustments in the treatment process. Accurate radiotherapy positioning is of great significance for improving tumor control rate, reducing normal tissue damage and reducing the risk of recurrence.

[0004] Therefore, how to provide more accurate positioning guidance for the radiotherapy process is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a portable virtual reality medical information display radiotherapy auxiliary system. When determining the positioning of a medical linear accelerator, a VR display device is used to display the virtual outline of the lesion area in real time, thereby providing more accurate positioning guidance for the radiotherapy process.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention discloses a portable virtual reality medical information display radiotherapy auxiliary system, comprising: an image acquisition module, a modeling module, a feature point recognition module, an analysis module, a lesion area determination module, and a VR display module;

[0008] The image acquisition module synchronously acquires the patient's CT image data and 3D structured light data;

[0009] The modeling module establishes a three-dimensional model of the target area according to the CT image data, and establishes a three-dimensional model of the patient's body surface according to the 3D structured light data;

[0010] The feature point recognition module recognizes the feature points of the target area three-dimensional model and the key points of the body surface three-dimensional model respectively;

[0011] The analysis module analyzes the association relationship between the feature points and the key points;

[0012] The lesion area determination module is used for the doctor to select a number of feature points from the feature points to determine the final lesion area;

[0013] The VR display module acquires the patient's real-time 3D structured light data and sends it to the modeling module for real-time modeling; the feature point recognition module recognizes the key point coordinates in the real-time modeling and sends them to the analysis module; the analysis module fits the human body surface contour according to the received key point coordinates, determines the feature point coordinates of the lesion area according to the received key point coordinates and the association relationship, and fits the lesion area contour; the VR display module synchronously displays the human body surface contour and the lesion area contour to assist doctors in adjusting the positioning of the medical linear accelerator.

[0014] Further, the CT image data and the 3D structured light data are obtained by collecting a CT image of the target area during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient, and a 3D structured light image of the body surface during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient;

[0015] The target area three-dimensional model and the body surface three-dimensional model are both three-dimensional point cloud models, and both include two periods: an exhalation pause and an inhalation pause.

[0016] Furthermore, the feature points and the key points are identified by the following steps:

[0017] Step 1: Select any point P in the 3D model. i , and determine its neighboring point set P i-k ;

[0018] Step 2: construct a triangulated mesh and calculate P according to the neighboring point set determined in step 1. i The discrete normal vector and discrete curvature of ;

[0019] Step 3, constructing a gradient matrix and calculating eigenvalues ​​according to the distance between each point in the neighboring point set, as well as the discrete normal vector and the discrete curvature, and identifying the feature points and the key points according to the set threshold.

[0020] Furthermore, the neighboring point set is obtained in the following manner:

[0021] Step 1.1, set the range of k values ​​and find the distance P i The nearest k+10 neighbors;

[0022] Step 1.2, calculate the relationship between any two adjacent points and P i The angles of the composed vectors are sorted;

[0023] Step 1.3: Delete the distance P between the two adjacent points with the smallest angle. i A point far away;

[0024] Step 1.4: Determine whether the current number of neighboring points is k. If so, take all the current k neighboring points as the neighboring point set P. i-k ; If not, return to step 1.2.

[0025] Furthermore, the construction of the triangulated mesh is specifically as follows: i-k Select and P i The nearest point P i-k-1 , and in P i-k Select the remaining points that match P i-k-1 The closest point P i-k-2 , with P i , P i-k-1 , P i-k-2 Form the first triangle; at P i-k Select the remaining points that match P i-k-2 The closest point P i-k-3 Form the second triangle; construct triangles one by one until the last point P i-k-k , and finally P i , P i-k-k , P i-k-1 Construct the final triangle.

[0026] Furthermore, the discrete normal vector is calculated in the following manner:

[0027] According to the side lengths of all triangles in the triangulated mesh, the shape weight of each triangle is calculated using the formula:

[0028]

[0029] Among them, Δ j represents the shape weight of the jth triangle, a, b, c are the lengths of the three sides of the jth triangle respectively;

[0030] According to the vertex coordinate values ​​of all triangles in the triangulated mesh, the normal vector of each triangle is calculated, and according to the shape weight of each triangle, P is calculated. i The discrete normal vector The formula is:

[0031]

[0032] Among them, N fj represents the normal vector of the j-th triangle, and k is the total number of triangles, which is the same as the total number of points in the neighboring point set.

[0033] Furthermore, the discrete curvature is calculated in the following manner:

[0034] According to N pi and P i-k Perform parametric surface fitting on all points in the curve to obtain the surface S(u,v), where u and v are the surface parameters obtained by the least squares method; perform Taylor expansion on the surface S(u,v) and calculate P i The discrete first-order derivative {S u ,S v} and the discrete second-order derivative {S uu ,S uv ,S vv};

[0035] According to the discrete first-order derivative, the discrete second-order derivative and the discrete normal vector N p , calculate the outer Engarten transformation matrix W, the formula is:

[0036]

[0037] Among them, E, F, G are the first type of basic quantities of the surface, and L, N, M are the second type of basic quantities of the surface; the eigenvalue κ of W is calculated based on the first type of basic quantities and the second type of basic quantities 1 , κ 2 , the discrete curvature κ is obtained, the formula is: κ=κ 1 κ 2 .

[0038] Furthermore, constructing a gradient matrix and calculating eigenvalues ​​includes:

[0039] According to the neighboring points, each point is concentrated and P i The distance, and the discrete normal vector and the discrete curvature calculation window function;

[0040] Calculate P according to the window function i The gradient change along the vector direction of any point in the neighboring point set;

[0041] Calculate P according to the gradient change and the discrete first-order derivative i The overall gradient of P i The gradient matrix of the point and calculate the three eigenvalues ​​λ′ of the gradient matrix 1 ,λ′ 2 ,λ′ 3 .

[0042] Furthermore, the identifying of the feature points and the key points according to the set threshold value is specifically as follows: setting the threshold value Ω and determining P according to the judgment condition i Whether it is a key point;

[0043] The judgment conditions are:

[0044] Condition 1, λ′ 1 ≥λ′ 2 ≥λ′ 3 ≥Ω;

[0045] Condition 2, λ′ 1 ≥λ′ 2 ,λ′ 1 ≥Ω,λ′ 3 ≠0;

[0046] Condition 3, λ′ 2 ≥λ′ 1 ,λ′ 2 ≥Ω,λ′ 3 ≠0;

[0047] When the eigenvalue λ′ 1 ,λ′ 2 ,λ′ 3 If any of conditions 1, 2, and 3 are met, then P i For the key point.

[0048] Furthermore, the analysis of the association relationship between the feature points and the key points is specifically as follows:

[0049] According to the three-dimensional model of the target area during exhalation pause and the three-dimensional model of the target area during inhalation pause, the feature point determination module is used to determine the feature points during exhalation pause and the feature points during inhalation pause and the corresponding time coordinates, and calculate the first displacement vector of the same feature point during exhalation pause and inhalation pause;

[0050] According to the three-dimensional body surface model during exhalation pause and the three-dimensional body surface model during inhalation pause, the feature point determination module is used to determine the key points during exhalation pause and the key points during inhalation pause and the corresponding time coordinates, and calculate the second displacement vector of the same key point during exhalation pause and inhalation pause;

[0051] The vector distance between each of the first displacement vectors and each of the second displacement vectors is calculated using the formula:

[0052] d′=sqrt((Ax-Bx) 2 +(Ay-By) 2 +(Az-Bz) 2 );

[0053] Wherein, (Ax, Ay, Az), (Bx, By, Bz) represent the first displacement vector and the second displacement vector respectively, and d′ is the distance between the displacement vectors;

[0054] The vector distances between any feature point and each key point are sorted, and the five key points with the smallest vector distances are selected as the associated key points of the feature point, and the three-dimensional coordinate difference between each associated key point and the feature point is confirmed; the associated key points and the corresponding three-dimensional coordinate differences of all feature points are determined one by one to obtain the association relationship.

[0055] Furthermore, the determining of the feature point coordinates of the lesion area according to the received key point coordinates and the association relationship is specifically:

[0056] According to the association relationship, the five associated key points corresponding to each feature point of the lesion area are determined; according to the coordinates of the five associated key points corresponding to the feature point of any lesion area, and the three-dimensional coordinate difference between the associated key point and the feature point, the five theoretical coordinates of the feature point are calculated, and the average coordinate of the five theoretical coordinates is calculated as the coordinate value of the feature point; the coordinate values ​​of all the feature points of the lesion area are calculated one by one.

[0057] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a portable virtual reality medical information display radiotherapy auxiliary system, which can accurately determine the characteristic points of the patient's target area and the key points of the body surface by synchronously acquiring the patient's CT image data and 3D structured light data, and establishing a three-dimensional model of the target area and a three-dimensional model of the patient's body surface; by analyzing the correlation between the characteristic points and the key points, when adjusting the position of the medical linear accelerator, the virtual lesion area contour can be displayed through the VR device to provide auxiliary guidance for the doctor. The present invention can provide more accurate positioning guidance for the radiotherapy process and improve the efficiency and accuracy of radiotherapy treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0059] Figure 1 It is a schematic diagram of the overall structure of an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] The embodiment of the present invention discloses a portable virtual reality medical information display radiotherapy auxiliary system, such as Figure 1 As shown, it includes: an image acquisition module, a modeling module, a feature point recognition module, an analysis module, a lesion area determination module, and a VR display module;

[0062] The image acquisition module synchronously acquires the patient's CT image data and 3D structured light data;

[0063] The modeling module builds a 3D model of the target area based on CT image data and a 3D model of the patient's body surface based on 3D structured light data;

[0064] The feature point recognition module recognizes the feature points of the three-dimensional model of the target area and the key points of the three-dimensional model of the body surface respectively;

[0065] The analysis module analyzes the correlation between feature points and key points;

[0066] The lesion area determination module is used for doctors to select several feature points from the feature points to determine the final lesion area;

[0067] The VR display module acquires the patient's real-time 3D structured light data and sends it to the modeling module for real-time modeling; the feature point recognition module identifies the key point coordinates in the real-time modeling and sends them to the analysis module; the analysis module fits the human body surface contour according to the received key point coordinates, determines the feature point coordinates of the lesion area according to the received key point coordinates and the correlation relationship, and fits the lesion area contour; the VR display module synchronously displays the human body surface contour and the lesion area contour to assist doctors in adjusting the position of the medical linear accelerator.

[0068] In a specific embodiment, CT image data and 3D structured light data are obtained by acquiring a CT image of a target area during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient, and a 3D structured light image of the body surface during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient;

[0069] The target area three-dimensional model and the body surface three-dimensional model are both three-dimensional point cloud models, and both include two periods: expiratory pause and inspiratory pause.

[0070] In a specific embodiment, the feature points and key points are identified by the following steps:

[0071] Step 1: Select any point P in the 3D model. i , and determine its neighboring point set P i-k ;

[0072] Step 2: Based on the neighboring point set determined in step 1, construct a triangulated mesh and calculate P i The discrete normal vector and discrete curvature of ;

[0073] Step 3: According to the distance between each point in the neighboring point set, as well as the discrete normal vector and discrete curvature, a gradient matrix is ​​constructed and the eigenvalue is calculated, and the feature points and key points are identified according to the set threshold.

[0074] In a specific embodiment, the neighboring point set is obtained in the following manner:

[0075] Step 1.1, set the range of k values ​​and find the distance P i The nearest k+10 neighbors;

[0076] Step 1.2, calculate the relationship between any two adjacent points and P i The angles of the composed vectors are sorted;

[0077] Step 1.3: Delete the distance P between the two adjacent points with the smallest angle. i A point far away;

[0078] Step 1.4: Determine whether the current number of neighboring points is k. If so, take all the current k neighboring points as the neighboring point set P. i-k ; If not, return to step 1.2.

[0079] By further determining the angle based on finding the nearest neighboring point, the neighboring point set can be more evenly distributed in P i Surroundings, improve the accuracy of feature point and key point recognition.

[0080] In a specific embodiment, constructing a triangulated mesh is as follows: i-k Select and P i The nearest point P i-k-1 , and in P i-k Select the remaining points that match P i-k-1 The closest point P i-k-2 , with P i , P i-k-1 , P i-k-2 Form the first triangle; at P i-k Select the remaining points that match P i-k-2 The closest point P i-k-3 Form the second triangle; construct triangles one by one until the last point P i-k-k , and finally P i , P i-k-k , P i-k-1 Construct the last triangle, that is, connect each triangle one by one to obtain a triangulated mesh consisting of a set of adjacent points.

[0081] In a specific embodiment, the discrete normal vector N p Calculated by:

[0082] According to the side lengths of all triangles in the triangulated mesh, the shape weight of each triangle is calculated using the formula:

[0083]

[0084] Among them, Δ j represents the shape weight of the jth triangle, a, b, c are the lengths of the three sides of the jth triangle respectively;

[0085] According to the vertex coordinates of all triangles in the triangulated mesh, the normal vector of each triangle is calculated, and according to the shape weight of each triangle, P is calculated. i The discrete normal vector The formula is:

[0086]

[0087] Among them, N fj represents the normal vector of the j-th triangle, and k is the total number of triangles, which is the same as the total number of points in the neighboring point set.

[0088] In a specific embodiment, the discrete curvature is calculated as follows:

[0089] According to N pi and P i-k Perform parametric surface fitting on all points in the curve to obtain the surface S(u,v), where u and v are the surface parameters obtained by the least squares method; perform Taylor expansion on the surface S(u,v) and calculate P i The discrete first-order derivative {S u ,S v} and the discrete second-order derivative {S uu ,S uv ,S vv};

[0090] According to the discrete first-order derivative, discrete second-order derivative and discrete normal vector N p , calculate the outer Engarten transformation matrix W, the formula is:

[0091]

[0092] Among them, E, F, G are the first basic quantities of the surface, and L, N, M are the second basic quantities of the surface. The eigenvalue κ of W is calculated based on the first and second basic quantities. 1 , κ 2 , we get the discrete curvature κ, the formula is: κ=κ 1 κ 2 .

[0093] Specifically, calculate the characteristic polynomial of W: det(W-λI)=0;

[0094] After expansion, we get: 2 -(E / L+G / N)λ+(EG / LN-F 2 / LM)=0;

[0095] After solving, we get:

[0096]

[0097] Among them, λ represents the eigenvalue, I represents the unit matrix, and λ 1 , 2 are two solutions of λ, eigenvalue κ 1 , κ 2 is: 1 =(λ 1 +λ 2 ) / 2,κ 2 =(λ 1 -λ 2 ) / 2;

[0098] In a specific embodiment, constructing a gradient matrix and calculating eigenvalues ​​includes:

[0099] According to the neighboring points, each point is concentrated and P i The distance, as well as the discrete normal vector and discrete curvature calculation window function;

[0100] Calculate P based on the window function i The gradient change along the vector direction of any point in the neighboring point set;

[0101] Calculate P based on gradient changes and discrete first-order derivatives i The overall gradient of P i The gradient matrix of the point and calculate the three eigenvalues ​​λ′ of the gradient matrix 1 ,λ′ 2 ,λ′ 3 .

[0102] Specifically, the window function formula is:

[0103]

[0104] Among them, w j (d j ,θ j ,κ j ) indicates P i-k The window function value of the jth point in ; d j ,θ j ,κ j are the distance parameter, normal vector parameter and discrete curvature parameter of the function respectively, d j =P i -P i-k-j ,θj =N pi -N pi-k-j , κ j =κ i -κ i-k-j , P i-k-j 、N pi-k-j , κ i-k-j P i-k The coordinates, discrete normal vector, and discrete curvature of the jth point in ; 2σ d =max(d i-j ) indicates P i-k The maximum distance between all points in the equation and Pi, 2σ N =π,2σ κ =max(κ i -κ i-k ) indicates P i-k All points in P i The maximum value of the discrete curvature difference.

[0105] Through the window function, calculate P i Along P i-k The gradient change E of the vector direction at any point j (u,v), the formula is:

[0106]

[0107] P i Substitute the discrete first-order derivative of into the above formula to calculate P i In P i-k The overall gradient within is given by:

[0108]

[0109] Rewrite the above formula to calculate the neighboring point set P i-k Direction cosines cosα, cosβ, cosγ in local coordinate system x, y, z, obtain P i The covariance matrix of the point gradient is as follows:

[0110]

[0111]

[0112] Among them, the local coordinate system constructed with the intersection of the discrete normal vector and the surface S(u,v) as the origin, the obtained P i Point coordinates (S x , S y , S z ), M(P i ) is the gradient matrix.

[0113] M(Pi ) is: det(M(P i )-λ′I)=0, after expansion, we get:

[0114] det(S x -S z -λ′,S x -S y ,S x -S s

[0115] S x -S y ,S y -S z -λ′,S y -S s ;

[0116] S x -S s ,S y -S s ,S s -S z -λ′)=0

[0117] After solving, we get three eigenvalues ​​λ′ 1 ,λ′ 2 ,λ′ 3 ;

[0118] Where λ′ represents M(P i ), I represents the identity matrix, λ′ 1 ,λ′ 2 ,λ′ 3 are the three solutions of λ′, namely the gradient matrix M(P i )’s eigenvalues.

[0119] In a specific embodiment, the feature points and key points are identified according to the set threshold, specifically: the threshold Ω is set and P is determined according to the judgment condition. i Whether it is a key point;

[0120] The judgment conditions are:

[0121] Condition 1, λ′ 1 ≥λ′ 2 ≥λ′ 3 ≥Ω;

[0122] Condition 2, λ′ 1 ≥λ′ 2 ,λ′ 1 ≥Ω,λ′ 3 ≠0;

[0123] Condition 3, λ′ 2 ≥λ′1 ,λ′ 2 ≥Ω,λ′ 3 ≠0;

[0124] When the eigenvalue λ′ 1 ,λ′ 2 ,λ′ 3 If any of conditions 1, 2, and 3 are met, then P i For the key point.

[0125] In a specific embodiment, the correlation relationship between the characteristic points and the key points is analyzed as follows:

[0126] According to the three-dimensional model of the target area during the exhalation pause and the three-dimensional model of the target area during the inhalation pause, the feature point determination module is used to determine the feature points during the exhalation pause and the feature points during the inhalation pause and the corresponding time coordinates, and the first displacement vector of the same feature point during the exhalation pause and the inhalation pause is calculated;

[0127] According to the three-dimensional body surface model during exhalation pause and the three-dimensional body surface model during inhalation pause, the key points during exhalation pause and the key points during inhalation pause and the corresponding time coordinates are determined by using the feature point determination module, and the second displacement vector of the same key point during exhalation pause and inhalation pause is calculated;

[0128] The vector distance between each first displacement vector and each second displacement vector is calculated using the formula:

[0129] d′=sqrt((Ax-Bx) 2 +(Ay-By) 2 +(Az-Bz) 2 );

[0130] Wherein, (Ax, Ay, Az), (Bx, By, Bz) represent the first displacement vector and the second displacement vector respectively, and d′ is the distance between the displacement vectors;

[0131] The vector distances between any feature point and each key point are sorted, and the five key points with the smallest vector distances are selected as the associated key points of the feature point, and the three-dimensional coordinate difference between each associated key point and the feature point is confirmed; the associated key points of all feature points and the corresponding three-dimensional coordinate differences are determined one by one to obtain the association relationship.

[0132] Specifically, the patient's breathing during radiotherapy will affect the patient's body surface and the position of the lesion area. The exhalation pause is the pause moment from the exhalation process to the inhalation process, and the inhalation pause is the pause moment from the inhalation process to the exhalation process. By analyzing and comparing the movement changes of each feature point and key point in the lesion area at the two moments, the key points on the body surface that are more relevant to the movement changes of the feature points in the lesion area can be effectively determined, so that the coordinates of the feature points in the lesion area can be more accurately determined from the coordinates of the key points on the body surface. In order to improve the accuracy of the fitted lesion area contour, the doctor can better adjust the positioning of the linear accelerator according to the lesion area contour and body surface contour displayed in real time by the VR device and the patient's breathing status, improve the positioning accuracy, so that the radiation can act on the lesion area as much as possible, improve the tumor control rate, and reduce normal tissue damage.

[0133] In a specific embodiment, the feature point coordinates of the lesion area are determined according to the received key point coordinates and the association relationship, specifically:

[0134] According to the association relationship, the five associated key points corresponding to each feature point of the lesion area are determined; according to the coordinates of the five associated key points corresponding to the feature point of any lesion area, and the three-dimensional coordinate difference between the associated key point and the feature point, the five theoretical coordinates of the feature point are calculated, and the average coordinate of the five theoretical coordinates is calculated as the coordinate value of the feature point; the coordinate values ​​of all the feature points of the lesion area are calculated one by one.

[0135] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0136] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A portable virtual reality medical information display radiotherapy auxiliary system, characterized in that: include: Image acquisition module, modeling module, feature point recognition module, analysis module, lesion area determination module, VR display module; The image acquisition module synchronously acquires CT image data and 3D structured light data of the patient; The modeling module establishes a three-dimensional model of the target area according to the CT image data, and establishes a three-dimensional model of the patient's body surface according to the 3D structured light data; The feature point recognition module recognizes the feature points of the target area three-dimensional model and the key points of the body surface three-dimensional model respectively; The analysis module analyzes the association relationship between the feature points and the key points; The lesion area determination module is used for the doctor to select a number of feature points from the feature points to determine the final lesion area; The VR display module acquires the real-time 3D structured light data of the patient and sends it to the modeling module for real-time modeling; The feature point recognition module recognizes the key point coordinates in real-time modeling and sends them to the analysis module; The analysis module fits the human body surface contour according to the received key point coordinates, determines the feature point coordinates of the lesion area according to the received key point coordinates and the association relationship, and fits the lesion area contour; the VR display module synchronously displays the human body surface contour and the lesion area contour to assist the doctor in adjusting the position of the medical linear accelerator.

2. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 1, characterized in that: The CT image data and the 3D structured light data are obtained by collecting a CT image of the target area during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient, and a 3D structured light image of the body surface during an expiratory pause and an inspiratory pause in at least one complete respiratory cycle of the patient. The target area three-dimensional model and the body surface three-dimensional model are both three-dimensional point cloud models, and both include two periods: an exhalation pause and an inhalation pause.

3. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 1, characterized in that: The feature points and the key points are identified by the following steps: Step 1: Select any point P in the 3D model. i , and determine its neighboring point set P i-k ; Step 2: construct a triangulated mesh and calculate P according to the neighboring point set determined in step 1. i The discrete normal vector and discrete curvature of ; Step 3, constructing a gradient matrix and calculating eigenvalues ​​according to the distance between each point in the neighboring point set, as well as the discrete normal vector and the discrete curvature, and identifying the feature points and the key points according to the set threshold.

4. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 3, characterized in that: The neighboring point set is obtained in the following way: Step 1.1, set the range of k values ​​and find the distance P i The nearest k+10 neighbors; Step 1.2, calculate the relationship between any two adjacent points and P i The angles of the composed vectors are sorted; Step 1.3, delete the two neighboring points with the smallest angle between them and P. i A point far away; Step 1.4: Determine whether the current number of neighboring points is k. If so, take all the current k neighboring points as the neighboring point set P. i-k ; If not, return to step 1.

2.

5. The portable virtual reality medical information display radiotherapy auxiliary system according to claim 3, characterized in that: The discrete normal vector is calculated as follows: According to the side lengths of all triangles in the triangulated mesh, the shape weight of each triangle is calculated using the formula: Among them, Δ j represents the shape weight of the jth triangle, a, b, c are the lengths of the three sides of the jth triangle respectively; According to the vertex coordinate values ​​of all triangles in the triangulated mesh, the normal vector of each triangle is calculated, and according to the shape weight of each triangle, P is calculated. i The discrete normal vector The formula is: Among them, N fj represents the normal vector of the j-th triangle, and k is the total number of triangles, which is the same as the total number of points in the neighboring point set.

6. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 5, characterized in that: The discrete curvature is calculated as follows: according to and P i-k Perform parametric surface fitting on all points in the curve to obtain the surface S(u,v), where u and v are the surface parameters obtained by the least squares method; perform Taylor expansion on the surface S(u,v) and calculate P i The discrete first-order derivative {S u ,S v } and the discrete second-order derivative {S uu ,S uv ,S vv }; According to the discrete first-order derivative, the discrete second-order derivative and the discrete normal vector N p , calculate the outer Engarten transformation matrix W, the formula is: Among them, E, F, and G are the first type of basic quantities of the surface, and L, N, and M are the second type of basic quantities of the surface; the eigenvalues ​​κ1 and κ2 of W are calculated according to the first type of basic quantities and the second type of basic quantities to obtain the discrete curvature κ, and the formula is: κ=κ1·κ2.

7. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 6, characterized in that: The step of constructing a gradient matrix and calculating eigenvalues ​​includes: According to the neighboring points, each point is concentrated and P i The distance, and the discrete normal vector and the discrete curvature calculation window function; Calculate P according to the window function i The gradient change along the vector direction of any point in the neighboring point set; Calculate P according to the gradient change and the discrete first-order derivative i The overall gradient of P i The gradient matrix of the point is obtained and the three eigenvalues ​​λ1′, λ2′, and λ3′ of the gradient matrix are calculated.

8. The portable virtual reality medical information display radiotherapy auxiliary system according to claim 7, characterized in that: The identifying of the feature points and the key points according to the set threshold value is specifically as follows: setting the threshold value Ω, and determining P according to the judgment condition i Whether it is a key point; The judgment conditions are: Condition 1, λ1′≥λ2′≥λ3′≥Ω; Condition 2, λ1′≥λ2′, λ1′≥Ω, λ3′≠0; Condition 3, λ2′≥λ1′, λ2′≥Ω, λ3′≠0; When the eigenvalues ​​λ1′, λ2′, and λ3′ meet any of the conditions 1, 2, and 3, then P is judged i For the key point.

9. The portable virtual reality medical information display radiotherapy auxiliary system according to claim 2, characterized in that: The analysis of the association relationship between the feature points and the key points is specifically as follows: According to the three-dimensional model of the target area during exhalation pause and the three-dimensional model of the target area during inhalation pause, the feature point determination module is used to determine the feature points during exhalation pause and the feature points during inhalation pause and the corresponding time coordinates, and calculate the first displacement vector of the same feature point during exhalation pause and inhalation pause; According to the three-dimensional body surface model during exhalation pause and the three-dimensional body surface model during inhalation pause, the feature point determination module is used to determine the key points during exhalation pause and the key points during inhalation pause and the corresponding time coordinates, and calculate the second displacement vector of the same key point during exhalation pause and inhalation pause; The vector distance between each of the first displacement vectors and each of the second displacement vectors is calculated using the formula: d′=sqrt((Ax-Bx) 2 +(Ay-By) 2 +(Az-Bz) 2 ); Wherein, (Ax, Ay, Az), (Bx, By, Bz) represent the first displacement vector and the second displacement vector respectively, and d′ is the distance between the displacement vectors; The vector distances between any feature point and each key point are sorted, and the five key points with the smallest vector distances are selected as the associated key points of the feature point, and the three-dimensional coordinate difference between each associated key point and the feature point is confirmed; the associated key points and the corresponding three-dimensional coordinate differences of all feature points are determined one by one to obtain the association relationship.

10. A portable virtual reality medical information display radiotherapy auxiliary system according to claim 9, characterized in that: The determining of the feature point coordinates of the lesion area according to the received key point coordinates and the association relationship is specifically: According to the association relationship, the five associated key points corresponding to each feature point of the lesion area are determined; according to the coordinates of the five associated key points corresponding to the feature point of any lesion area, and the three-dimensional coordinate difference between the associated key point and the feature point, the five theoretical coordinates of the feature point are calculated, and the average coordinate of the five theoretical coordinates is calculated as the coordinate value of the feature point; the coordinate values ​​of all the feature points of the lesion area are calculated one by one.

Images