A C-arm calibration method, device and system based on image feature extraction

Abstract

This invention discloses a C-arm calibration method, apparatus, and system based on image feature extraction. The method includes: precisely placing the calibration phantom on... C The imaging area of ​​the shape-arm is exposed under the initial pose to acquire the initial image and extract the initial two-dimensional coordinates of the marker points. Combined with the known three-dimensional coordinates of the marker points, the initial projection parameters are calculated; the driving... C The telescopic arm rotates around at least one axis of its coordinate system to multiple calibration poses, records the angles and exposures, and extracts the calibration 2D coordinates. Based on the rotation angles and initial 3D coordinates, the calibration 3D coordinates under each calibration pose are determined, and then the corresponding projection parameters are calculated. Based on each calibration pose and its projection parameters, a system is established. C The precise mapping relationship between the pose of the manipulator arm and its projection parameters. This invention achieves image feature processing-based results by acquiring images of the calibrated manipulator under different poses and calculating the projection parameters. C The boom provides a precise pose-projection parameter mapping relationship through one-time rapid positioning.

Description

Technical Field

[0001] This invention relates to a C-arm calibration method, apparatus, and system based on image feature extraction, belonging to the field of medical device image calibration technology. Background Technology

[0002] C-arms are widely used X-ray imaging devices in clinical surgeries such as orthopedics. They typically consist of an X-ray source, a flat panel detector, and a C-shaped support arm connecting the two. The C-arm can slide along a slip ring track (move along a C-shaped trajectory) and / or rotate around its own axis to obtain fluoroscopic images of the patient's surgical area from different angles.

[0003] In existing technologies, the positioning of the C-arm during surgery mainly relies on manual adjustment by the surgeon. The surgeon repeatedly adjusts the sliding angle, rotation angle, and spatial position of the C-arm based on experience until a surgical field image that meets the surgeon's requirements is obtained. This manual adjustment method has obvious drawbacks: First, the adjustment process depends on personal experience and often requires multiple trials and errors, which significantly prolongs the operation time; second, each trial and error requires X-ray exposure, increasing the radiation exposure dose for patients and medical staff; and third, the adjustment accuracy is difficult to guarantee, affecting the quality and efficiency of the surgery.

[0004] To achieve intelligent and automated adjustment of the C-arm—that is, to drive the C-arm to the ideal position in one go based on the target image pose—the technological foundation lies in establishing a precise mapping relationship between the C-arm pose (including sliding and rotation angles) and its imaging projection parameters, i.e., completing the C-arm calibration. However, there are two main structural types of C-arms: isocentric and anisocentric. The sliding and rotational movements of an isocentric C-arm revolve around a fixed geometric center (isocentric point), with a standard circular trajectory; while the motion center of an anisocentric C-arm shifts with changes in pose, resulting in an approximately elliptical trajectory. These two structures have fundamental differences in physical kinematics, leading to different patterns in the variation of their projection parameters (such as the distance from the X-ray source to the rotation center, and the distance from the rotation center to the detector) with pose. Existing calibration methods are often designed for C-arms with specific structures, lacking a unified, efficient, and highly accurate universal calibration scheme applicable to both isocentric and anisocentric C-arms.

[0005] Therefore, there is an urgent need for a new C-arm calibration method that can solve the above problems and lay the foundation for achieving precise, rapid, and automatic adjustment of the C-arm. Summary of the Invention

[0006] The purpose of this invention is to provide a C-arm calibration method, device, and system based on image feature extraction. By acquiring images of the calibration phantoms in different poses and calculating the projection parameters, the invention achieves one-time rapid positioning of the C-arm based on image feature processing, providing an accurate pose-projection parameter mapping relationship.

[0007] To achieve the above objectives / to solve the above technical problems, the present invention is implemented using the following technical solution.

[0008] On one hand, the present invention provides a C-arm calibration method based on image feature extraction, comprising the following steps:

[0009] A calibration phantom with pre-arranged marker points is placed in the imaging area of ​​a C-arm. Exposure is performed with the C-arm in its initial pose to acquire an initial image and construct a two-dimensional coordinate system. The initial two-dimensional coordinates of the target marker points in the initial image are then extracted.

[0010] A three-dimensional coordinate system for the C-arm is constructed, and the initial three-dimensional coordinates of the target marker points in the three-dimensional coordinate system are extracted. Based on the initial two-dimensional coordinates and the corresponding initial three-dimensional coordinates of each target marker point, the projection parameters of the C-arm under the initial pose are calculated.

[0011] The C-arm is driven to rotate around at least one of the X-axis, Y-axis and Z-axis of its three-dimensional coordinate system. When it rotates to a calibration pose, the rotation direction and rotation angle are recorded, and the calibration phantom is exposed to obtain a calibration two-dimensional image under the corresponding pose. The calibration two-dimensional coordinates of the target marker points are extracted from the image.

[0012] Based on the recorded rotation direction and rotation angle, combined with the initial three-dimensional coordinates, the calibration three-dimensional coordinates corresponding to the target marker point in the calibration pose are determined. Based on the correspondence between the extracted calibration two-dimensional coordinates and the calibration three-dimensional coordinates, the projection parameters of the C-arm in each calibration pose are calculated.

[0013] Based on the projection parameters calculated under the initial pose and each calibration pose, and combined with the rotation direction and rotation angle corresponding to each calibration pose, a mapping relationship between the C-arm pose and the projection parameters is established.

[0014] Furthermore, the calibration mold with pre-set regularly arranged marker points is specifically composed of several marker points arranged around the circumference of the cylindrical surface. The marker points are composed of different sizes to distinguish their positions. The marker points are made of steel balls.

[0015] Furthermore, the calibration phantom placed within the imaging area of ​​the C-arm must satisfy the following condition: the intersection of the multiple sets of positioning laser beams set on the C-arm coincides with the geometric center of the calibration phantom;

[0016] The multiple sets of positioning laser beams include three sets of line lasers, which are used to indicate the left and right positions of the calibration phantom on the flat panel detector side, the up and down positions on the C-arm slip ring track, and the left and right positions on the X-ray source tube side.

[0017] Furthermore, the method for obtaining the two-dimensional coordinates of the target marker point is as follows:

[0018] First, obtain the row and column coordinates of the target marker points in the initial image. Then, convert the row and column coordinates into two-dimensional coordinates in the image coordinate system according to a preset transformation relationship. The expression is:

[0019] ;

[0020] in: Let be the x-coordinate of the i-th target marker point in the image coordinate system. Let be the y-coordinate value of the i-th target marker point in the image coordinate system.

[0021] Let be the coordinates of the i-th target marker point in the u-direction within the row and column coordinate system. Let be the coordinate value of the i-th target marker point in the v-direction in the row and column coordinate system.

[0022] , These represent the offsets of the image row and column coordinate systems relative to the image coordinate system in the x and y directions, respectively.

[0023] The method for constructing the image coordinate system is as follows: with the center of the exposed image as the origin, the positive x-axis is to the right along the image column direction, and the positive y-axis is to the upward direction along the image row direction, thus establishing a right-handed coordinate system;

[0024] The method for constructing the row and column coordinate system is as follows: with the upper left corner of the exposed image as the origin, the positive direction of the column coordinate is to the right along the column direction of the image, and the positive direction of the row coordinate is to the downward direction along the row direction of the image.

[0025] Furthermore, the method for obtaining the projection parameters includes:

[0026] Methods for obtaining C-arm projection parameters under isocentric and non-isocentric conditions.

[0027] Furthermore, the specific method for obtaining the C-arm projection parameters under the condition of isocenter is as follows:

[0028] Step 1: Obtain the projection parameters of the C-arm in its initial pose under isocentric conditions. The expression is:

[0029] ;

[0030] Where: R is the distance from the X-ray source to the center of the C-arm, and D is the distance from the center of the C-arm to the flat panel detector. for Magnification of the image projected onto the flat panel detector. Let x be the x-coordinate of the i-th target marker point in the image coordinate system captured by the flat panel detector, representing the initial pose. Let y be the y-coordinate of the i-th target marker point in the image coordinate system captured by the flat panel detector, representing the initial pose. Let x be the x-coordinate of the initial pose marker point in the C-arm's three-dimensional coordinate system. Let y be the initial pose marker point in the C-arm's three-dimensional coordinate system. Let z be the z-coordinate of the initial pose marker point in the C-arm's three-dimensional coordinate system;

[0031] Step 2: Drive the C-arm to rotate around the X-axis, Y-axis, and Z-axis respectively. , , When the angle reaches different calibration poses, the rotation transformation matrix around each coordinate axis is calculated based on the corresponding rotation angle and the projection parameters under the initial pose, and then the complete projection matrix under each calibration pose is obtained.

[0032] Furthermore, the specific method for obtaining the C-arm projection parameters under non-isocentric conditions is as follows:

[0033] Step 1: Within the initial pose and small-angle rotation range, approximate the C-arm as an isocentric structure. Using the method for obtaining the projection parameters of the C-arm under isocentric conditions, solve for the initial projection matrix and the rotation transformation matrix around each coordinate axis of the C-arm under non-isocentric conditions. The expressions are as follows:

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] ;

[0039] Where: θx, θy, and θz are the rotation angles of the C-arm about its three-dimensional coordinate system's x-axis, y-axis, and z-axis, respectively; θ is the sum of θx, θy, and θz; P θ R is the initial projection matrix of the non-equicenter C-arm; θ D is the distance from the X-ray source to the center of the C-arm after the non-isocentric C-arm is rotated by an angle θ. θT is the distance from the center of the non-isocentric C-arm to the flat panel detector after rotating the C-arm by an angle θ. x T is the rotation matrix of the non-isocentric C-arm about the x-axis. y T is the rotation matrix of the non-isocentric C-arm about the y-axis. z The rotation matrix of the non-isocentric C-arm about the z-axis. Indicates the positive and negative directions of the image coordinates rotation around the x-axis. Indicates the positive and negative directions of the image coordinates rotation around the y-axis. Indicates the positive and negative directions of the image coordinate rotation around the z-axis; m iθ After rotating the non-equicenter C-arm by an angle θ Magnification of the image projected onto the flat panel detector. These are the x and y coordinates of the i-th marker point in the image coordinate system of the flat panel detector after the non-isocentric C-arm is rotated by an angle θ; , and These represent the x, y, and z coordinates of the marked point in the three-dimensional coordinate system of the C-arm after the non-isocentric C-arm is rotated by an angle θ.

[0040] Step 2: Divide the entire range of motion of the C-arm into several angular intervals, select one or more representative poses in each angular interval, and solve for the projection parameters under each representative pose;

[0041] Step 3: Based on the series of discrete poses and their corresponding projection parameters obtained in steps 1 and 2, construct a pose-projection parameter mapping relationship covering the entire usage range of the C-arm.

[0042] Secondly, the present invention provides a C-arm calibration device based on image feature extraction, comprising:

[0043] The image acquisition module is used to place a calibration phantom with pre-set regularly arranged marker points into the imaging area of ​​the C-arm, expose the C-arm in its initial pose, acquire the initial image and construct a two-dimensional coordinate system, and extract the initial two-dimensional coordinates of the target marker points in the initial image.

[0044] The initial projection parameter module is used to construct a three-dimensional coordinate system for the C-arm, extract the initial three-dimensional coordinates of the target marker points in the three-dimensional coordinate system, and calculate the projection parameters of the C-arm under the initial pose based on the initial two-dimensional coordinates and the corresponding initial three-dimensional coordinates of each target marker point.

[0045] The calibration module is used to drive the C-arm to rotate around at least one of the X-axis, Y-axis and Z-axis of its three-dimensional coordinate system. When rotating to a calibration pose, the rotation direction and rotation angle are recorded, the calibration phantom is exposed, the calibration two-dimensional image under the corresponding pose is obtained, and the calibration two-dimensional coordinates of the target marker points are extracted.

[0046] The calibration projection parameter module is used to determine the calibration three-dimensional coordinates of the target marker point in the calibration pose based on the recorded rotation direction and rotation angle, combined with the initial three-dimensional coordinates, and to calculate the projection parameters of the C-arm in each calibration pose based on the correspondence between the extracted calibration two-dimensional coordinates and the calibration three-dimensional coordinates.

[0047] The mapping relationship establishment module is used to establish a mapping relationship between the C-arm pose and the projection parameters based on the projection parameters calculated under the initial pose and each calibration pose, combined with the rotation direction and rotation angle corresponding to each calibration pose.

[0048] Thirdly, the present invention provides a C-arm calibration system based on image feature extraction, comprising:

[0049] Memory, used to store computer programs / instructions;

[0050] A processor is used to execute the computer program / instructions to implement the steps of the above-described C-arm calibration method based on image feature extraction.

[0051] Fourthly, the present invention provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of the above-described C-arm calibration method based on image feature extraction.

[0052] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention designs a special calibration phantom with regularly arranged and differentiated size markers, and combines high-precision laser positioning to achieve precise centering of the phantom, which significantly improves the recognition accuracy and anti-interference ability of markers in the image, and provides high-quality two-dimensional coordinate input for subsequent projection parameter calculation; on this basis, the present invention, through a systematic process of image coordinate extraction, coordinate system transformation and projection parameter calculation, can be applied to both isocentric and non-isocentric C-arm structures.

[0053] Through the above-mentioned calibration method based on image feature processing, the C-arm can be quickly positioned to the required surgical pose in one go. This not only significantly improves the accuracy and efficiency of pose adjustment, but also greatly reduces the radiation dose caused by repeated exposure and shortens the operation time, thus having good clinical application value. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the isocentric C-arm calibration process;

[0055] Figure 2 This is a schematic diagram of the calibration process for a non-isocentric C-arm.

[0056] Figure 3 A design diagram for the calibration phantom is shown below;

[0057] Figure 4 This is a schematic diagram of the image coordinate system and the three-dimensional coordinate system of the C-arm. Detailed Implementation

[0058] It should be noted that:

[0059] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0060] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0061] Example 1

[0062] like Figure 1 The embodiment shown provides a C-arm calibration method based on image feature extraction. This embodiment details the calibration process applied to isocentric C-arms, specifically including:

[0063] Step S1: System preparation and model placement

[0064] Place the calibration phantom with pre-arranged markers on the operating table or a special support, turn on the three sets of positioning lasers on the C-arm, and adjust the spatial position of the calibration phantom until the intersection of the three laser lines coincides with the geometric center of the calibration phantom. At this time, the central axis of the calibration phantom is perpendicular to the X-axis of the three-dimensional coordinate system of the C-arm (pointing from the rotation center to the flat plate detector) and basically aligned with the Z-axis of the three-dimensional coordinate system of the C-arm.

[0065] The three-dimensional coordinate system of the C-arm is defined as follows: with the geometric center (i.e., the isocenter point) of the C-arm as the origin, and the direction from the X-ray source to the flat panel detector as... Positive direction of the axis, perpendicular to The axis and the direction pointing towards the opening of the C-arm are The positive direction of the axis is determined by the right-hand rule. Positive direction of the axis.

[0066] like Figure 3 As shown, the calibration mold with pre-set regular arrangement of marker points is specifically composed of several marker points, which are arranged in one or two circles along the circumference of the cylindrical surface. The marker points are composed of different sizes to distinguish their positions, and the marker points are made of steel balls.

[0067] This embodiment improves the accuracy and precision of steel ball extraction by designing special arrangements of steel balls of different sizes. Through these special arrangements, when extracting marker points, it is easier to identify which steel balls were effectively extracted and which were not, based on their combination and relative position. Simultaneously, it allows for correction of marker point coordinates with excessive extraction errors. The steel ball design scheme is shown in Table 1 below.

[0068] Table 1: Design dimensions and numbering of steel balls

[0069]

[0070] As shown in the table above, steel balls with diameters of 2mm and 4mm were selected as markers, arranged in the order shown in the table, totaling 21 markers. The distance between adjacent markers remained constant, and the relative positions of the five large steel balls (4mm) were fixed. The number and relative positions of the small steel balls (2mm) between adjacent large steel balls were also fixed. Therefore, when extracting the center coordinates of the steel balls, if some large steel balls were not effectively extracted, it was easy to identify which positions or numbers of the large steel balls were not extracted. Similarly, if some small steel balls were not effectively extracted, their positions or numbers could be identified based on their relative positions to the large steel balls. The relative positional relationship of the steel balls identifies the location and number of the small steel balls that were not extracted. By identifying and eliminating markers that were not effectively identified, the center coordinates of the effectively identified markers on the image are accurately matched one-to-one with the three-dimensional markers on the calibration phantom, which can effectively improve the accuracy and speed of calibration. Compared with using markers of the same size, if a marker is not identified or is misidentified, the coordinates of the marker on the image cannot be matched one-to-one with the three-dimensional markers on the calibration phantom, thus making the calibration image invalid and requiring re-image until all markers are effectively identified.

[0071] If there are misidentified points, resulting in the mismatch between the marker points on the image and the three-dimensional marker points on the calibration phantom, it will lead to excessive calibration accuracy error. At the same time, after the coordinates of the large steel ball in the image are effectively extracted, the small steel ball can be locally optimized according to its positional relationship with the large steel ball, thereby improving the extraction accuracy of the center coordinates of the small steel ball.

[0072] Step S2: Initial pose (orthogonal) image acquisition and coordinate extraction

[0073] Adjust the C-arm to its initial pose, i.e., when both the sliding angle and rotation angle are zero (positive position). Expose the calibration phantom in this pose to obtain the initial image Io;

[0074] Establish an image coordinate system: with the center of the initial image Io as the origin, the positive x-axis is to the right along the column direction of the image, the positive y-axis is upward along the row direction of the image, and the positive z-axis is upward perpendicular to the image plane.

[0075] Establish a row and column coordinate system: with the top left corner of the initial image Io as the origin, the positive direction of the column coordinate is to the right along the column direction of the image, and the positive direction of the row coordinate is downward along the row direction of the image.

[0076] Identify and extract the coordinates of the centers of all visible steel balls in the initial image Io in the row and column coordinate system. , i=1, 2, ..., N, where N is the number of steel balls;

[0077] like Figure 4 As shown, the row and column coordinates are transformed using a preset transformation relationship. Convert to image physical coordinates The transformation relationship can be expressed as:

[0078] ;

[0079] in: Let be the x-coordinate of the i-th target marker point in the image coordinate system. Let be the y-coordinate value of the i-th target marker point in the image coordinate system.

[0080] Let be the coordinates of the i-th target marker point in the u-direction within the row and column coordinate system. Let be the coordinate value of the i-th target marker point in the v-direction in the row and column coordinate system.

[0081] , These are the offsets of the image row and column coordinate systems relative to the image coordinate system in the x and y directions, respectively. These parameters can be obtained in advance through detector calibration.

[0082] Step S3: Solving for initial projection parameters

[0083] Based on the known geometric dimensions of the calibration phantom and its placement orientation in the C-arm's three-dimensional coordinate system (ensuring alignment via laser), the initial three-dimensional coordinates of each steel ball in the C-arm's three-dimensional coordinate system can be calculated. For an isocentric C-arm, let the coordinates of the X-ray source focus S in the three-dimensional coordinate system be (-DR, 0, 0), where R is the distance from the X-ray source to the center of the C-arm, D is the distance from the center of the C-arm to the flat panel detector, and the flat panel detector plane is located at x = R.

[0084] For the i-th steel ball, its world coordinates and image coordinates have the following relationship:

[0085] ;

[0086] in, for Magnification of the image projected onto the flat panel detector;

[0087] The above relationships can be expressed as a matrix expression as follows:

[0088] ;

[0089] ;

[0090] because For unknown parameters D and R, a system of equations can be constructed. Using the coordinate relationships of multiple (N≥2, usually more for greater accuracy) steel balls, an optimization algorithm is employed to solve for the optimal values ​​of D and R, thus obtaining the projection parameters under the initial pose. Simultaneously, the initial matrix... Equivalent to parameters D and R, i.e., from the initial matrix D and R can be solved, and the matrix can also be solved given D and R. ;

[0091] Step S4: Single-axis rotation calibration and orientation determination

[0092] Keeping the calibration phantom's position absolutely unchanged, rotate it around the X-axis for calibration: drive the C-arm around its three-dimensional coordinate system. Rotate the axis by a known angle Record the rotation angle and observed rotation direction. Under this new pose, expose the calibration phantom to obtain the first calibration image Ix, and extract the calibration two-dimensional coordinates of the same steel ball. .

[0093] Since it is an isocentric C-arm, the parameters R and D remain unchanged after rotation. The three-dimensional coordinates of the steel ball in the new pose can be obtained by multiplying the initial three-dimensional coordinates by the rotation matrix around the X-axis. We obtain the rotation matrix. It's an angle. The function, with the expression:

[0094] ;

[0095] in: Indicates the positive and negative directions of the image coordinates rotation around the x-axis. The angle of rotation about the x-axis is the magnitude of the rotation. The readings can be directly read from the scale on the C-arm machine;

[0096] Substitute the rotated 3D coordinates into the projection equation, and use the known D, R, and The direction of rotation can be deduced (i.e.) (positive and negative signs).

[0097] Rotation calibration around the Y and Z axes: Repeat a similar process around the X axis; drive the C-arm individually around the X and Z axes. Axis rotation | | Angle and circumference Axis rotation | The angle is used to perform exposure, coordinate extraction, and direction determination to obtain the rotation matrices around the Y and Z axes. and and its direction parameters , ;

[0098] Step S5: Establishing and Verifying Mapping Relationships

[0099] After the above steps, we have obtained:

[0100] Initial projection matrix (Including parameters D, R).

[0101] Rotation transformation matrices about the X, Y, and Z axes , and ,angle With symbols;

[0102] For any C-arm pose, its rotation angles about the X, Y, and Z axes The only certainty is that the complete projection matrix M in this pose can be calculated as:

[0103] ;

[0104] This formula establishes the formula from the "isocentric C-arm pose". The precise mapping relationship from “projection matrix M” to “projection matrix M”.

[0105] Calibration accuracy verification:

[0106] Rotate the C-arm's pose around the X, Y, and Z axes by a certain angle, and record the magnitude and direction of the rotation angle around the X, Y, and Z axes respectively. Substituting these values ​​into the rotation transformation matrix yields the rotation transformation matrix. , and Simultaneously, based on the projection matrix obtained from calibration Obtain the projection matrix M under the current pose; based on the known 3D coordinates of the N steel balls on the calibration phantom and the projection matrix M, the 2D image coordinates of the N steel balls under the current pose on the projected image can be obtained (the 2D image coordinates of the N steel balls are calculated based on the projection matrix); expose the calibration phantom under the current pose to obtain the exposed image, and extract the image coordinates of the N steel balls under the current pose (the 2D image coordinates of the N steel balls are extracted based on the actual exposed image); extract the 2D image coordinates of the target points in the exposed image. Error analysis is performed on the obtained image coordinates. If the error meets the threshold δ, the calibration accuracy meets the requirements.

[0107] The error analysis expression is as follows:

[0108] ;

[0109] in: To calibrate the number of steel balls on the phantom, and The coordinates of the two-dimensional image are obtained by solving the projection matrix M.

[0110] For isocentric C-arms, this embodiment can establish an accurate analytical mapping relationship through initial pose and single-axis rotation calibration.

[0111] Example 2:

[0112] Non-isocentric C-arm calibration completed

[0113] like Figure 2 As shown, this embodiment provides a non-isocentric C-arm calibration method. The core of the non-isocentric C-arm calibration is: approximating isocentric processing within a small angle range, specifically including:

[0114] Step T1: Orthogonal small-angle range reference calibration

[0115] This step is exactly the same as steps S1~S4 in Example 1. The C-arm is approximated as an isocentric structure. Using the method for obtaining the projection parameters of the C-arm under isocentric conditions, the initial projection matrix of the C-arm and the rotation transformation matrix around each coordinate axis under non-isocentric conditions are obtained. The expressions are:

[0116] ;

[0117] ;

[0118] ;

[0119] ;

[0120] ;

[0121] Where: θx, θy, and θz are the rotation angles of the C-arm around the x-axis, y-axis, and z-axis of the three-dimensional coordinate system of the C-arm, respectively; θ is the sum of θx, θy, and θz; P θ R is the initial projection matrix of the non-equicenter C-arm; θ D is the distance from the X-ray source to the center of the C-arm after the non-isocentric C-arm is rotated by an angle θ. θ T is the distance from the center of the non-isocentric C-arm to the flat panel detector after rotating the C-arm by an angle θ. x T is the rotation matrix of the non-isocentric C-arm about the x-axis. y T is the rotation matrix of the non-isocentric C-arm about the y-axis. z The rotation matrix of the non-isocentric C-arm about the z-axis. Indicates the positive and negative directions of the image coordinates rotation around the x-axis. Indicates the positive and negative directions of the image coordinates rotation around the y-axis. Indicates the positive and negative directions of the image coordinate rotation around the z-axis; m iθ After rotating the non-equicenter C-arm by an angle θ Magnification of the image projected onto the flat panel detector. These are the x and y coordinates of the i-th marker point in the image coordinate system of the flat panel detector after the non-isocentric C-arm is rotated by an angle θ; , and These represent the x, y, and z coordinates of the marked point in the three-dimensional coordinate system of the C-arm after rotating the non-isocentric C-arm by an angle θ.

[0122] Step T2: Calibration of small-angle references at other locations

[0123] This step is exactly the same as steps S1 to S3 in Example 1. The parameters solved in step S4 for each position are the same, so only one solution for the position parameters in S4 is needed in the positive position. The C-arm is approximated as an isocentric structure within a small angle range. The projection parameter acquisition method under the isocentric condition is used to solve for the initial projection matrix and the rotation transformation matrix around each coordinate axis.

[0124] Establishment and verification of mapping relationships.

[0125] The reference calibration within a small angular range at the correct position has yielded the following results:

[0126] Projection matrix at each position (Includes parameter D) θ , R θ ).

[0127] Rotation transformation matrices about the X, Y, and Z axes , and ,angle With symbols;

[0128] For any C-arm pose, its rotation angles about the X, Y, and Z axes The only certainty is that the complete projection matrix M in this pose can be calculated as:

[0129] ;

[0130] This formula establishes the position of the "non-isocentric C-arm". The precise mapping relationship from “projection matrix M” to “projection matrix M”.

[0131] Additionally: Keeping the positional relationship between the calibration phantom and the C-arm unchanged, slide the C-arm to 5-10 specific angles within the calibration grid coverage area, record the angle values, and expose to acquire phantom images;

[0132] The verification scheme for the non-isocentric C-arm is the same as that for the isocentric scheme, except that it needs to be based on the projection parameters D of different intervals. θ , R θ Simply perform the calculation;

[0133] If the calibration accuracy meets the requirements, continue with the subsequent surgical procedures; if the accuracy requirements are not met, repeat the calibration operation to optimize the accuracy.

[0134] For non-equicenter C-arms, this embodiment innovatively adopts a "small angle approximate equicenter" strategy to effectively fit its nonlinear motion trajectory, solving the problem of the lack of a universal calibration method in the prior art.

[0135] Example 3:

[0136] This embodiment provides a C-arm calibration device based on image feature extraction, including:

[0137] The image acquisition module is used to place a calibration phantom with pre-set regularly arranged marker points into the imaging area of ​​the C-arm, expose the C-arm in its initial pose, acquire the initial image and construct a two-dimensional coordinate system, and extract the initial two-dimensional coordinates of the target marker points in the initial image.

[0138] The initial projection parameter module is used to construct a three-dimensional coordinate system for the C-arm, extract the initial three-dimensional coordinates of the target marker points in the three-dimensional coordinate system, and calculate the projection parameters of the C-arm under the initial pose based on the initial two-dimensional coordinates and the corresponding initial three-dimensional coordinates of each target marker point.

[0139] The calibration module is used to drive the C-arm to rotate around at least one of the X-axis, Y-axis and Z-axis of its three-dimensional coordinate system. When rotating to a calibration pose, the rotation direction and rotation angle are recorded, the calibration phantom is exposed, the calibration two-dimensional image under the corresponding pose is obtained, and the calibration two-dimensional coordinates of the target marker points are extracted.

[0140] The calibration projection parameter module is used to determine the calibration three-dimensional coordinates of the target marker point in the calibration pose based on the recorded rotation direction and rotation angle, combined with the initial three-dimensional coordinates, and to calculate the projection parameters of the C-arm in each calibration pose based on the correspondence between the extracted calibration two-dimensional coordinates and the calibration three-dimensional coordinates.

[0141] The mapping relationship establishment module is used to establish a mapping relationship between the C-arm pose and the projection parameters based on the projection parameters calculated under the initial pose and each calibration pose, combined with the rotation direction and rotation angle corresponding to each calibration pose.

[0142] Example 4

[0143] This embodiment provides a C-arm calibration system based on image feature extraction, including:

[0144] Memory, used to store computer programs / instructions;

[0145] A processor is used to execute the computer program / instructions to implement the steps of the above-described C-arm calibration method based on image feature extraction.

[0146] Example 5

[0147] This embodiment provides a computer-readable storage medium storing a computer program / instructions thereon. When the computer program / instructions are executed by a processor, they implement the steps of the above-described C-arm calibration method based on image feature extraction.

[0148] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0149] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0150] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0151] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0152] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A C-arm calibration method based on image feature extraction, characterized in that, Includes the following steps: A calibration phantom with pre-arranged marker points is placed in the imaging area of ​​a C-arm. Exposure is performed with the C-arm in its initial pose to acquire an initial image and construct a two-dimensional coordinate system. The initial two-dimensional coordinates of the target marker points in the initial image are then extracted. A three-dimensional coordinate system for the C-arm is constructed, and the initial three-dimensional coordinates of the target marker points in the three-dimensional coordinate system are extracted. Based on the initial two-dimensional coordinates and the corresponding initial three-dimensional coordinates of each target marker point, the projection parameters of the C-arm under the initial pose are calculated. The C-arm is driven to rotate around at least one of the X-axis, Y-axis and Z-axis of its three-dimensional coordinate system. When it rotates to a calibration pose, the rotation direction and rotation angle are recorded, and the calibration phantom is exposed to obtain a calibration two-dimensional image under the corresponding pose. The calibration two-dimensional coordinates of the target marker points are extracted from the image. Based on the recorded rotation direction and rotation angle, combined with the initial three-dimensional coordinates, the calibration three-dimensional coordinates corresponding to the target marker point in the calibration pose are determined. Based on the correspondence between the extracted calibration two-dimensional coordinates and the calibration three-dimensional coordinates, the projection parameters of the C-arm in each calibration pose are calculated. Based on the projection parameters calculated under the initial pose and each calibration pose, and combined with the rotation direction and rotation angle corresponding to each calibration pose, a mapping relationship between the C-arm pose and the projection parameters is established.

2. The C-arm calibration method based on image feature extraction according to claim 1, characterized in that, The calibration phantom with pre-set, regularly arranged marker points is specifically composed of several marker points arranged along the circumference of the cylindrical surface. The marker points are composed of different sizes to distinguish their orientation.

3. The C-arm calibration method based on image feature extraction according to claim 1, characterized in that, The calibration phantom placed within the imaging area of ​​the C-arm must satisfy the following condition: the intersection of the multiple sets of positioning laser beams set on the C-arm coincides with the geometric center of the calibration phantom. The multiple sets of positioning laser beams include three sets of line lasers, which are used to indicate the left and right positions of the calibration phantom on the flat panel detector side, the up and down positions on the C-arm slip ring track, and the left and right positions on the X-ray source tube side.

4. The C-arm calibration method based on image feature extraction according to claim 1, characterized in that, The specific method for obtaining the two-dimensional coordinates of the target marker point is as follows: First, obtain the row and column coordinates of the target marker points in the initial image. Then, convert the row and column coordinates into two-dimensional coordinates in the image coordinate system according to a preset transformation relationship. The expression is: ； in: Let be the x-coordinate of the i-th target marker point in the image coordinate system. Let be the y-coordinate value of the i-th target marker point in the image coordinate system. Let be the coordinates of the i-th target marker point in the u-direction within the row and column coordinate system. Let be the coordinate value of the i-th target marker point in the v-direction in the row and column coordinate system. , These represent the offsets of the image row and column coordinate systems relative to the image coordinate system in the x and y directions, respectively. The method for constructing the image coordinate system is as follows: with the center of the exposed image as the origin, the positive x-axis is to the right along the image column direction, and the positive y-axis is to the upward direction along the image row direction, thus establishing a right-handed coordinate system; The method for constructing the row and column coordinate system is as follows: with the upper left corner of the exposed image as the origin, the positive direction of the column coordinate is to the right along the column direction of the image, and the positive direction of the row coordinate is to the downward direction along the row direction of the image.

5. The C-arm calibration method based on image feature extraction according to claim 4, characterized in that, The method for obtaining the projection parameters includes: Methods for obtaining C-arm projection parameters under isocentric and non-isocentric conditions.

6. The C-arm calibration method based on image feature extraction according to claim 5, characterized in that, The specific method for obtaining the projection parameters of the C-arm under the condition of isocenter is as follows: Step 1: Obtain the projection parameters of the C-arm in its initial pose under isocentric conditions. The expression is: ； Where: R is the distance from the X-ray source to the center of the C-arm, and D is the distance from the center of the C-arm to the flat panel detector. for Magnification of the image projected onto the flat panel detector. Let x be the x-coordinate of the i-th target marker point in the image coordinate system captured by the flat panel detector, representing the initial pose. Let y be the y-coordinate of the i-th target marker point in the image coordinate system captured by the flat panel detector, representing the initial pose. Let x be the x-coordinate of the initial pose marker point in the C-arm's three-dimensional coordinate system. Let y be the initial pose marker point in the C-arm's three-dimensional coordinate system. Let z be the z-coordinate of the initial pose marker point in the C-arm's three-dimensional coordinate system; Step 2: Drive the C-arm to rotate around the X-axis, Y-axis, and Z-axis respectively. , , When the angle reaches different calibration poses, the rotation transformation matrix around each coordinate axis is calculated based on the corresponding rotation angle and the projection parameters under the initial pose, and then the complete projection matrix under each calibration pose is obtained.

7. The C-arm calibration method based on image feature extraction according to claim 5, characterized in that, The specific method for obtaining the projection parameters of the C-arm under non-isocentric conditions is as follows: Step 1: Within the initial pose and small-angle rotation range, approximate the C-arm as an isocentric structure. Using the method for obtaining C-arm projection parameters under isocentric conditions, solve for the initial projection matrix and the rotation transformation matrix around each coordinate axis. The expressions are: ； ； ； ； ； Where: θx, θy, and θz are the rotation angles of the C-arm about its three-dimensional coordinate system's x-axis, y-axis, and z-axis, respectively, θ is the sum of θx, θy, and θz, and P θ R is the initial projection matrix of the non-equicenter C-arm; θ D is the distance from the X-ray source to the center of the C-arm after the non-isocentric C-arm is rotated by an angle θ. θ T is the distance from the center of the non-isocentric C-arm to the flat panel detector after rotating the C-arm by an angle θ. x T is the rotation matrix of the non-isocentric C-arm about the x-axis. y T is the rotation matrix of the non-isocentric C-arm about the y-axis. z The rotation matrix of the non-isocentric C-arm about the z-axis. Indicates the positive and negative directions of the image coordinates rotation around the x-axis. Indicates the positive and negative directions of the image coordinates rotation around the y-axis. Indicates the positive and negative directions of the image coordinate rotation around the z-axis, m iθ After rotating the non-equicenter C-arm by an angle θ Magnification of the image projected onto the flat panel detector. These are the x and y coordinates of the i-th marker point in the image coordinate system of the flat panel detector after the non-isocentric C-arm is rotated by an angle θ; , and These represent the x, y, and z coordinates of the marked point in the three-dimensional coordinate system of the C-arm after the non-isocentric C-arm is rotated by an angle θ. Step 2: Divide the entire range of motion of the C-arm into several angular intervals, select several representative poses in each angular interval, and solve for the projection parameters under each representative pose; Step 3: Based on the series of discrete poses and their corresponding projection parameters obtained in steps 1 and 2, construct a pose-projection parameter mapping relationship covering the entire usage range of the C-arm.

8. A C-arm calibration device based on image feature extraction, characterized in that, include: The image acquisition module is used to place a calibration phantom with pre-set regularly arranged marker points into the imaging area of ​​the C-arm, expose the C-arm in its initial pose, acquire the initial image and construct a two-dimensional coordinate system, and extract the initial two-dimensional coordinates of the target marker points in the initial image. The initial projection parameter module is used to construct a three-dimensional coordinate system for the C-arm, extract the initial three-dimensional coordinates of the target marker points in the three-dimensional coordinate system, and calculate the projection parameters of the C-arm under the initial pose based on the initial two-dimensional coordinates and the corresponding initial three-dimensional coordinates of each target marker point. The calibration module is used to drive the C-arm to rotate around at least one of the X-axis, Y-axis and Z-axis of its three-dimensional coordinate system. When rotating to a calibration pose, the rotation direction and rotation angle are recorded, the calibration phantom is exposed, the calibration two-dimensional image under the corresponding pose is obtained, and the calibration two-dimensional coordinates of the target marker points are extracted. The calibration projection parameter module is used to determine the calibration three-dimensional coordinates of the target marker point in the calibration pose based on the recorded rotation direction and rotation angle, combined with the initial three-dimensional coordinates, and to calculate the projection parameters of the C-arm in each calibration pose based on the correspondence between the extracted calibration two-dimensional coordinates and the calibration three-dimensional coordinates. The mapping relationship establishment module is used to establish a mapping relationship between the C-arm pose and the projection parameters based on the projection parameters calculated under the initial pose and each calibration pose, combined with the rotation direction and rotation angle corresponding to each calibration pose.

9. A C-arm calibration system based on image feature extraction, characterized in that, include: Memory, used to store computer programs / instructions; A processor for executing the computer program / instructions to implement the steps of the C-arm calibration method based on image feature extraction as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the C-arm calibration method based on image feature extraction as described in any one of claims 1-7.

Images