A calibration system and calibration method for a dual-plane ultrasonic probe

Abstract

The present invention discloses a calibration system based on a dual-plane ultrasonic probe, comprising an orthogonal dual-plane ultrasonic probe, a magnetic field generator, a magnetic navigation positioning receiver, a calibration phantom, and a container. The calibration phantom is a rectangular parallelepiped fixed to the bottom of the container and immersed in the ultrasonic imaging medium in the container; the calibration phantom is divided into an upper and lower part, each of which is interspersed with an N-shaped wire group; the two imaging planes of the dual-plane ultrasonic probe are perpendicular to each other and can intersect with the two wire groups of the calibration phantom respectively; the magnetic navigation positioning receiver is fixed to the middle of the dual-plane ultrasonic probe; the magnetic navigation positioning probe is used to select the surface points of the phantom, so as to locate the calibration phantom; the magnetic field generator is placed on one side of the container to ensure that its magnetic field range can cover the magnetic navigation positioning receiver and the magnetic navigation positioning probe on the ultrasonic probe. This scheme is conducive to the standardized calibration data acquisition and annotation process; and is conducive to the high-precision calculation of the coordinate conversion matrix between the magnetic navigation positioning receiver and the ultrasonic image.

Description

Technical Field

[0001] The invention relates to a calibration system and a calibration method for an orthogonal dual-plane ultrasonic probe, and belongs to the technical field of medical equipment and computer vision. Background Art

[0002] Prostate biopsy is the gold standard for the diagnosis of prostate cancer. Ultrasound images are often used to guide prostate biopsy surgery because of their convenience in acquisition and real-time image quality. At present, the clinical use of prostate biopsy systems that combine preoperative MR and intraoperative ultrasound fusion requires the acquisition of spatial position information of two-dimensional ultrasound images. By fixing the electromagnetic positioning sensor on the ultrasound probe, the spatial position information of the sensor can be obtained in real time. In addition, the ultrasound probe needs to be calibrated to obtain the conversion relationship from the sensor coordinate system to the ultrasound image coordinate system, and then obtain the spatial position information of the ultrasound image.

[0003] At present, the common ultrasound image calibration method is to scan the N-line body membrane with an ultrasound image in a water tank to form several bright spots, identify and calculate the spatial position information of the bright spots, and combine it with its pixel coordinate information in the ultrasound image to calculate the calibration matrix.

[0004] In the existing technology, the ultrasonic calibration phantom is mainly designed for a single-plane ultrasonic probe. For a dual-plane ultrasonic probe, since its two imaging planes are orthogonal to each other, the field of view is limited, and it cannot be fully immersed in the water tank, it is difficult to find a suitable angle and position to scan all points of the N-line phantom. Due to the poor imaging quality of ultrasonic images, there are problems such as reflections and artifacts, and the bright spot area is too large when the ultrasonic section is not completely perpendicular to the N line, so the accuracy of feature point recognition is poor. The above reasons lead to difficulties in calibrating image acquisition and low accuracy of the calculated calibration matrix. Summary of the invention

[0005] In view of the above problems, an object of the present invention is to provide a convenient and easy-to-use dual-plane ultrasound probe calibration system and a robust and high-precision calibration method.

[0006] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a dual-plane ultrasonic probe calibration system, comprising an orthogonal dual-plane ultrasonic probe, which has two ultrasonic imaging sections, and the two sections are orthogonal to each other; a magnetic field generator, which is used to generate a magnetic field for electromagnetic positioning; a magnetic navigation positioning receiver, whose rigid body is fixed on the dual-plane ultrasonic probe, and can obtain its own spatial position and posture parameters in real time in the magnetic field generated by the magnetic field generator, so that after calibration, it can be used to obtain the spatial posture parameters of the two ultrasonic sections; a magnetic navigation positioning probe, which can obtain the spatial position parameters of the needle tip in real time in the magnetic field generated by the magnetic field generator, for positioning the calibration phantom; the calibration phantom is a rectangular parallelepiped; a container and a fixing device, which are used to hold the ultrasonic imaging medium.

[0007] Furthermore, the calibration phantom is made of resin material and is a rectangular parallelepiped as a whole, divided into two upper and lower areas, each with a square hole, and the side walls of the holes are provided with a number of inverted triangular through holes, which are interspersed with nylon threads.

[0008] Furthermore, the holes in the upper and lower areas of the calibration phantom are all punched on the side walls, and the wires are inserted to form 5 groups of N-shaped wires. And for the upper and lower areas of the phantom, the 5 groups of N-shaped wires respectively formed by nylon wires have the same shape and size and are perpendicular to each other. The upper area is used to collect the data required for the sagittal calibration of the ultrasound probe, and the lower area is used to collect the data required for the cross-sectional calibration of the ultrasound probe.

[0009] Furthermore, the wires passing through the triangular through holes of the calibration phantom are made of nylon, and the diameter is similar to the ultrasonic wave length used by the ultrasonic probe, so that the intersection with the nylon wire is clearer in the ultrasonic image.

[0010] Furthermore, the container and the fixing device are high enough to hold the calibration phantom and the fixing device. The fixing device can fix the calibration phantom to the bottom of the container. The container can be used to hold hot water at human body temperature and immerse the calibration phantom. The four walls of the container are provided with absorbing materials to reduce ultrasonic reflection.

[0011] A calibration method based on a dual-plane ultrasonic probe calibration system comprises the following steps:

[0012] Step (1) calibration data acquisition: using the above-mentioned dual-plane ultrasonic probe calibration system to collect the raw data required for calibration;

[0013] Step (2) calibration data annotation: automatically or non-automatically, for a number of points formed by the intersection of the ultrasound cross section and the N-shaped line of the calibration phantom, automatically or non-automatically, according to a specific geometric relationship, identify and annotate their coordinates in the image;

[0014] Step (3) calibration data preprocessing: according to the distance geometric relationship between a number of points formed by the intersection of the ultrasonic cross section and the N-shaped lines of the calibration phantom, the coordinates of a number of feature points formed by the intersection with the hypotenuse of each group of N lines in the phantom coordinate system are calculated;

[0015] Step (4) Iterative optimization using the least squares method to remove outliers and calculate the calibration matrix: Calculate the coordinate transformation matrix from the magnetic navigation receiver coordinate system to the ultrasound image coordinate system;

[0016] Step (5) averaging the dual-plane calibration matrices: based on the prior information that the two imaging planes of the ultrasound probe are orthogonal to each other, averaging the calibration matrices obtained for the two probes;

[0017] Furthermore, the specific process of step (1) is as follows:

[0018] Step (1.1) vertically fix the calibration phantom in a container so that the imaging medium immerses the phantom;

[0019] Step (1.2) establishes a magnetic field generator coordinate system based on the magnetic field generator;

[0020] Step (1.3) uses a needle probe to collect phantom positioning data, including the coordinates of several points on the right surface, upper surface, and front surface of the rectangular parallelepiped calibration phantom;

[0021] Step (1.4) ensure that the magnetic navigation positioning receiver is fixed in the middle of the ultrasound probe;

[0022] Step (1.5) collects cross-sectional calibration data: the ultrasonic cross-section is made to intersect with the N line of the lower half of the calibration phantom.

[0023] Under the premise of ensuring that the ultrasound image intersects with all N-shaped lines and the intersection points are clearly visible, adjust the three translational degrees of freedom and three rotational degrees of freedom of the ultrasound probe respectively to collect several images;

[0024] Step (1.6) collects sagittal calibration data: the ultrasonic sagittal plane is made to intersect with the N line of the upper part of the calibration phantom.

[0025] Under the premise of ensuring that the ultrasound image intersects with all N-shaped lines and the intersection points are clearly visible, adjust the three translational degrees of freedom and three rotational degrees of freedom of the ultrasound probe respectively to collect several images;

[0026] Furthermore, the specific process of step (2) is as follows: for the foreground points in a single frame image, if ABC, DEF, GHI, JKL, MNO, ADG, BEH, and CFI are collinear, and AC|DF|GI, and AG|BH|CI are present, then the interference of other noise points that do not meet the above rules can be excluded during annotation, and the foreground points can be identified more accurately according to the collinearity and parallelism rules.

[0027] Furthermore, the specific process of step (4) is as follows:

[0028] Step (4.1), establish the magnetic field generator coordinate system T (transmitter) based on the magnetic field transmitter, and establish the phantom reconstruction coordinate system C (reconstruction) based on the calibration phantom. Figure 2 As shown in the figure, the receiver coordinate system R (receiver) is established based on the magnetic navigation positioning receiver, and the image coordinate system I (image) is established based on the transverse and sagittal ultrasound images. For any feature point, there is a coordinate transformation: p c =T ct *T tr*T ri *p i ;

[0029] Step (4.2), given the calibration phantom parameter information, use the acquired coordinates of the magnetic field generator coordinate system of 30 points on the right surface, upper surface, and front of the cuboid phantom to calculate the transformation matrix Tct from the transmitter coordinate system to the phantom reconstruction coordinate system;

[0030] Step (4.3), based on the coplanarity of the five points, filter out outliers. The five feature points in the calibration image should belong to the same plane. Use the coordinates of the magnetic field generator coordinate system of these five points to fit the plane. If the calculated distance between any point and the plane is greater than a specific threshold, the frame of image data is discarded;

[0031] Step (4.4), use iterative least squares fitting to transform the matrix Tri from the image coordinate system to the sensor / receiver coordinate system;

[0032] Step (4.5), use the calibration matrix T obtained in step 6 above ri , transform the existing key points and reconstruct their coordinates in the phantom coordinate system;

[0033] Step (4.6), for the calculated coordinates of the phantom coordinate system, the points that differ greatly from the true values ​​are removed as outliers;

[0034] Step (4.7), repeat (4.4) to (4.6) until no outliers are filtered out, and obtain the calibration matrix T of the cross section or sagittal plane ri .

[0035] Furthermore, the specific process of step (5) is as follows:

[0036] Step (5.1), for the cross section and sagittal plane, the calibration matrix is ​​in the form of a 4*4 matrix in Its physical meanings are the direction vectors of the ultrasonic probe axis, directly above, and directly to the right in the receiver coordinate system;

[0037] Step (5.2), let Construct new cross-sectional and sagittal calibration matrices

[0038] The present invention adopts the above technical solution, which has the following advantages:

[0039] 1. The calibration phantom in the present invention is standardized in design and has low production cost, and the design of the inverted triangular holes on the side walls ensures the accuracy of the coordinates of the interlaced nylon lines;

[0040] 2. The calibration phantom in the present invention has an N-shaped line that is divided into two areas, upper and lower, and is perpendicular to each other, and is suitable for the case where the cross section and sagittal plane of the dual-plane calibration phantom are perpendicular to each other. After the calibration phantom is fixed, the ultrasound probe is vertically inserted into the imaging medium, and the N-shaped lines in the upper and lower areas of the calibration phantom can be used to collect calibration images required for the two planes respectively. The ultrasound imaging plane is roughly perpendicular to the N-shaped line, so the feature points in the image are clearer and easier to identify;

[0041] 3. The calibration image acquisition process proposed in the present invention systematically specifies the movement mode of the probe during the image acquisition process, and adjusts the three translational degrees of freedom and the three rotational degrees of freedom respectively, so as to collect a large amount of ultrasound data in different positions, which is beneficial to the accuracy of the calibration matrix solution;

[0042] 4. The calibration phantom positioning method proposed in the present invention avoids the difficulty of manually positioning specific points and improves the stability of phantom position calculation by using multi-point coordinates to fit the three planes of the rectangular phantom by least squares respectively;

[0043] 5. The threading method of the five groups of N-shaped lines of the calibration phantom provided by the present invention provides a priori information about the verticality of the N-shaped lines, which is helpful for identifying key points when manually annotating the calibration image;

[0044] 6. The calibration image preprocessing method proposed in the present invention improves the calculation accuracy of the calibration matrix by screening out outliers;

[0045] 7. The dual-plane vertical calibration post-processing proposed in the present invention reduces the calculation error of the calibration matrix by introducing the information of the dual-plane verticality of the ultrasonic probe. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A schematic diagram of the placement of all hardware of the calibration system described in the present invention;

[0047] Figure 2 Using the N-shaped line of the lower half of the calibration phantom, the transverse calibration data of the ultrasound probe is collected (left picture); using the N-shaped line of the sang half of the calibration phantom, the sagittal calibration data of the ultrasound probe is collected (right picture)

[0048] Figure 3 Schematic diagram of several coordinate systems involved in the calibration process;

[0049] Figure 4 A three-dimensional view of the calibration phantom described in the present invention, excluding the thread group interspersed therein;

[0050] Figure 5 It is the front view of the calibration phantom described in the present invention;

[0051] Figure 6 It is a right view of the calibration phantom described in the present invention;

[0052] Figure 7 It is a schematic diagram of all the silk thread groups interspersed in the phantom;

[0053] Figure 8 A schematic diagram of a group of N-shaped lines interspersed in the phantom;

[0054] Fig. 9 This is an example of a single frame sagittal calibration image acquired;

[0055] Fig.10 Flowchart of the whole calibration process.

[0056] In the figure: 1. orthogonal dual-plane ultrasound probe, 2. magnetic field generator, 3. magnetic navigation positioning receiver, 4. ultrasound scanning section, 5. calibration phantom, 6. container. DETAILED DESCRIPTION

[0057] The present invention is described in detail below in conjunction with the accompanying drawings. However, the accompanying drawings are provided only for a better understanding of the present invention and should not be understood as limiting the present invention.

[0058] Example 1: See Figure 1 , 4 , 5, 6, 7, a dual-plane ultrasound probe calibration system, comprising an orthogonal dual-plane ultrasound probe 1, which has two ultrasound imaging sections 4, and the two sections are orthogonal to each other; a magnetic field generator 2, used to generate a magnetic field for electromagnetic positioning; a magnetic navigation positioning receiver 3, whose rigid body is fixed on the dual-plane ultrasound probe, and can obtain its own spatial position and posture parameters in real time in the magnetic field generated by the magnetic field generator, so that after calibration, it can be used to obtain the spatial posture parameters of the two ultrasound sections; a magnetic navigation positioning probe, which can obtain the spatial position parameters of the needle tip in real time in the magnetic field generated by the magnetic field generator, for positioning the calibration phantom; a calibration phantom 5, which is a rectangular parallelepiped; a container 6 and a fixing device, used to hold an ultrasound imaging medium.

[0059] The calibration phantom 5 is made of resin material, such as Figure 4 , 5 As shown in , 6 and 7, the whole is a rectangular parallelepiped, which is divided into two upper and lower areas, each with a square hole. The side walls of the hole are provided with a number of inverted triangular through holes, which can be used to insert silk threads.

[0060] The calibration phantom 5, such as Figure 7As shown, the holes in the upper and lower regions are all punched in the side walls, and the wires are inserted to form 5 groups of N-shaped wires. And the 5 groups of N-shaped wires formed by nylon wires inserted in the upper and lower regions of the phantom are of the same shape and size and are perpendicular to each other.

[0061] The calibration phantom 5, such as Figure 7 As shown, the nylon lines are interlaced to form 5 groups of N-shaped lines, of which 3 groups of N-shaped lines are parallel to each other, namely Figure 7 There are three groups of N-shaped lines, C, D, and E; the other two groups of N-shaped lines have opposite hypotenuse directions to the aforementioned N-shaped lines, i.e. Figure 7 There are two groups of N-shaped lines, A and B. Figure 7 As shown, from inside to outside, there are 5 groups of N-shaped lines, namely A, B, C, D, and E.

[0062] The wires passing through the triangular through holes of the calibration phantom 5 are made of nylon and have a diameter similar to that of the ultrasonic wave used by the ultrasonic probe.

[0063] The container 6 and the fixing device are high enough to hold the calibration phantom and the fixing device. The fixing device can fix the calibration phantom to the bottom of the container. The container can be used to hold hot water at human body temperature and immerse the calibration phantom. The four walls of the container are provided with absorbing materials to reduce ultrasonic reflection.

[0064] Example 2: See Figure 2 , 3 , 8, 9, 10, based on the above dual-plane ultrasonic probe calibration system, the present invention also provides a calibration method based on the dual-plane ultrasonic probe calibration system, comprising the following steps:

[0065] (1) Calibration data acquisition: Use the above-mentioned dual-plane ultrasonic probe calibration system to collect the raw data required for calibration, as follows:

[0066] (1.1) Fix the calibration phantom vertically in a container and immerse it in clean water. Heat the water to reach human body temperature of 37°C;

[0067] (1.2) Taking the magnetic field generator as the reference, establish the magnetic field generator coordinate system;

[0068] (1.3) Use a needle probe to collect the phantom positioning data. That is, arbitrarily collect the coordinates of 30 points on the right surface, upper surface, and front surface of the rectangular phantom using the magnetic field generator coordinate system;

[0069] (1.4) Fix the magnetic navigation positioning receiver in the middle of the ultrasound probe;

[0070] (1.5) Collect cross-sectional calibration data, such as Figure 2 (left) as shown;

[0071] (1.5.1) Insert the ultrasound probe vertically into the water, with the front half submerged in the water. Adjust the position of the ultrasound probe so that its cross-sectional field of view intersects with the five groups of N lines in the lower half of the phantom, and a total of 15 bright spots corresponding to the five groups of N lines can be clearly seen in the cross-sectional imaging;

[0072] (1.5.2) Collect a total of 360 images at a frequency of 0.3s per frame, and collect the position data of the magnetic navigation positioning receiver at the same time. During this period, translate the ultrasound probe in the x, y, and N directions of the calibration phantom coordinate system, rotate the ultrasound probe with the ultrasound probe axis as the axis, and the ultrasound probe imaging plane facing the phantom, pitch the ultrasound probe, and the ultrasound probe imaging plane facing the phantom, rotate the ultrasound probe left and right, and ensure that 15 bright spots are clearly visible;

[0073] (1.6) Collect sagittal calibration data, such as Figure 2 (right) shows:

[0074] (1.6.1) Insert the ultrasound probe vertically into the water, with the front half submerged in the water. Adjust the position of the ultrasound probe so that its sagittal view intersects with the five groups of N lines in the lower upper part of the phantom, and a total of 15 bright spots corresponding to the five groups of N lines can be clearly seen in the cross-sectional imaging;

[0075] (1.6.2) Collect a total of 360 images at a frequency of 0.3s per frame, and collect the position data of the magnetic navigation positioning receiver at the same time. During this period, translate the ultrasound probe in the x, y, and N directions of the calibration phantom coordinate system, rotate the ultrasound probe with the ultrasound probe axis as the axis, and the ultrasound probe imaging plane facing the phantom, pitch the ultrasound probe, and the ultrasound probe imaging plane facing the phantom, rotate the ultrasound probe left and right, and ensure that 15 bright spots are clearly visible;

[0076] (2) Calibration data labeling: After the calibration data collection is completed, the image needs to be manually labeled. Fig. 9 As shown in the figure, taking the cross-sectional ultrasound section calibration image as an example, the ultrasound section intersects with 5 groups of N lines to form 15 bright spots A to O, and the pixel coordinates of these 15 points in the image coordinate system are manually annotated. Figure 7 The specified 5 groups of N lines are interlaced. Fig. 9 The foreground points in the sagittal plane annotation data shown are collinear with ABC, DEF, GHI, JKL, MNO, ADG, BEH, and CFI, and AC∥DF∥GI, and AG∥BH∥CI. When manually annotating, the interference of other noise points that do not meet the above rules can be excluded, and the foreground points can be identified more accurately according to the collinearity and parallelism rules;

[0077] (3) Calibration data preprocessing: For the 15 points formed by the intersection of the ultrasound cross section and the 5 groups of N lines, calculate the coordinates of the 5 points formed by the intersection with the hypotenuse of each group of N lines in the calibration phantom coordinate system. Figure 6 As shown in the figure, a set of N lines is taken as an example, including 3 edges, AC, EG, and IK, where points A, C, E, G, I, and K are the intersection points of the N lines and the inner wall of the calibration phantom. During the calibration image acquisition process, the ultrasonic section BFJ intersects with this set of N lines at points B, F, and J. In addition, the extended lines of the edges AC and EG intersect at point D, and the extended lines of the edges EG and IK intersect at point H, then △BDF~△JHF, therefore, Under the premise that the shape of the calibration phantom is known, it is easy to calculate the coordinates of points D and H. Then we have

[0078] (4) Least squares iterative optimization, removing outliers and calculating the calibration matrix: calculating the coordinate transformation matrix from the magnetic navigation receiver coordinate system to the ultrasound image coordinate system;

[0079] (4.1) The magnetic field generator coordinate system T (transmitter) is established with the magnetic field transmitter as the reference, and the phantom reconstruction coordinate system C (reconstruction) is established with the calibration phantom as the reference. Figure 3 As shown, the receiver coordinate system R (receiver) is established based on the magnetic navigation positioning receiver, and the ultrasound image coordinate system I (image) is established based on the transverse and sagittal ultrasound images. For any feature point, there is a coordinate transformation: p c =T ct *T tr *T ri *p i ;

[0080] (4.2) Given the calibration phantom parameter information, use the magnetic field generator coordinates of 30 points on the right, top, and front surfaces of the cuboid phantom that have been collected to calculate the transformation matrix T from the transmitter coordinate system to the phantom reconstruction coordinate system. ct ;

[0081] (4.2.1) For 30 points on the right surface, upper surface, and front surface of the phantom, the least squares method is used to fit the plane equation A*x+B*y+C*N-1=0, and three sets of plane parameters A1, B1, C1, A2, B2, C2, A3, B3, and C3 are calculated for the three planes respectively;

[0082] (4.2.2) Using the three sets of plane parameters, calculate the coordinates of the magnetic field generator coordinate system of the upper right corner of the front face of the rectangular calibration phantom;

[0083] (4.2.3) Using the three sets of plane parameters, calculate the coordinates of the magnetic field generator coordinate system in the positive directions of the x, y, and z axes of the rectangular calibration phantom;

[0084] (4.2.4) Using the 1 point coordinate and 3 direction vectors shown in 2.2 and 2.3, a total of 4 homogeneous coordinates, the least squares method is used to fit and calculate the transformation matrix Tct from the magnetic field generator T coordinate system to the phantom reconstruction coordinate system C;

[0085] (4.3) Based on the coplanarity of the five points, filter out outliers. For each of the five points in the above picture, they should belong to the same plane. Use the five-point coordinates to fit the plane:

[0086] (4.3.1) For each calibration image, the coordinates of the magnetic field generator coordinate system of the 5 N-line hypotenuse intersection points calculated are fitted to the plane;

[0087] (4.3.2) Calculate the distances between the five points and the plane respectively;

[0088] (4.3.3) If the distance between any point and the plane is greater than a certain threshold, all data of the calibration image of this frame will be discarded and will no longer participate in the calculation;

[0089] (4.4) Use iterative least squares to fit the transformation matrix T from the image coordinate system to the sensor / receiver coordinate system ri :

[0090] (4.4.1) For all the feature points marked in the calibration data preprocessing step, their magnetic field generator coordinates are calculated from their phantom reconstruction coordinates. t = inverse(T ct )*p c ;

[0091] (4.4.2) For each magnetic field generator coordinate system in the above steps, calculate its receiver coordinate system coordinate p r = inverse(T tr )*p t , where T tr It is the conversion matrix between the magnetic navigation positioning receiver and the transmitter, which is obtained by the magnetic navigation positioning device;

[0092] (4.4.3) Choose a set of 6 parameters to construct the initial rigid body transformation matrix, including rotation of angle γ along the x-axis, rotation of angle β along the y-axis, rotation of angle α along the N-axis, and translation (x, y, z);

[0093]

[0094] (4.4.4) Modify the matrix transformation parameters α, β, γ, x, y, N, and reconstruct the transformation matrix;

[0095] (4.4.5) Calculation

[0096] (4.4.6), calculate the difference

[0097] (4.4.7), if the difference is smaller, keep the iterative modification of this step;

[0098] (4.4.8) Repeat steps (4.4.4) to (4.4.7) until the error converges.

[0099] (4.5) Using the calibration matrix T obtained in step 6 above ri , transform the existing key points and reconstruct their coordinates in the phantom coordinate system;

[0100] (4.6) For the calculated coordinates of the phantom reconstruction coordinate system, the points that differ greatly from the true values ​​are screened out as outliers;

[0101] (4.7) Repeat (4.4) to (4.6) until no outliers are eliminated, and the calibration matrix T of the cross section or sagittal plane is obtained. ri ;

[0102] (5) Averaging the dual-plane calibration matrices: Based on the prior information that the two imaging planes of the ultrasound probe are orthogonal to each other, the calibration matrices obtained for the two probes are averaged.

[0103] (5.1) For the cross section and sagittal plane, the calibration matrix is ​​in the form of a 4*4 matrix. in Its physical meanings are the direction vectors of the ultrasonic probe axis, directly above, and directly to the right in the receiver coordinate system;

[0104] (5.2), let Construct new cross-sectional and sagittal calibration matrices

[0105] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

Claims

1. A calibration method for a dual-plane ultrasonic probe calibration system, characterized in that: The following steps are involved: Step (1) calibration data acquisition: using a dual-plane ultrasonic probe calibration system to collect raw data required for calibration; Step (2) calibration data annotation: automatically or non-automatically, for a number of points formed by the intersection of the ultrasound cross section and the N-shaped line of the calibration phantom, automatically or non-automatically, according to the geometric relationship, identify and annotate their coordinates in the image; Step (3) calibration data preprocessing: according to the distance geometric relationship between a number of points formed by the intersection of the ultrasonic cross section and the N-shaped lines of the calibration phantom, the coordinates of a number of feature points formed by the intersection with the hypotenuse of each group of N lines in the phantom coordinate system are calculated; Step (4) Iterative optimization using the least squares method to remove outliers and calculate the calibration matrix: Calculate the coordinate transformation matrix from the magnetic navigation receiver coordinate system to the ultrasound image coordinate system; Step (5) averaging the dual-plane calibration matrices: based on the prior information that the two imaging planes of the ultrasound probe are orthogonal to each other, averaging the calibration matrices obtained for the two probes; The specific process of step (1) is as follows: Step (1.1) fix the calibration phantom vertically in a container so that the imaging medium immerses the phantom; Step (1.2) establishes a magnetic field generator coordinate system based on the magnetic field generator; Step (1.3) uses a needle probe to collect phantom positioning data, including the coordinates of several points on the right surface, upper surface, and front surface of the rectangular parallelepiped calibration phantom; Step (1.4) ensure that the magnetic navigation positioning receiver is fixed in the middle of the ultrasound probe; Step (1.5) collecting cross-sectional calibration data: making the ultrasound cross section intersect with the N-shaped lines of the lower half of the calibration phantom; under the premise of ensuring that the ultrasound image intersects with all the N-shaped lines and the intersection points are clearly visible, adjusting the three translational degrees of freedom and three rotational degrees of freedom of the ultrasound probe respectively, and collecting a number of images; Step (1.6) collecting sagittal calibration data: the ultrasound sagittal plane is made to intersect with the N-shaped lines of the upper part of the calibration phantom. Under the premise that the ultrasound image intersects with all the N-shaped lines and the intersection points are clearly visible, the three translation degrees of freedom and three rotation degrees of freedom of the ultrasound probe are adjusted respectively to collect a number of images; The specific process of step (2) is as follows: considering that foreground points in a single-frame image, points ABC, DEF, GHI, JKL, MNO, ADG, BEH, and CFI are collinear, and AC|DF|GI, and AG|BH|CI are collinear, then marking is performed; when marking, interference from other noise points that do not conform to the rules is excluded, and the foreground points are identified more accurately according to the collinearity and parallelism rules; The specific process of step (4) is as follows: Step (4.1) establishes the magnetic field generator coordinate system T (transmitter) based on the magnetic field transmitter, establishes the phantom reconstruction coordinate system C (reconstruction) based on the calibration phantom, establishes the receiver coordinate system R (receiver) based on the magnetic navigation positioning receiver, and establishes the image coordinate system I (image) based on the transverse and sagittal ultrasound images. For any feature point, there is a coordinate transformation: p c =T ct *T tr *T ri *pi; Step (4.2) Given the calibration phantom parameter information, use the acquired coordinates of the magnetic field generator coordinate system of 30 points on the right surface, top surface, and front surface of the cuboid phantom to calculate the transformation matrix Tct from the transmitter coordinate system to the phantom reconstruction coordinate system; Step (4.3) is based on the coplanarity of the five points to filter out outliers. For the five feature points in the calibration image, which belong to the same plane, the coordinates of the magnetic field generator coordinate system of these five points are used to fit the plane. If the calculated distance between any point and the plane is greater than a specific threshold, the calibration image data of this frame is discarded. Step (4.4) uses iterative least squares fitting to obtain the calibration matrix Tri from the image coordinate system to the sensor / receiver coordinate system; Step (4.5) uses the calibration matrix Tri obtained in step (4.4) to transform the coordinates of the existing key points and reconstruct their coordinates in the phantom coordinate system; In step (4.6), for the calculated coordinates of the phantom coordinate system, the points that differ greatly from the true values ​​are removed as outliers; Step (4.7) repeats (4.4) to (4.6) until no outliers are screened out, and the calibration matrix Tri of the cross section or sagittal plane is obtained.

2. The calibration method of the dual-plane ultrasonic probe calibration system according to claim 1, characterized in that: The specific process of step (5) is as follows: Step (5.1), for the cross section and sagittal plane, the calibration matrix is ​​in the form of a 4*4 matrix in Its physical meanings are the direction vectors of the ultrasonic probe axis, directly above, and directly to the right in the receiver coordinate system; Step (5.2), let Construct new cross-sectional and sagittal calibration matrices 3. A dual-plane ultrasound probe calibration system, characterized in that: A calibration method for implementing the dual-plane ultrasonic probe calibration system according to claim 1 or 2, wherein the calibration system comprises An orthogonal dual-plane ultrasound probe has two ultrasound imaging sections, and the two imaging sections are perpendicular to each other and can intersect with the two groups of wires of the calibration phantom respectively; A magnetic field generator, used to generate a magnetic field for electromagnetic positioning, is placed on one side of the container to ensure that its magnetic field range can cover the magnetic navigation positioning receiver and the magnetic navigation positioning probe on the ultrasonic probe; The magnetic navigation positioning receiver, whose rigid body is fixed in the middle of the dual-plane ultrasound probe, can obtain its own spatial position and posture parameters in real time in the magnetic field generated by the magnetic field generator; The magnetic navigation positioning probe can obtain the spatial position parameters of the needle tip in real time in the magnetic field generated by the magnetic field generator, which is used to locate and calibrate the phantom; The calibration phantom is a rectangular parallelepiped fixed to the bottom of the container and immersed in the ultrasonic imaging medium in the container; the calibration phantom is divided into two parts, an upper part and an lower part, each of which is interspersed with an N-shaped wire group; A container and a fixture for holding an ultrasound imaging medium; The calibration phantom is made of resin material, and is a rectangular block divided into two areas, each with a square hole. The side wall of the hole is provided with an inverted triangular through hole, through which nylon wire is inserted. The holes in the upper and lower areas of the calibration phantom are all punched in the side walls and interspersed with lines to form 5 groups of N-shaped lines. The upper and lower areas of the phantom are each formed by 5 groups of N-shaped lines interspersed with nylon lines. The upper and lower areas have the same shape and size and are perpendicular to each other.

4. The dual-plane ultrasound probe calibration system according to claim 3, characterized in that: The nylon lines are interlaced to form 5 groups of N-shaped lines, 3 of which are parallel to each other, and the other 2 groups of N-shaped lines are in the opposite direction of the hypotenuse of the aforementioned N-shaped lines.

5. The dual-plane ultrasound probe calibration system according to claim 4, characterized in that: The wires passing through the triangular through holes are made of nylon and have a diameter similar to that of the ultrasonic wave used in the ultrasonic probe.

Images