A compact ultrasonic probe calibration system and method

Abstract

The application discloses a simplified ultrasonic probe calibration system and a method thereof. A model registration tool is designed, one end of the tool is inserted into a sharp needle, and the other end is fixed with a positioning sensor. The model registration tool becomes a positioning pen through simple calibration. The model registration tool is used to establish a calibration model coordinate system without an additional positioning sensor, so that the calibration of the ultrasonic probe is completed by using one positioning sensor, and the calibration cost is reduced. An octagonal calibration model with small volume and simple structure is designed. The model obtains multiple effective positioning points through one-time ultrasonic imaging, more calibration targets can be obtained by using less calibration times, and the calibration efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of medical devices, and more specifically to a simplified ultrasound probe calibration system and method. Background Technology

[0002] Due to the advantages of real-time imaging and radiation-free operation, 3D ultrasound is increasingly being used in surgical navigation and scoliosis screening. Existing 3D ultrasound images are primarily obtained by reconstructing sections of 2D ultrasound images. The accuracy of the reconstruction depends mainly on the accuracy of the 2D image localization. 3D ultrasound shows great potential in surgical navigation and localization.

[0003] Localization of two-dimensional images can be divided into two parts: one is tracking the ultrasonic probe, where a positioning sensor is typically attached to the probe to establish a probe coordinate system. The main tracking methods include optical positioning, magnetic positioning, mechanical positioning, and acoustic positioning. The second part is calibrating the ultrasonic imaging plane and the probe coordinate system. Traditional calibration models, such as point-based, cross-line-based, and N-line-based models, while simple in structure, only yield one valid data point per imaging, reducing calibration efficiency. Some calibration models specifically designed to improve efficiency, while shortening calibration time, suffer from complex structures, manufacturing difficulties, and large footprints. Furthermore, existing calibration methods generally require two sensors, one fixed to the ultrasonic probe and the other to the calibration model, increasing calibration costs. Summary of the Invention

[0004] To address the aforementioned problems in existing technologies, this invention proposes a simplified ultrasonic probe calibration system and method. This calibration system features a simple calibration model, high calibration efficiency, and requires only one positioning sensor.

[0005] The technical solution of this invention is: a simplified ultrasonic probe calibration system, comprising:

[0006] The calibration model includes two identical and symmetrically fitted calibration pools. Each calibration pool is hollow inside and has a convex octagonal edge with two mutually perpendicular axes of symmetry. Each of the four non-adjacent sides of the convex octagon has k conical holes, and the conical holes on opposite sides are centrally symmetrical about the center of the convex octagon. The diameter of the conical holes gradually decreases along the depth direction. A cylindrical hole penetrating the conical holes is opened on the outer surface of the calibration pool.

[0007] The model registration tool is used to register the calibration model, including a cylindrical pencil-shaped calibration rod and a hand grip end at the end of the calibration rod; the hand grip end has a groove for fixing the positioning sensor; the first end of the calibration rod is a frustum designed in the shape of a cone; a central hole for placing a fine needle is opened at the center of the top of the frustum.

[0008] An ultrasonic probe is used to image a calibration model.

[0009] As a preferred embodiment of the present invention, the side of the frustum is provided with a glue injection groove communicating with the central hole. The glue injection groove is used to add light-curing adhesive into the central hole of the frustum. The light-curing adhesive can fix the fine needle placed in the central hole.

[0010] This invention also provides a simplified calibration method for an ultrasonic probe calibration system, comprising the following steps:

[0011] A. The model registration tool is calibrated, and the position information of the needle tip relative to the positioning sensor is obtained after calibration;

[0012] B. Register the calibration model: Align the tip of the model registration tool from step A with the conical hole on the model to obtain the transformation matrix of the model coordinate system relative to the base coordinate system; the position of the calibration model should remain unchanged after registration, otherwise repeat step B;

[0013] C. Calibrate the ultrasound image: Remove the positioning sensor from the model registration tool and fix it on the ultrasound probe; the ultrasound probe images the calibration model to ensure that 2(k-1) points appear simultaneously in one ultrasound image, and the external control terminal reads the position information collected by the positioning device and the ultrasound image information collected by the ultrasound probe; change the probe pose and perform multiple imaging operations on the calibration model to obtain the transformation matrix of the ultrasound image coordinate system relative to the positioning sensor.

[0014] As a preferred embodiment of the present invention, step A specifically includes the following steps:

[0015] A1. Align the tip of the fine needle with a fixed point on a flat table, adjust the angle of the model registration tool, and the positioning device collects multiple sets of different position and orientation information. The position of the needle tip relative to the positioning sensor is then determined using the matrix least squares method.

[0016] A2. After completing the calibration operation in A1, use the data collected in A1 to perform calibration. Perform an error assessment; if the error is greater than 1 mm, repeat step A1 until the error is within 1 mm.

[0017] As a preferred embodiment of the present invention, step B specifically includes the following steps:

[0018] B1. Thread 2(k-1) thin wires through the cylindrical holes made on the side. Divide the sides with cylindrical holes and which are centrally symmetrical into a group. Then there are two groups on the model: Group 1 and Group 2. The two thin wires pass through the two outermost cylindrical holes of Group 1 respectively, connecting the two parallel sides of Group 1. The remaining cylindrical holes of Group 1 are connected to the cylindrical holes on Group 2 through thin wires. These thin wires are parallel to each other.

[0019] B2. Insert the tip of the fine needle into the center of the conical hole on the calibration model, adjust the angle of the model registration tool, and the positioning device collects multiple sets of position and orientation information. Continue inserting the tip of the fine needle into the next conical hole, repeating the above steps until all conical holes have been measured; obtain the transformation matrix of the base coordinate system relative to the model coordinate system through the Iterative Closest Point (ICP) algorithm.

[0020] B3. Compare the deviation between the calibrated position of the conical hole and the theoretical position of the conical hole. If the deviation is greater than 1mm, repeat step B1 until an accuracy of less than 1mm is obtained.

[0021] As a preferred embodiment of the present invention, step C specifically includes the following steps:

[0022] C1. After fixing the position sensor to the ultrasonic probe, image the calibration model; segment the k positioning points on each ultrasonic image to obtain the position P of the positioning points in the ultrasonic image coordinate system. image <4,1>, and it is assumed that the thickness of the ultrasound image is 0, i.e., P image The third row reading of e4,1> is 0;

[0023] C2. Perform linear fitting on the 2(k-1) positioning points segmented in step C1, and find the proportional relationship of these points forming 2-3 line segments;

[0024] C3. Based on the proportional relationship calculated in step C2, and combined with the principle of triangle similarity in the calibration model, since the model design parameters are known, the spatial position P of the positioning point in the model coordinate system can be determined. model <4,1>, and then the transformation matrix between the ultrasound image coordinate system and the positioning sensor coordinate system is obtained through ICP.

[0025] Compared with the prior art, the beneficial effects of the present invention are:

[0026] A model registration tool was designed, in which a needle is inserted at one end and a positioning sensor is fixed at the other. Through simple calibration, the model registration tool becomes a positioning pen. This tool enables the establishment of the model's coordinate system without the need for an additional positioning sensor, allowing for the calibration of an ultrasonic probe using only one sensor, thus reducing calibration costs. A small and simple calibration model was also designed, which obtains multiple effective positioning points through a single ultrasonic imaging, allowing for more calibration targets to be obtained with fewer calibration cycles, thereby improving calibration efficiency. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of coordinate system transformation.

[0028] Figure 2 A schematic diagram of the model registration tool.

[0029] Figure 3 This is a schematic diagram of the calibration model for an octagonal ultrasonic probe.

[0030] Figure 4 This is a schematic diagram of the calibration process.

[0031] In the diagram: 1. Hand grip; 2. Calibration rod; 3. Frustum; 4. Calibration pool; 5. Conical hole; 6. Cylindrical hole. Detailed Implementation

[0032] The present invention will be further described and illustrated below with reference to specific embodiments. The embodiments described are merely examples of the content of this disclosure and do not limit the scope of the invention. The technical features of each embodiment in the present invention can be combined accordingly, provided that there is no mutual conflict.

[0033] like Figure 1 The diagram illustrates a coordinate system transformation using a simplified ultrasonic probe calibration system and method according to the present invention. The present invention first provides a simplified ultrasonic probe calibration system, comprising:

[0034] The calibration model includes two identical and symmetrically fitted calibration cells 4, such as Figure 3 As shown, the calibration cell is hollow inside and has a convex octagonal edge. The convex octagon has two mutually perpendicular axes of symmetry. Each of the four non-adjacent sides of the convex octagon has k conical holes 5, and the conical holes 5 on opposite sides are centrally symmetrical about the center of the convex octagon. The diameter of the conical holes 5 gradually decreases along the depth direction. A cylindrical hole 6 is provided on the outer surface of the calibration cell, which passes through the conical holes 5.

[0035] Model registration tools are used to register calibration models, such as... Figure 2As shown, it includes a cylindrical pencil-shaped calibration rod 2 and a hand grip end 1 at the end of the calibration rod. The hand grip end 1 has a groove for fixing the positioning sensor. The first end of the calibration rod is a frustum 3 designed in the shape of a cone. The center of the top of the frustum 3 has a central hole for placing a fine needle. The side of the frustum 3 has a glue injection groove communicating with the central hole. The glue injection groove is used to add UV-curable glue into the central hole of the frustum 3. The UV-curable glue can fix the fine needle placed in the central hole.

[0036] An ultrasonic probe is used to image a calibration model.

[0037] This invention also provides a simplified calibration method for an ultrasonic probe calibration system, specifically including the following steps:

[0038] Step A calibrates the model registration tool, obtaining the position information of the needle tip relative to the positioning sensor; specifically, it includes the following steps:

[0039] A1. Align the tip of the fine needle with a fixed point on a flat table, adjust the angle of the model registration tool, and the positioning device collects multiple sets of different position and orientation information. The position of the needle tip relative to the positioning sensor is then determined using the matrix least squares method. The specific formula for calibration using the matrix least squares method is as follows:

[0040]

[0041] in, It is the positioning sensor coordinate system {O prcbe} relative to the base coordinate system {O base The transformation matrix of} is obtained directly through the positioning device; A <3n-3,3> B is the position matrix of the fixed point on the desktop in the base coordinate system; <3n-3，1＞ The estimated position matrix of the desktop fixed point in the base coordinate system; P probe <4,1> indicates that the tip of the fine needle is relative to {O} probe The position of}; P base <4,1> represents a fixed point on the desktop at {O base The position within}; since the position of the needle tip is always fixed, i.e., P base <4,1> remains unchanged, therefore it can be obtained through multiple measurements. To obtain calibration Where n is the total number of calibration samples; A <3n-3，3> T It is matrix A <3n-3，3> The transpose of .

[0042] A2. Align the tip of the fine needle with the fixed point on the desktop in step A1, adjust the angle of the model registration tool, and the positioning device collects multiple sets of different position and orientation information. Then, analyze the obtained... Perform an error assessment. If the error is greater than 1 mm, repeat step A1 until an accuracy of less than 1 mm is obtained. The specific error assessment formula is as follows:

[0043]

[0044] in, It is the position of the needle tip relative to the positioning sensor obtained in step A1; The estimated position of a fixed point on the desktop in the base coordinate system is calculated based on the n sets of data calibrated and sampled in step A1. It is the position of the desktop fixed point calibrated in step A1 in the base coordinate system and the estimated distance of the desktop fixed point in the base coordinate system, where i indicates that this is the data collected in the i-th time; This represents the position of the i-th calibrated desktop fixed point in the base coordinate system and the estimated distance of the desktop fixed point in the base coordinate system; n is the total number of calibration samples in step A1.

[0045] B. Register the calibration model: Align the tip of the model registration tool from step A with the conical hole on the model to obtain the transformation matrix of the model coordinate system relative to the base coordinate system; the position of the calibration model should remain unchanged after registration, otherwise repeat step B;

[0046] The calibration model requires 2(k-1) thin threads to be threaded through it, with each thread being parallel to the next. Two of the parallel threads pass through the two outer cylindrical holes 6 of the same group. The remaining threads connect the middle of this group and the other group.

[0047] Step B specifically includes the following steps:

[0048] B1. After installing the fine thread, insert the tip of the fine needle into the center of the conical hole 5 on the calibration model. Adjust the angle of the model registration tool. The positioning device collects multiple sets of position and orientation information. Continue inserting the tip of the fine needle into the next conical hole 5, repeating the above steps until all conical holes 5 have been measured. The transformation matrix between the base coordinate system and the model coordinate system is obtained through the Iterative Closest Point (ICP) calibration algorithm. The specific formula is as follows:

[0049]

[0050] Where n is the number of samples taken for a single conical hole, and m represents the number of conical holes. P model <4,1> is the theoretical position of the conical hole in the calibration model, which is determined by the SolidWorks model. To determine the position of the j-th conical hole in the base coordinate system; The distance between the calibrated j-th conical hole in the model coordinate system and the theoretical j-th conical hole.

[0051] B2. Compare the deviation between the calibrated position of conical hole 5 and the theoretical position of conical hole 5. If the deviation is greater than 1mm, repeat step B1 until an accuracy of less than 1mm is obtained. The error is calculated after IPC iteration in step B1.

[0052] C. Calibrate the ultrasound images. Remove the positioning sensor from the model registration tool and fix it to the ultrasound probe. Image the calibration model, ensuring that 2(k-1) points appear simultaneously in one image, and simultaneously read the position information acquired by the positioning device and the ultrasound image information. Change the probe pose and image the calibration model multiple times to calibrate the transformation matrix of the ultrasound image coordinate system relative to the positioning sensor. After fixing the position sensor to the ultrasound probe, step C specifically includes the following steps:

[0053] C1. Divide each ultrasound image into 2(k-1) positioning points to obtain the position P of the positioning points in the ultrasound image coordinate system. image Furthermore, it is assumed that the thickness of the ultrasound image is 0, i.e., P image The third row reading of <3,1> is 0;

[0054] C2. Perform linear fitting on the 2(k-1) positioning points segmented in step C1, and find the proportional relationship of these points forming 2-3 line segments;

[0055] C3. Based on the proportional relationship calculated in step C2, and combined with the principle of triangle similarity in the calibration model, since the model design parameters are known, the spatial position P of the positioning point in the model coordinate system can be determined. model Then, the transformation matrix between the ultrasound image coordinate system and the positioning sensor coordinate system is obtained through ICP. The specific formula is as follows:

[0056]

[0057]

[0058] in, It is the scaling factor of the ultrasound image; The ultrasound image coordinate system is relative to {O} probe The transformation matrix of}. m is the number of positioning points. Transformation is performed via ICP. The parameters of the matrix are optimized to obtain the final calibration result.

[0059]

Example 1

[0060] The position sensor magnetic marker was fixed to the model registration tool. The tip of the fine needle of the model registration tool was aligned with a point on the table. The magnetic marker recorded 8 sets of data, and the position of the tip of the fine needle relative to the positioning sensor was calibrated. The calibration error was calculated to be 0.48 mm. Figure 3 The calibration model was placed in an aqueous environment. The operator held the model registration tool, aligning the tip of the tool with the conical holes of the calibration model, and calibrated all 12 conical holes sequentially. Five sets of sensor position information were recorded for each conical hole, for a total of 60 sets of data. The transformation matrix between the base coordinate system and the model coordinate system was obtained through ICP. Remove the magnetic marker and fix it to the ultrasound probe. Image the calibration model, ensuring four marker points are present in the imaging plane for each image. Figure 4 As shown in (a), linear optimization is performed on the marker points, as follows: Figure 4 As shown in (b), the proportional relationship of the line segment formed by the four points is obtained. Since the positions of each conical hole in the model are known in the model coordinate system, according to... Figure 4 (c) The principle of similarity of triangles can be used to determine the theoretical position of the marker point in the model coordinate system. Furthermore, the transformation matrix between the ultrasound image coordinate system and the positioning sensor coordinate system can be obtained through ICP. Figure 4 (d) Displaying the calibration results Transform the points in the ultrasound image coordinate system to the model coordinate system.

[0061] The above-described embodiments are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. Those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. An ultrasonic image calibration method based on an ultrasonic probe calibration system, characterized by, The ultrasonic probe calibration system comprises: The calibration model comprises two identical and symmetrically fitted calibration pools (4), which are hollow inside and have a convex octagonal edge with two mutually perpendicular and symmetric axes; each of the four edges of the convex octagon that are not adjacent to each other is provided with k conical holes (5), and the conical holes (5) on the opposite edges are centrally symmetric about the center of the convex octagon, and the diameters of the conical holes (5) gradually decrease in the depth direction; a cylindrical hole (6) is formed in the outer side of the calibration pool and penetrates the conical hole (5); The model registration tool is used for registering the calibration model, and comprises a cylindrical pencil-shaped calibration rod (2) and a hand-held end (1) arranged at the end of the calibration rod, the hand-held end (1) is provided with a groove for fixing a positioning sensor, and the first end of the calibration rod is a circular truncated cone (3) designed in a circular truncated cone shape, and a central hole for placing a fine needle is formed in the center of the top end of the circular truncated cone (3); The ultrasonic probe is used for imaging the calibration model; The calibration method comprises the following steps: A, calibrating the model registration tool to obtain the position information of the fine needle tip relative to the positioning sensor; B, registering the calibration model: sequentially aligning the fine needle tip of the model registration tool in step A with the conical holes on the model to calibrate the conversion matrix of the model coordinate system relative to the base coordinate system; after the registration is completed, the position of the calibration model should be kept unchanged, otherwise step B is repeated; Step B specifically comprises the following steps: B1, threading through the cylindrical holes (6) formed on the side, a total of 2(k-1) fine lines are threaded; the edges with the cylindrical holes and the central symmetry are divided into a group, so there are two groups, group 1 and group 2, on the model; two fine lines respectively penetrate the two outermost cylindrical holes (6) of group 1, and group 1 is connected in parallel through the two edges; the remaining cylindrical holes of group 1 are connected to the cylindrical holes of group 2 through fine lines, and these fine lines are parallel to each other; B2, insert the fine needle tip into the center of the conical hole (5) on the calibration model, adjust the model registration tool angle, and position the device to collect multiple sets of position and direction information. Continue to insert the fine needle tip into the next conical hole (5) and repeat the previous steps until all the conical holes (5) are measured. Through the iterative closest point algorithm (ICP), the conversion matrix of the base coordinate system relative to the model coordinate system is calibrated B3, compare the deviation between the calibrated conical hole (5) position and the theoretical conical hole position, if the deviation is greater than 1mm, repeat step B1 until the accuracy is below 1mm; C, calibrating the ultrasonic image: removing the positioning sensor from the model registration tool and fixing it on the ultrasonic probe; the ultrasonic probe images the calibration model to ensure that 2(k-1) points appear in one ultrasonic image at the same time, and the external control end reads the position information collected by the positioning device and the ultrasonic image information collected by the ultrasonic probe; changing the position of the probe, imaging the calibration model multiple times, and calibrating the conversion matrix of the ultrasonic image coordinate system relative to the positioning sensor; Step C specifically comprises the following steps: C1. Image the phantom with the position sensor fixed behind the ultrasound probe; segment the k landmarks on each ultrasound image to get their positions P in the ultrasound image coordinate system image <4,1> and consider the ultrasound image thickness to be 0, i.e. P image The third line reading of <4,1> is 0; C2, linearly fitting the 2(k-1) positioning points segmented in step C1, and calculating the proportional relationship of the 2k-3 line segments composed of these points; C3, according to the proportional relationship calculated in step C2, combined with the triangle similarity principle existing in the calibration model, since the model design parameters are known, the spatial position P of the positioning point in the model coordinate system can be solved model <4,1>, and then the conversion matrix of the ultrasound image coordinate system relative to the positioning sensor coordinate system is solved by ICP 2. The ultrasonic image calibration method of claim 1, wherein, The side of the circular truncated cone (3) is provided with a glue injection groove in communication with the central hole, and the glue injection groove is used for adding light-curing glue into the central hole of the circular truncated cone (3), and the light-curing glue can fix the fine needle placed in the central hole.

3. The ultrasonic image calibration method of claim 1, wherein, Step A specifically comprises the following steps: A1, the fine needle tip is aimed at a fixed point on a flat table, the model registration tool angle is adjusted, a plurality of different position and direction information is obtained by the positioning device, and the position of the fine needle tip relative to the positioning sensor is calibrated by matrix least square method A2, after completing the calibration operation of A1, using the data collected in A1 to calibrate the calibration obtained in A1 error evaluation; if the error is greater than 1 mm, repeat step A1 until the error is within 1 mm.

4. The ultrasonic image calibration method of claim 3, wherein, The basic formula for calibration in A1 is: Wherein, is the conversion matrix of the positioning sensor coordinate system {O probe} relative to the base coordinate system {O base}, which is directly collected by the positioning device; A <3n-3,3> is the position matrix of the desktop fixed point in the base coordinate system; B <3n-3,1> is the estimated position matrix of the desktop fixed point in the base coordinate system; P probe <4,1> represents the position of the fine needle tip relative to {O probe}; P base <4,1> represents the position of the fixed point on the desktop in {O base}; since the position of the fine needle tip is always fixed, i.e. P base <4,1> remains unchanged, so can be obtained by multiple measurements to calibrate where n is the total number of calibration samples; A <3n-3,3> T is the transpose matrix of matrix A <3n-3,3> .

5. The ultrasonic image calibration method of claim 3, wherein, The error evaluation formula STD in A2 is specifically: wherein, is the position of the fine needle tip relative to the positioning sensor obtained from the calibration in step A1 ; is the estimated position of the fixed point on the table in the base coordinate system calculated from the n sets of data sampled in the calibration in step A1 ; is the distance between the position of the fixed point on the table in the base coordinate system obtained from the i-th calibration in step A1 and the estimated position of the fixed point on the table in the base coordinate system, wherein i represents that it is the i-th collected data; represents the distance between the position of the fixed point on the table in the base coordinate system obtained from the i-th calibration in step A1 and the estimated position of the fixed point on the table in the base coordinate system; n is the total number of calibration sampling in step A1.

6. The ultrasonic image calibration method of claim 1, wherein, The calibration formula by ICP in Step B2 is: Wherein, n is the sampling number of single conical hole (5), m represents the number of conical holes (5); P model <4,1> is the theoretical position of the calibration model conical hole, which is determined by the SolidWorks model; is the position of the jth conical hole in the base coordinate system obtained by calibration; is the distance between the jth conical hole obtained by calibration and the theoretical jth conical hole in the model coordinate system.

7. The ultrasonic image calibration method of claim 1, wherein, The calibration formula in Step C3 is: in, It is the scaling factor of the ultrasound image; The ultrasound image coordinate system is relative to {O} probe The transformation matrix of}; m is the number of positioning points, which is achieved through ICP. The parameters of the matrix are optimized to obtain the final calibration result.

Images