A CT image reconstruction precision analysis device and method

Abstract

The application discloses a CT image reconstruction precision analysis device and method, wherein the CT image reconstruction precision analysis device comprises: splicing pieces, which are detachably connected between adjacent splicing pieces to form an expandable detection tool; and calibration pieces, which are distributed on exposed surfaces of the splicing pieces; the exposed surfaces of the detection tool comprise the same calibration pieces and newly-added calibration pieces before and after each expansion; a tool coordinate system is constructed through the expandable detection tool, and is registered with an image coordinate system in a CT image to obtain error distribution between image coordinates and tool coordinates of the calibration pieces. The application simulates a human body scanning part in the CT image by using the expandable detection tool, and installs the calibration pieces on the detection tool to simulate a human body lesion, so that actual coordinates of the calibration pieces on the exposed surfaces of the detection tool are gradually measured, and the obtained actual coordinates are registered with the CT image coordinates, thereby effectively improving the reliability of CT image precision analysis.

Description

Technical Field

[0001] This invention relates to the field of auxiliary medical device technology, and in particular to a CT image reconstruction accuracy analysis device and method. Background Technology

[0002] In recent years, surgical robots have developed rapidly and have been widely studied worldwide. By fully utilizing medical imaging information, surgical robots can provide doctors with more auxiliary functions, such as navigation and positioning, and are characterized by high operational precision, good repeatability, and high stability. To fully leverage the above-mentioned performance of surgical robots, it is necessary to ensure the accuracy of CT image reconstruction, thereby guaranteeing the accurate location of identified target lesions. Therefore, the assessment of CT image accuracy is crucial.

[0003] To analyze the accuracy of CT image reconstruction, a detection fixture capable of simulating scanned areas of the human body is needed. This fixture must also contain a calibration component that can simulate the location of lesions. Only by comparing the actual position of the calibration component on the detection fixture with the calibration position determined through CT image reconstruction can the error between the image calibration position and the actual calibration position be obtained.

[0004] Due to manufacturing precision limitations, the actual positions of calibration components will vary for each testing fixture. Therefore, precise measurements are required to determine the actual positions of the calibration components on each fixture. If the fixture is designed as a single piece with calibration components embedded internally, the actual positions of all calibration components can only be measured externally using a measuring instrument. This method yields inaccurate measurements with significant distance errors, which in turn makes it difficult to guarantee the reliability of CT image accuracy analysis. Summary of the Invention

[0005] Purpose of the invention: To address the problems existing in the prior art, the present invention provides a CT image reconstruction accuracy analysis device and method, which can accurately measure the actual position of the calibration parts on the testing fixture, thereby improving the reliability of CT image accuracy analysis.

[0006] Technical solution: To achieve the above objectives, the present invention provides a CT image reconstruction accuracy analysis device, comprising:

[0007] The splicing components are detachably connected to each other to form an expandable testing fixture;

[0008] Calibration components are distributed on the exposed surface of the splicing components;

[0009] Before and after each expansion, a temporary coordinate system is constructed based on the calibration components on the exposed surface of the testing fixture;

[0010] The calibration components on the exposed surface of the testing fixture include the same calibration components before and after expansion, as well as newly added calibration components. The same calibration components are used to register each temporary coordinate system to obtain the fixture coordinate system. The obtained fixture coordinate system is then registered with the image coordinate system in the CT image to obtain the error distribution between the calibration component image coordinates and the fixture coordinates.

[0011] Furthermore, the same calibration component includes a calibration component that is always exposed during the expansion process, and the registration of the temporary coordinate system is performed based on the always exposed calibration component.

[0012] Specifically, after each expansion, the exposed surface of the testing fixture only contains the always-exposed calibration component and the newly added calibration component.

[0013] Specifically, the always exposed calibration components are distributed on a splicing component that serves as a reference component, and the remaining splicing components serve as extension components. The reference component is used as the initial testing fixture, and the extension components are added one by one to form a progressively expanding testing fixture.

[0014] Specifically, the always-exposed calibration element is distributed on one reference surface of the reference element.

[0015] Furthermore, the registration of the obtained tooling coordinate system with the image coordinate system in the CT image specifically involves:

[0016] Determine the correspondence between the tooling coordinates and image coordinates of some or all calibration parts, and perform registration between the tooling coordinate system and the image coordinate system.

[0017] Specifically, the calibration components include a first calibration component and a second calibration component, and the tooling coordinate system and the image coordinate system are initially registered based on the tooling coordinates and image coordinates of the first calibration component.

[0018] Specifically, the correspondence between the tooling coordinates and image coordinates of the second calibration part is determined based on the results of the preliminary registration, and a second registration between the tooling coordinate system and the image coordinate system is performed based on the tooling coordinates and image coordinates of all calibration parts.

[0019] Furthermore, it also includes a fastener, which is detachably connected to the splicing component, for protecting the exposed calibration components on the testing fixture after the expansion of the testing fixture is completed.

[0020] A method for analyzing the accuracy of CT image reconstruction, based on the aforementioned CT image reconstruction accuracy analysis device, includes the following steps:

[0021] S1. Before and after each expansion, obtain the temporary coordinates of the calibration part on the exposed surface of the testing fixture, and register the temporary coordinates according to the same calibration part to obtain the fixture coordinates of the calibration part in the fixture coordinate system.

[0022] S2. After completing the expansion of the testing fixture, extract the image coordinates of the calibration parts on the testing fixture through CT image reconstruction.

[0023] S3. Register the tooling coordinates and image coordinates of the calibration part to the same coordinate system to obtain the error distribution between the image coordinates and tooling coordinates of the calibration part.

[0024] Furthermore, the registration of temporary coordinates based on the same calibration component specifically includes:

[0025] The same calibration components include those that are always exposed during the expansion process, and temporary coordinates are registered based on the always exposed calibration components.

[0026] Furthermore, after completing the expansion of the testing fixture, a fixing component is further assembled on the testing fixture. The fixing component is detachably connected to the splicing component and is used to protect the exposed calibration component on the testing fixture.

[0027] Furthermore, the process of registering the tooling coordinates and image coordinates of the calibration part to the same coordinate system specifically involves:

[0028] Determine the correspondence between the tooling coordinates and image coordinates of some or all calibration parts, and perform registration between the tooling coordinate system and the image coordinate system.

[0029] Specifically, the calibration components include a first calibration component and a second calibration component, and the tooling coordinate system and the image coordinate system are initially registered based on the tooling coordinates and image coordinates of the first calibration component.

[0030] Specifically, the correspondence between the tooling coordinates and image coordinates of the second calibration part is determined based on the results of the preliminary registration, and a second registration between the tooling coordinate system and the image coordinate system is performed based on the tooling coordinates and image coordinates of all calibration parts.

[0031] Beneficial effects:

[0032] This invention employs an expandable detection fixture to simulate human body scanning areas in CT images, and installs calibration components on the detection fixture to simulate human lesions. This allows for the gradual measurement and registration of temporary coordinates of the calibration components on the exposed surface of the fixture, ultimately obtaining the actual coordinates of all calibration components. Compared to directly measuring the actual coordinates of calibration components within a single fixture, this significantly improves measurement accuracy. By registering the obtained actual coordinates with CT image coordinates, the reliability of CT image reconstruction accuracy analysis is effectively improved, facilitating the full utilization of the superior performance of the surgical robot. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the CT image reconstruction accuracy analysis device in an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of the structure of the reference component in an embodiment of the present invention;

[0035] Figure 3 This is a schematic diagram of the structure of the extension component in an embodiment of the present invention;

[0036] Figure 4 a to 4c are schematic diagrams illustrating the process of calibrating the target ball using a coordinate measuring machine in an embodiment of the present invention;

[0037] Figure 5 This is an overall flowchart of the CT image reconstruction accuracy analysis method in an embodiment of the present invention;

[0038] Figure 6 This is a flowchart illustrating the extraction of target sphere image coordinates in an embodiment of the present invention;

[0039] The diagram includes: 1. reference component, 2. extension component, 3. fixing component, 4. target ball, 41. large ball, 42. small ball, 5. receiving groove, 6. receiving cavity, 7. hexagonal thread mounting groove, 8. hexagonal plastic screw, 9. center surface. Detailed Implementation

[0040] The preferred embodiments of the present invention will now be described in conjunction with the accompanying drawings to more clearly and completely illustrate the technical solution of the present invention.

[0041] Reference Figure 1 The CT image reconstruction accuracy analysis device provided by the present invention includes several splicing components and calibration components distributed on the exposed surface of the splicing components, wherein adjacent splicing components are detachably connected to form an expandable testing fixture.

[0042] In this invention, the detection fixture is used to simulate the scanning area of ​​the human body, and the calibration part is used to simulate the lesion of the human body. In order to distinguish between the two in CT images, they are made of different non-metallic materials (to prevent metal artifacts).

[0043] For example, the testing fixture can be made of plexiglass, which has advantages such as high mechanical strength, good light transmission, good insulation, and dimensional stability, while the calibration parts can be made of ceramic.

[0044] Furthermore, before and after each expansion, the exposed surface of the testing fixture has both the same calibration parts and newly added calibration parts. This allows for the creation of a temporary coordinate system by using the temporary coordinates of the calibration parts on the exposed surface of the testing fixture before and after each expansion. By registering the temporary coordinate system pairwise with the identical calibration parts before and after the expansion, the fixture coordinates of all calibration parts under the same coordinate system can be obtained; this coordinate system is called the fixture coordinate system.

[0045] Furthermore, the same calibration component includes a calibration component that is always exposed during the expansion process, so that a temporary coordinate system can be registered based on the always exposed calibration component, thereby greatly reducing the amount of computation.

[0046] Furthermore, after each expansion, the exposed surface of the testing fixture only contains the always-exposed calibration parts and the newly added calibration parts, in order to avoid repeated measurement and registration of the calibration parts.

[0047] Furthermore, the always exposed calibration components are distributed on a splicing component that serves as a reference component 1, and the remaining splicing components serve as extension components 2. Extension components 2 are added one by one on the basis of the reference component 1 to form a progressively expanding testing fixture.

[0048] Specifically, the testing fixture can be designed as a cuboid structure, and is composed of several cuboid splicing parts assembled sequentially along the same direction.

[0049] For example, the testing fixture is made up of one reference piece 1 and six extension pieces 2 assembled from left to right, wherein the one reference piece 1 is set as a cuboid A, and the six extension pieces 2 are set as cuboids B to G from left to right.

[0050] Furthermore, the always exposed calibration components are distributed on a reference surface of the reference component 1.

[0051] Reference Figures 2-3 Calibration elements are arranged on both the left and right sides of the reference element 1, while calibration elements are arranged only on the right side of the extension element 2. That is, the left side of the reference element 1 is the reference surface, and the calibration elements on the reference surface are always exposed, while the calibration elements on the right side of the extension element 2 are newly added calibration elements each time it is extended.

[0052] In this invention, after all expansion is completed, the testing fixture still has exposed calibration components, namely the calibration components on the reference surface and the calibration components added during the last expansion. These exposed calibration components are easily bumped or knocked during the storage and transfer of the testing fixture, thus affecting its service life and testing accuracy. Furthermore, since lesions requiring identification via CT images are mostly located inside the human body, a fixing component 3 needs to be added to the outside of the testing fixture. This fixing component is detachably connected to the splicing component to protect the exposed calibration components on the testing fixture.

[0053] For example, a fixing component 3 is added to the left and right sides of the testing fixture, namely the cuboid H on the left and the cuboid I on the right. No calibration component is set on the fixing component 3. It mainly serves to fix and protect, so that the testing fixture after assembly does not have any exposed calibration components.

[0054] Specifically, the testing fixture is used to simulate the chest area in a CT chest X-ray. Its width and height are designed according to chest circumference and need to conform to the width and thickness of the human body. The overall length of the testing fixture needs to be within the CT scanning area.

[0055] For example, the length of the detection fixture is designed to be 320 mm, and the height and width are designed to be 140 mm, ensuring that its length is a multiple of 20 mm for the minimum scanning area of ​​a 32-slice CT scanner with a slice thickness of 0.625 mm. This invention can also be used to simulate other parts of the human body or lateral radiographs, with the shape and size of the detection fixture determined according to the simulated human body scanning area.

[0056] Furthermore, in order to conform to the human body structure and facilitate the observation of the error distribution of CT image reconstruction in the later stage, the calibration parts are symmetrically distributed about the center plane 9 (dashed box in the figure) about the length direction of the testing fixture.

[0057] For example, the height and width of cuboids A to I are all 140mm, the length of cuboids A to G is all 40mm, and the length of cuboids H to I is all 20mm. Therefore, there are 4 splicing surfaces on both sides of the center surface 9 of the inspection fixture, and the calibration parts on the 8 splicing surfaces are symmetrically distributed about the center surface 9.

[0058] Specifically, the calibration components on all splicing surfaces are distributed uniformly. This eliminates the need to distinguish the splicing order of the six expansion components 2 and the two fixing components 3 during expansion and assembly. Furthermore, when using a coordinate measuring machine, only one set of parameters (such as measurement spacing) needs to be entered to complete the coordinate measurement of all calibration components. (Refer to...) Figures 2-3 The calibration parts on the eight splicing surfaces are evenly distributed in a 4×4 pattern, and the spacing between adjacent calibration parts is 40mm.

[0059] Furthermore, to facilitate the transfer of the tooling, a lifting structure, such as a handle or a protrusion, can be added to the inspection tooling.

[0060] Furthermore, in order to reproduce the placement of the detection fixture during CT scanning in the future (so that the experiment can be repeated), a positioning structure can be set on the detection fixture. For example, a crosshair can be set at the center of the outer side of the fixing part 3 to make it easy to check whether the detection fixture is aligned with the CT laser positioning line.

[0061] The calibration fixture coordinates obtained using the above methods and the calibration component image coordinates obtained from CT images are in different coordinate systems. Therefore, it is necessary to register them to the same coordinate system before obtaining the error distribution between the calibration component image coordinates and the fixture coordinates. Since the calibration components on the inspection fixture are symmetrically distributed, and the distribution of calibration components on each splicing surface is the same, the resulting point set lacks directionality. In particular, the calibration component images obtained from CT images are out of order, making it difficult to determine the correspondence between the calibration component image coordinates and the fixture coordinates.

[0062] To achieve registration between two coordinate systems, it is first necessary to determine the correspondence between point sets in the two coordinate systems according to certain criteria. Otherwise, all correspondences can only be listed by permutation and combination. Then, registration is performed according to different combination methods, and the coordinate error after registration is further calculated. Finally, the registration method with the smallest error is selected. It can be said that the amount of calculation is enormous.

[0063] Furthermore, the calibration components include a first calibration component and a second calibration component. The first calibration component determines the correspondence between the image coordinates and tooling coordinates of some or all of the calibration components. Then, based on the image coordinates and tooling coordinates of the calibration components with the determined correspondence, the tooling coordinate system and the image coordinate system are registered, which greatly reduces the amount of calculation.

[0064] In this invention, the first calibration component and the second calibration component need to be distinguishable in the CT image. This distinction can be achieved through any form, such as shape, size, or distribution. Based on the differences between the two calibration components in the image, the correspondence between the image coordinates and the tooling coordinates of some or all of the calibration components can be determined, thereby achieving efficient registration between the tooling coordinate system and the image coordinate system.

[0065] In addition, the first calibration piece can also be used to determine the placement direction and scanning direction of the detection fixture during CT scanning, ensuring that CT images can be reproduced later (repeatable experiments), and at the same time, it can be used to preliminarily determine whether the coordinates of the calibration piece image extracted from the CT image are consistent with the positional distribution of the calibration piece in the CT image.

[0066] For example, the calibration component is designed as a target ball 4 structure, and the first calibration component and the second calibration component are a large ball 41 and a small ball 42, respectively. The diameter of the large ball 41 is 12.7 mm and the diameter of the small ball 42 is 7.1 mm.

[0067] Preferably, the testing fixture is provided with at least three first calibration components at unequal distances, and the first calibration components are distributed on the reference component 1, so as to determine the correspondence between the image coordinates and the fixture coordinates based on the relative positional relationship between the first calibration components.

[0068] Reference Figures 2-3 The inspection fixture has four large spheres 41 spaced at unequal intervals, and the rest are small spheres 42. To distinguish the left and right order of the reference component 1 and to facilitate determining the placement and scanning direction of the inspection fixture during CT scanning, three large spheres 41 are arranged on the reference surface, with the remaining large sphere 41 located to the right of the reference component 1. By arranging the large spheres 41 on the reference component 1, the positional relationship between the large spheres 41 is fixed. Otherwise, as the splicing order of the extension components 2 changes, the position of the large spheres 41 would be random and not fixed, making it difficult to utilize the registration and orientation functions of the large spheres 41.

[0069] Furthermore, the exposed surface of the splice is provided with a receiving groove 5 for fixing the calibration component, which can increase the contact area between the calibration component and the splice and improve the firmness.

[0070] Reference Figures 2-3 The left and right sides of the reference part 1 and the right side of the extension part 2 are provided with 4×4 cylindrical grooves (small cylinders in the figure). The interval between adjacent grooves is 40mm and the groove diameter is 15mm. All target balls 4 are fixed in the corresponding grooves by hot melt glue (or other fixing methods).

[0071] Furthermore, in order to enable the splicing of the testing fixture, a receiving cavity 6 is provided on the splicing surface of the calibration component for receiving the calibration component.

[0072] For example, the inner side of the fastener 3 and the left side of the extension 2 are also provided with 4×4 cylindrical cavities (large cylinder in the figure), the diameter of the cylindrical cavity is larger than the diameter of the large ball 41, so as to reduce wear.

[0073] Furthermore, there are no restrictions on the detachable connection method, but it is necessary to ensure that the spliced ​​testing fixture does not become loose, so as to prevent changes in the position of the calibration parts and affect the testing accuracy.

[0074] For example, adjacent splicing parts and splicing parts and fixing parts 3 are fixedly connected by four hexagonal plastic screws 8 (to prevent metal artifacts). Each splicing part and fixing part 3 has four hexagonal threaded mounting slots 7 with a diameter of 10mm and a length of 30mm.

[0075] The present invention also provides an analysis method based on the above-mentioned CT image reconstruction accuracy analysis device, referring to... Figure 5 This includes the following steps:

[0076] S1. Based on the reference component 1, add extension components 2 one by one to realize the gradual expansion of the inspection fixture. During the expansion process, use a coordinate measuring machine to gradually measure the temporary coordinates of the target ball 4 on the exposed surface of the inspection fixture.

[0077] Reference Figure 4 a. First, using reference part 1 as the initial inspection fixture, a coordinate measuring machine is used to measure the temporary coordinates of the target ball 4 on its exposed surface. At this time, the target balls 4 on the left and right sides of the cuboid A are in a temporary coordinate system, namely coordinate system A1.

[0078] Reference Figure 4b. Then, based on the reference component 1, add the first extension component 2, that is, fix the cuboid B on the right side of the cuboid A, and use a coordinate measuring machine to measure the temporary coordinates of the target ball 4 on the left side of the cuboid A and the right side of the cuboid B. At this time, the target ball 4 on the left side of the cuboid A and the right side of the cuboid B are in a temporary coordinate system, namely coordinate system A2.

[0079] Reference Figure 4 c. Continue to add a second extension 2 on the basis of the reference part 1 and the extension part 2, that is, fix the cuboid C on the right side of the cuboid B, and use a coordinate measuring machine to measure the temporary coordinates of the target ball 4 on the left side of the cuboid A and the right side of the cuboid C. At this time, the target ball 4 on the left side of the cuboid A and the right side of the cuboid C are in a temporary coordinate system, namely coordinate system A3.

[0080] By analogy, the temporary coordinates of all target balls 4 in different temporary coordinate systems can be obtained. Since the newly added target balls 4 and the target balls 4 on the reference plane are in the same temporary coordinate system during each calibration, all temporary coordinate systems can be registered based on the target balls 4 on the reference plane, and finally the temporary coordinates of all target balls 4 can be registered to the same tooling coordinate system.

[0081] Specifically, in order to determine the correspondence between the target ball 4 and the temporary coordinates, the calibration order of the target ball 4 on all splicing surfaces is the same using a coordinate measuring machine. For example, in this embodiment, the calibration is performed row by row from top to bottom, and each row is calibrated from left to right.

[0082] For example, for coordinate systems A1 and A2, the coordinates u of the target sphere 4 on the reference surface are determined according to the first and second calibrations. i and u i The transformation matrix H between coordinate systems A1 and A2 is calculated using the ICP-SVD decomposition algorithm. 21 Then, the coordinates of the newly added target sphere 4 in coordinate system A2 are multiplied by the transformation matrix H. 21 Then you can transfer it to coordinate system A1.

[0083] And so on, each time using coordinate system A1 as the reference, coordinate system A... i The newly added target sphere's 4 coordinates are multiplied by the corresponding H. i1 The matrix is ​​used to transform it into coordinate system A1, thereby obtaining the tooling coordinates of all target balls 4 in tooling coordinate system A1 (in other embodiments, any temporary coordinate system can be used as a reference, and other temporary coordinate systems can be registered to the same tooling coordinate system).

[0084] Furthermore, in order to assess the accuracy of the coordinate calibration, the positional error between the two calibration processes can be obtained by subtracting the coordinates of the target sphere 4 on the reference surface from those of the two calibrations.

[0085] For example, each error calculation compares the newly calibrated reference surface target sphere 4 coordinates with the first calibrated reference surface target sphere 4 coordinates to ensure that the position error after each expansion is less than 0.2. Otherwise, it indicates that the temporary coordinate error obtained by the coordinate measuring machine is large and cannot be used for error analysis of subsequent image coordinates, and the temporary coordinate measurement needs to be repeated.

[0086] Specifically, the rigid registration process of the ICP-SVD decomposition algorithm is as follows:

[0087] ① Assume that there exists a corresponding set of point cloud coordinates x in coordinate systems X and Y. i and y i Then the centroid coordinates of this point cloud in the X and Y coordinate systems are respectively

[0088] ② Solve for the displacement vectors of each point relative to the centroid in coordinate systems X and Y: x i ′=x i -x, y i ′=y i -y;

[0089] ③ Using the centroid displacement vector, according to the formula Calculate the 3x3 transformation matrix M, which represents the positional relationship between point clouds in coordinate systems X and Y, similar to the covariance matrix;

[0090] ④ Perform SVD decomposition on matrix M: M = U∧V T Where U and V are both 3x3 orthogonal matrices;

[0091] ⑤ Calculate the rotation matrix R that rotates the positive direction of coordinate system Y to be consistent with the positive direction of coordinate system X: Where R * For all rotation matrices R, x i ′-Ry i The matrix with the smallest determinant value of ′;

[0092] ⑥ Calculate the displacement matrix t that translates the point space of coordinate system Y to the point space of coordinate system X: t = x i -R * y i .

[0093] In this process, the transformation matrix M for converting the point cloud in coordinate system Y to coordinate system X can be calculated through steps ①②③. Furthermore, the transformation process required to convert the point cloud in coordinate system Y to coordinate system X, namely the rotation and translation process, can be obtained through steps ④⑤⑥.

[0094] S2. After completing the expansion of the inspection fixture, continue to assemble the two side fixing parts 3, and then place the inspection fixture according to the CT laser positioning line for chest scanning imaging. Extract the image coordinates of all target balls 4 through CT image reconstruction.

[0095] Specifically, the ITK algorithm is mainly used to extract spherical coordinates. The specific implementation is to extract spherical coordinates from CT images using NP100_3D_System_Fixture software (internal software). Other software developed based on the ITK algorithm can also be used for coordinate extraction.

[0096] Reference Figure 6 The algorithm for extracting the image coordinates of target sphere 4 is as follows: To reduce the workload of subsequent region growing (OTSU algorithm), the image is first adjusted to a suitable window width and window level; then, the OTSU algorithm is used to segment the CT image to obtain the foreground and background. This operation achieves automatic segmentation, reduces the error of manually setting the threshold segmentation, and ensures that the target sphere 4 coordinates remain unchanged each time; in order not to change the size of the segmented target, and to separate the foreground and background and eliminate small objects, morphological opening operation is performed on the segmentation results; in order to optimize the segmentation results and smooth the sphere boundaries, connected component analysis is performed on the morphological results, and finally the image coordinates and radii of 128 ceramic spheres are extracted.

[0097] Furthermore, a sphere can be rendered using the radius to initially observe whether the positional relationship between the extracted spheres and the original sphere position in the CT image has shifted.

[0098] S3. Calculate the error between the tooling coordinates and the image coordinates of the target ball 4. Since they are in different coordinate systems, they need to be transformed into the same coordinate system.

[0099] Specifically, if we do not distinguish between large and small spheres, we can arrange and combine the image coordinates of all target spheres 4, resulting in a total of 128! sorting combinations. We then rigidly register the point set under each sorting with the tool coordinates, select the sorting combination with the smallest error as the final image coordinate sorting, and obtain the registration matrix with the smallest error. Then, we multiply the image coordinates of all target spheres 4 by this matrix and transform them into the tool coordinate system, thereby calculating the error between the two. It can be said that the amount of computation is enormous.

[0100] Furthermore, in order to reduce the amount of computation, we can first perform preliminary registration of the two coordinate systems based on the tooling coordinates and image coordinates of the four large spheres 41.

[0101] Specifically, the coordinates of the four large spheres 41 extracted from the image are out of order, so it is necessary to first determine the correspondence between the image coordinates and the tooling coordinates of the four large spheres 41.

[0102] For example, the coordinate correspondence of the large spheres 41 can be determined based on their relative positions. Specifically, the actual distance between each large sphere 41 can be obtained from its tooling coordinates, and the image distance between each large sphere 41 can be obtained from its image coordinates. Since the distances between the four large spheres 41 are not equal, comparing the two can determine the correspondence between the image coordinates and tooling coordinates of the large spheres 41. Then, the registration matrix T can be initially obtained by registering the large spheres 41 based on their image coordinates and tooling coordinates.

[0103] For example, the coordinate correspondence of the large sphere 41 can also be determined by permutation and combination. Specifically, the image coordinates (n1, n2, n3, n4) of the four large spheres 41 are arranged into 24 possible combinations using permutation and combination (i.e., n1n2n3n4, n1n2n4n3, n1n4n2n3, n1n4n3n2, n1n3n4n2, n1n3n2n4, n2n1n3n4, n2n1n4n3, n2n3n4n1, n2n2n3n4n1, n2n3n4n4n3, n ...3, n2n3n4n4n3, n2n3n4n4n3, n2n3n4n4n3, n2n3n4n4n3, n2n3n4n4n3, n2n3n4n 3n1n4, n2n4n1n3, n2n4n3n1, n3n1n2n4, n3n1n4n2, n3n2n1n4, n3n2n4n1, n3n4n1n2, n3n4n2n1, n4n1n2n3, n4n1n3n2, n4n2n1n3, n4n2n3n1, n4n3n1n2, n4n3n2n1). The image coordinates v of the large sphere 41 under each combination. i The tooling coordinates v of the four large spheres 41. i Perform rigid registration, select the combination with the smallest error after registration as the final sort, and obtain the registration matrix T corresponding to the smallest error.

[0104] Furthermore, the image coordinates q of all target spheres 4 i Multiplying by matrix T initially transforms it into the tooling coordinate system, yielding the first registration coordinates q for all target spheres 4 in the tooling coordinate system. i At this point, the first registration coordinates of ball 42 are q. i The coordinates are still disordered and do not yet correspond one-to-one with the tooling coordinates of ball 42.

[0105] Furthermore, for each first registration coordinate q i ′, calculate its tooling coordinates p with all target balls 4 according to the Euclidean distance formula. i The distance between them, based on the minimum distance, is used to register the first coordinates q of all target balls 4. i Reordering yields the coordinates p of the target sphere and tooling. i The second registration coordinate q determines the correspondence. i ″.

[0106] Furthermore, due to the second registration coordinate q i "This was initially registered based on the tooling coordinates and image coordinates of the four large spheres 41, but this registration process is not entirely accurate. Therefore, after determining the coordinate correspondence of all target spheres 4, a second registration q can be performed based on all target spheres 4." i "and tooling coordinates p" i Rigid registration was performed again to obtain a very accurate third registration coordinate q. i "′, from which the coordinate error of all target balls 4 is calculated. i =p i -q i ″′.

[0107] Specifically, by observing Error i The distribution parameters, such as the maximum, minimum, median, and mean values, can be used to evaluate the reconstruction accuracy of CT images. For example, we observe the Error... i The maximum and median values ​​are considered. If the value is within 0.4, it indicates that the application requirements are met. Otherwise, the CT reconstruction accuracy does not meet the requirements of the puncture surgery machine, and the CT images cannot be used.

[0108] In this embodiment, the tooling coordinates and image coordinates of the target sphere 4 are registered twice: the first registration of the large sphere 41 is a preliminary registration to reduce the amount of computation, and the second registration of all target spheres 4 is to improve the registration accuracy. In other embodiments, only one registration process can be performed, and the registration between the tooling coordinate system and the image coordinate system can be achieved based on some or all of the image coordinates and tooling coordinates of the target spheres 4.

[0109] This invention employs an expandable detection fixture to simulate human body scanning areas in CT images, and installs calibration components on the fixture to simulate human lesions. This allows for the gradual measurement and registration of temporary coordinates of the calibration components on the exposed surface of the fixture, ultimately obtaining the actual coordinates of all calibration components. Compared to directly measuring the actual coordinates of calibration components within a single fixture, this significantly improves measurement accuracy. By registering the measured actual coordinates with the CT image coordinates, the reliability of CT image reconstruction accuracy analysis is effectively enhanced, facilitating the full utilization of the superior performance of the surgical robot.

[0110] The above-described specific embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Various modifications, substitutions, and improvements made by those skilled in the art to the technical solutions of the present invention based on the provided textual description and drawings, without departing from the design concept and spirit of the present invention, should all fall within the scope of protection of the present invention.

Claims

1. A CT image reconstruction accuracy analysis device, characterized in that, include: The splicing components are detachably connected to each other to form an expandable testing fixture; Calibration components are distributed on the exposed surface of the splicing components; Before and after each expansion, a temporary coordinate system is constructed based on the calibration components on the exposed surface of the testing fixture; The calibration components on the exposed surface of the testing fixture include the same calibration components before and after expansion, as well as newly added calibration components. The same calibration components are used to register each temporary coordinate system to obtain the fixture coordinate system. The obtained fixture coordinate system is then registered with the image coordinate system in the CT image to obtain the error distribution between the calibration component image coordinates and the fixture coordinates.

2. The CT image reconstruction accuracy analysis device according to claim 1, characterized in that, The same calibration components include those that are always exposed during the expansion process, and the temporary coordinate system is registered based on the always exposed calibration components.

3. The CT image reconstruction accuracy analysis device according to claim 2, characterized in that, After each expansion, the exposed surface of the testing fixture only contains the always-exposed calibration component and the newly added calibration component.

4. The CT image reconstruction accuracy analysis device according to claim 3, characterized in that, The always-exposed calibration components are distributed on a splicing component that serves as a reference component. The remaining splicing components serve as extension components. The reference component is used as the initial testing fixture, and the extension components are added one by one to form a progressively expanding testing fixture.

5. The CT image reconstruction accuracy analysis device according to claim 4, characterized in that, The always-exposed calibration components are distributed on one reference surface of the reference component.

6. The CT image reconstruction accuracy analysis device according to claim 1, characterized in that, The process of registering the obtained tooling coordinate system with the image coordinate system in the CT image is as follows: Determine the correspondence between the tooling coordinates and image coordinates of some or all calibration parts, and perform registration between the tooling coordinate system and the image coordinate system.

7. The CT image reconstruction accuracy analysis device according to claim 6, characterized in that, The calibration components include a first calibration component and a second calibration component. The tooling coordinate system and the image coordinate system are initially registered based on the tooling coordinates and image coordinates of the first calibration component.

8. The CT image reconstruction accuracy analysis device according to claim 7, characterized in that, Based on the results of the initial registration, the correspondence between the tooling coordinates and the image coordinates of the second calibration part is determined, and a second registration between the tooling coordinate system and the image coordinate system is performed based on the tooling coordinates and image coordinates of all calibration parts.

9. The CT image reconstruction accuracy analysis device according to claim 1, characterized in that, It also includes a fastener, which is detachably connected to the splicing component, for protecting the exposed calibration components on the testing fixture after the expansion of the testing fixture is completed.

10. A method for analyzing the accuracy of CT image reconstruction, implemented based on the CT image reconstruction accuracy analysis device of claim 1, characterized in that, Includes the following steps: S1. Before and after each expansion, obtain the temporary coordinates of the calibration part on the exposed surface of the testing fixture, and register the temporary coordinates according to the same calibration part to obtain the fixture coordinates of the calibration part in the fixture coordinate system. S2. After completing the expansion of the testing fixture, extract the image coordinates of the calibration parts on the testing fixture through CT image reconstruction. S3. Register the tooling coordinates and image coordinates of the calibration part to the same coordinate system to obtain the error distribution between the image coordinates and tooling coordinates of the calibration part.

11. The CT image reconstruction accuracy analysis method according to claim 10, characterized in that, The registration of temporary coordinates based on the same calibration component specifically includes: The same calibration components include those that are always exposed during the expansion process, and temporary coordinates are registered based on the always exposed calibration components.

12. The CT image reconstruction accuracy analysis method according to claim 10, characterized in that, After completing the expansion of the testing fixture, a fixing component is further assembled on the testing fixture. The fixing component is detachably connected to the splicing component and is used to protect the exposed calibration components on the testing fixture.

13. The CT image reconstruction accuracy analysis method according to claim 10, characterized in that, The process of registering the tooling coordinates and image coordinates of the calibration part to the same coordinate system specifically involves: Determine the correspondence between the tooling coordinates and image coordinates of some or all calibration parts, and perform registration between the tooling coordinate system and the image coordinate system.

14. The CT image reconstruction accuracy analysis method according to claim 13, characterized in that, The calibration components include a first calibration component and a second calibration component. The tooling coordinate system and the image coordinate system are initially registered based on the tooling coordinates and image coordinates of the first calibration component.

15. The CT image reconstruction accuracy analysis method according to claim 14, characterized in that, Based on the results of the initial registration, the correspondence between the tooling coordinates and the image coordinates of the second calibration part is determined, and a second registration between the tooling coordinate system and the image coordinate system is performed based on the tooling coordinates and image coordinates of all calibration parts.

Images