CN112837246A - Geometric correction method for image guidance system - Google Patents

Abstract

The invention discloses a geometric correction method of an image guidance system, which comprises the following steps: firstly, establishing a high-precision die body, wherein metal mark points under a known three-dimensional space coordinate system are fixed on the die body; secondly, respectively projecting and imaging in the emitting directions of two X-ray light sources of the projection system based on the phantom, and recording projection coordinates of phantom mark points on a detector under a two-dimensional coordinate system; then, respectively obtaining correct mapping relations in two directions, namely mapping matrixes, by a linear direct transformation method; and finally, applying the matching mapping relation from the three-dimensional coordinates to the two-dimensional coordinates to a projection reconstruction algorithm to obtain a reconstructed image based on a corrected reconstruction algorithm. The invention overcomes the problem of geometric artifacts of reconstructed images of an image guide system, improves the quality of the reconstructed images and greatly improves the precision based on image guide registration.

Description

Geometric correction method for image guidance system

Technical Field

The invention relates to the field of medical treatment, in particular to a geometric correction method of an image guidance system.

Background

In image-guided radiotherapy, registration of pre-treatment data with image data of a patient acquired during treatment is a crucial step in providing a physician with comprehensive and useful information. Generally, two-dimensional images (such as an ultrasound image and an X-ray image) are easier to acquire during a treatment process, can reflect the position information of a patient at present, are fast in acquisition speed, and can reduce the radiation exposure time of the patient and a doctor, but three-dimensional spatial information and anatomical information contained in CT and MRI are lacked in the two-dimensional information, so that three-dimensional data are required to be subjected to interventional therapy so as to obtain more visual and detailed image information similar to a three-dimensional system, and in the process, the two-dimensional X-ray image needs to be registered with three-dimensional CT volume data, namely 2D/3D image registration.

The image guidance system is a hardware basis for 2D/3D image registration, the image guidance system mainly adopts a ray algorithm to carry out two-dimensional reconstruction on an object, and an ideal reconstruction geometric model requires that a central ray must pass through a rotating shaft and fall on the central position of an imaging area and be vertical to a detector. However, in the actual mechanical installation process, because the precision of the artificial machine is not up to the requirement, some geometric position deviations often occur, and these geometric errors can be attributed to geometric errors of the displacement and inclination of the detector, including a horizontal deviation Δ U and a vertical deviation Δ V of the detector, a rotation angle θ of the detector in the direction of the winding row, a rotation angle Φ of the detector in the direction of the winding column, and a rotation angle η of the detector in the direction of the winding Z axis. Geometrical parameters describing geometrical characteristics of the image-guided system include: (1) distance SAD from ray source to isocenter, (2) distance SDD from ray source to detector, and (3) abscissa U of vertical foot position of ray source on detector₀(4) ordinate V of vertical foot position of radiation source on detector₀(5) the angle theta of rotation of the detector along the U-axis, (6) the angle phi of rotation of the detector along the V-axis, and (7) the angle eta of rotation of the detector along the center point.

These geometric deviations can cause the actual image to be mismatched with the ideal reconstructed geometric model, thereby causing geometric artifacts on the reconstructed image, which severely affect the quality of the reconstructed image and significantly reduce the resolution of the reconstructed image.

Disclosure of Invention

The purpose of the invention is as follows: the invention aims to provide an image guidance system geometric correction method which can overcome the problem of geometric artifact of a reconstructed image of an image guidance system, improve the quality of the reconstructed image and improve the accuracy based on image guidance registration.

The technical scheme is as follows: the invention relates to a geometric correction method of an image guidance system, which comprises the following steps:

the method comprises the following steps: establishing a high-precision die body, wherein metal mark points under a known three-dimensional space coordinate system are fixed on the die body;

step two: respectively projecting and imaging in the emitting directions of two X-ray light sources of a projection system based on the phantom, and recording projection coordinates of phantom marking points on a detector under a two-dimensional coordinate system;

step three: respectively obtaining correct mapping relations in two directions, namely mapping matrixes, by a linear direct transformation method;

step four: and applying the matching mapping relation from the three-dimensional coordinates to the two-dimensional coordinates to a projection reconstruction algorithm to obtain a reconstructed image based on a corrected reconstruction algorithm.

The die body is made of organic glass with a pentahedron shape, a plurality of metal small balls are embedded on the surface of the die body, and a clear metal shadow can be projected under the irradiation of X rays.

In the invention, the mapping matrix satisfies the formula [ uw vw w]^T＝P[x y z 1]^TAnd P is a 3 multiplied by 4 mapping matrix, w is a distance weight factor, x, y and z are point coordinates of a three-dimensional space, u and v are point coordinates on a detector, a correction data set is generated through the marker points of the phantom spheres, an over-determined equation set related to the P matrix is created, and the mapping matrix is obtained through solving.

The mapping matrix P satisfies the formula P ═ K [ R | t ═ K]Where K is the 3 × 3 intra-system parameter matrix, R is the 3 × 3 rotation matrix, t is the 3 × 1 translation matrix, R is represented by RzRyRx, and the left 3 × 3 sub-matrix of the mapping matrix P is obtained by multiplying the intra-parameter matrix K and the rotation matrix, i.e. P_3×3KR, the geometric parameter of the actual detector is denoted U₀＝K₁₃，V₀＝K₂₃， SDD＝K₁₁P_xWhere Px is the detector pixel size, and the rotation angle θ is atan2 (R)₃₂，R₃₃)，

η＝atan2(R₁₂，R₁₁) Translation vector T_z＝P₃₄，T_y＝(P₂₄-K₂₃P₃₄)/K₂₂， T_x＝(P₁₄-T_yP₁₂-K₂₃P₃₄)/K₁₁。

The invention has the beneficial effects that: compared with the prior art, the invention has the following remarkable advantages: the invention considers the shortage of the manual installation precision of the system to cause the system to generate geometric deviation, overcomes the problem of geometric artifact of the reconstructed image of the image guide system, improves the quality of the reconstructed image and greatly improves the precision based on image guide registration.

Drawings

FIG. 1 is a flow chart of the present invention;

FIG. 2 is an idealized geometric model of the present invention;

FIG. 3 is a practical geometric model of the present invention;

FIG. 4 is a schematic view of the detector translational misalignment of the present invention;

FIG. 5 is a schematic view of the detector of the present invention rotated about a central longitudinal axis;

FIG. 6 is a schematic view of the detector of the present invention rotated about a central horizontal axis;

FIG. 7 is a schematic view of the detector of the present invention rotated about the Z-axis;

FIG. 8 is a diagram illustrating the distribution of the mark points according to the present invention.

Detailed Description

The technical solution of the present invention is further explained with reference to fig. 1-8.

An image-guided system geometry correction method as shown in the figure, comprising the steps of:

the method comprises the following steps: establishing a high-precision die body, wherein metal mark points under a known three-dimensional space coordinate system are fixed on the die body;

step two: respectively projecting and imaging in the emitting directions of two X-ray light sources of a projection system based on the phantom, and recording projection coordinates of phantom marking points on a detector under a two-dimensional coordinate system;

step three: respectively obtaining correct mapping relations in two directions, namely mapping matrixes, by a linear direct transformation method;

step four: and applying the matching mapping relation from the three-dimensional coordinates to the two-dimensional coordinates to a projection reconstruction algorithm to obtain a reconstructed image based on a corrected reconstruction algorithm.

The mold body is made of organic glass with a pentahedron shape, a plurality of metal small balls are embedded on the surface of the mold body, clear metal shadows can be projected under the irradiation of X rays, as shown in figure 8, the centers of the metal shadows can be accurately marked on a projection drawing, images of metal mark points cannot be overlapped, and the positions of corresponding mold body mark points can be conveniently calculated according to the positions of the projected image mark points.

Mapping the matrix to satisfy the formula [ uw vw w]^T＝P[x y z 1]^TAnd P is a 3 multiplied by 4 mapping matrix, w is a distance weight factor, x, y and z are point coordinates of a three-dimensional space, u and v are point coordinates on a detector, a correction data set is generated through the marker points of the phantom spheres, an over-determined equation set related to the P matrix is created, and the mapping matrix is obtained through solving. The mapping matrix P satisfies the formula P ═ K [ R | t ═ K]Where K is the 3 × 3 intra-system parameter matrix, R is the 3 × 3 rotation matrix, t is the 3 × 1 translation matrix, R is represented by RzRyRx, and the left 3 × 3 sub-matrix of the mapping matrix P is obtained by multiplying the intra-parameter matrix K and the rotation matrix, i.e. P_3×3KR, the geometric parameter of the actual detector is denoted U₀＝K₁₃，V₀＝K₂₃，SDD＝K₁₁P_xWhere Px is the detector pixel size, and the rotation angle θ is atan2 (R)₃₂，R₃₃)，

η＝atan2(R₁₂，R₁₁) Translation vector T_z＝P₃₄，T_y＝(P₂₄-K₂₃P₃₄)/K₂₂，T_x＝(P₁₄-T_yP₁₂-K₂₃P₃₄)/K₁₁。

Claims (5)

1. An image-guided system geometry correction method, comprising the steps of:

the method comprises the following steps: establishing a high-precision die body, wherein metal mark points under a known three-dimensional space coordinate system are fixed on the die body;

step two: respectively projecting and imaging in the emitting directions of two X-ray light sources of a projection system based on the phantom, and recording projection coordinates of phantom marking points on a detector under a two-dimensional coordinate system;

step three: respectively obtaining correct mapping relations in two directions, namely mapping matrixes, by a linear direct transformation method;

step four: and applying the matching mapping relation from the three-dimensional coordinates to the two-dimensional coordinates to a projection reconstruction algorithm to obtain a reconstructed image based on a corrected reconstruction algorithm.

2. An image guidance system geometry correction method according to claim 1, characterized in that the phantom is organic glass in the shape of a pentahedron.

3. An image-guided system geometry correction method according to claim 1, characterized in that said mapping matrix satisfies the formula [ uw vw w]^T＝P[x y z 1]^TAnd P is a 3 multiplied by 4 mapping matrix, w is a distance weight factor, x, y and z are point coordinates of a three-dimensional space, u and v are point coordinates on a detector, a correction data set is generated through the marker points of the phantom spheres, an over-determined equation set related to the P matrix is created, and the mapping matrix is obtained through solving.

4. An image-guided system geometry correction method according to claim 1, characterized in that the mapping matrix P satisfies the formula P ═ K [ R | t ═ K]Where K is the 3 × 3 intra-system parameter matrix, R is the 3 × 3 rotation matrix, t is the 3 × 1 translation matrix, R is represented by RzRyRx, and the left 3 × 3 sub-matrix of the mapping matrix P is obtained by multiplying the intra-parameter matrix K and the rotation matrix, i.e. P_3×3KR, the geometric parameter of the actual detector is denoted U₀＝K₁₃，V₀＝K₂₃，SDD＝K₁₁P_xWhere Px is the detector pixel size, and the rotation angle θ is atan2 (R)₃₂，R₃₃)，

η＝atan2(R₁₂，R₁₁) Translation vector T_z＝P₃₄，T_y＝(P₂₄-K₂₃P₃₄)/K₂₂，T_x＝(P₁₄-T_yP₁₂-K₂₃P₃₄)/K₁₁。

5. An image guidance system geometric correction method according to claim 2, characterized in that, a plurality of metal balls are embedded on the surface of the model body, and a clear metal shadow can be projected under the irradiation of X-rays.

Images