A joint registration system and method based on magnetocardiogram and electrocardiogram

Abstract

The application provides a magnetocardiogram and electrocardiogram combined registration system, comprising a magnetocardiogram acquisition module, an electrocardiogram acquisition module, an optical scanning device, a CT scanning device and a CT registration module. Before the acquisition of the magnetocardiogram (MCG) and the electrocardiogram (ECG), the MCG array panel and the ECG electrode are digitized by using the optical scanner respectively; the MCG array is aligned to the human body coordinate system according to the MCG array marker; the CT scanning is performed on the subject's torso, and the human body coordinate system is aligned to the CT coordinate system according to the CT registration module worn by the subject. The system realizes the combined registration of the MCG and the ECG, and improves the registration accuracy through the optical scanning device. The application also provides a registration method corresponding to the system, which simplifies the registration process under the premise of ensuring the registration accuracy and promotes the combined imaging research of the MCG and the ECG.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of biomedical engineering, and particularly relates to a combined registration system and method based on magnetocardiogram and electrocardiogram. BACKGROUND

[0002] Arrhythmia is one of the cardiovascular diseases with high morbidity and mortality in the world. After more than a hundred years of development, electrocardiogram is currently the most important and most widely used technology in clinical diagnosis and treatment. Conventional 12-lead electrocardiogram has low spatial and temporal resolution, so the clear diagnosis, evaluation and treatment of arrhythmia diseases in clinic often need invasive cardiac electrophysiological mapping. This technology not only increases the risk of treatment, but also has limitations in understanding the mechanism of complex arrhythmia. At present, the emerging magnetocardiogram (MCG) and electrocardiogram (ECG) three-dimensional imaging technology can measure the electromagnetic field information generated by the cardiac electrophysiological activity near the body surface, and through the combination of human structure image information, a three-dimensional visual cardiac electrophysiological activity model is reversely reconstructed, so as to provide accurate lesion point position and arrhythmia mechanism characteristic analysis.

[0003] Studies have shown that MCG is more sensitive to tangential and eddy current, ECG is more sensitive to axial current, and the information of the two is complementary, which can jointly assist clinicians in diagnosis and treatment. SUMMARY

[0004] In view of the registration requirement of MCG and ECG combined imaging, the application provides a combined registration system and method based on magnetocardiogram and electrocardiogram, which is used for multi-modal MCG and ECG combined imaging.

[0005] In order to achieve the above purpose, the application uses the following technical scheme:

[0006] A combined registration system based on magnetocardiogram and electrocardiogram, comprising a magnetocardiogram acquisition module, an electrocardiogram acquisition module, an optical scanning device, a CT registration module and a CT scanning device.

[0007] Further, the magnetocardiogram acquisition module comprises a magnetic shielding barrel, a magnetocardiogram atomic magnetometer array panel, an atomic magnetometer, a mechanical arm and a magnetocardiogram data acquisition system; the magnetocardiogram atomic magnetometer array panel is arranged with a plurality of atomic magnetometers arranged in an array; the mechanical arm is used to adjust the height and angle of the magnetocardiogram atomic magnetometer array panel and the human body; the magnetocardiogram signal detected by the atomic magnetometer is recorded by the magnetocardiogram data acquisition system.

[0008] Further, the electrocardio acquisition module comprises a plurality of flexible electrode strips integrated with electrocardio electrodes and preamplifiers, and an electrocardio data acquisition system; each of the flexible electrode strips integrated with electrocardio electrodes and preamplifiers is provided with a plurality of electrodes, the electrodes are marked with blue markers consistent with the size of the electrodes, the preamplifier shell is subjected to magnetic shielding treatment so that the internal circuit does not generate magnetic noise, the electrocardio signals collected by the electrocardio electrodes are amplified by the preamplifiers, and the electrocardio data acquisition system is used for recording.

[0009] Further, the optical scanning device comprises a structured light scanner and a host computer system. Before performing magnetocardiogram and electrocardiogram acquisition, a CT registration module is fixed on one side of the torso of the subject, an electrocardio acquisition module is arranged on the surface of the torso of the subject, the flexible electrode strips in the electrocardio acquisition module and the CT registration module are manually scanned by using the structured light scanner, and the scanning is recorded by the host computer system.

[0010] Further, the CT registration module is made of a non-magnetic light-sensitive resin material, and comprises an arc-shaped marker and a spherical marker point located above the arc-shaped marker.

[0011] Further, the CT scanning device collects the structural information of the CT registration module during CT scanning.

[0012] Further, the CT registration module is registered by manually selecting a reference point and using an iterative closest point method (ICP).

[0013] The application provides a combined registration method based on magnetocardiogram and electrocardiogram, comprising the following steps:

[0014] Step 1: the subject lies flat, the flexible electrode strips are arranged on the surface of the torso of the subject, blue markers are placed at the electrodes, the subject is connected to the electrocardio data acquisition system, and then the magnetometer array panel is adjusted to a suitable position and height relative to the torso by using a mechanical arm;

[0015] Step 2: the subject is sent into a magnetic shielding barrel, and electrocardio and magnetocardiogram signals are collected synchronously by the electrocardio acquisition module and the magnetocardiogram acquisition module;

[0016] Step 3: after the electrocardio and magnetocardiogram signals are collected, the subject is pushed out of the magnetic shielding barrel, the CT registration module is fixed at a suitable position on one side of the body according to the waist circumference of the subject, and the CT registration module is completely attached to the surface of the torso;

[0017] Step 4: the structured light scanner is started, the scanning of the magnetometer array panel is paused after the scanning is quickly completed, the mechanical arm is used to move away the panel, the structured light scanner is started again to complete the scanning of the subsequent flexible electrode strips and the CT registration module, the scanning data is processed and packaged by the host computer system, and the point cloud data file is saved.

[0018] Step 5: The subject wears the CT registration module to perform CT scanning, completes the acquisition of the torso CT image, saves it in DICOM format, and performs segmentation and three-dimensional reconstruction on the CT image;

[0019] Step 6: Extract the magnetometer array panel point cloud and the CT registration module point cloud from the point cloud data file obtained by scanning. According to the panel marker points, coarsely register the panel point cloud and the panel model by manually selecting points, and finely register them by using the ICP algorithm to obtain the conversion matrix T1. Similarly, according to the CT registration module marker points, coarsely register the CT registration module point cloud and the CT image by manually selecting points, and finely register them by using the ICP algorithm to obtain the conversion matrix T2.

[0020] Step 7: According to the color of the electrode marker, set the color threshold to extract the point cloud corresponding to the electrode position. For all point clouds that meet the color standard, use the preset electrode spacing to perform Euclidean clustering, and then use the RANSAC algorithm to fit the clustered electrode point cloud according to the electrode shape to obtain the position of the electrode in the CT coordinate system.

[0021] Step 8: Since the magnetometer array panel is a custom device, the positions and directions of the atomic magnetometers are known. By applying the conversion matrix T = T2*T1, the atomic magnetometers are aligned to the CT coordinate system.

[0022] Further, in step 6, the scanned point cloud data is downsampled, and four groups of non-coplanar points are manually selected to solve the rigid transformation matrix, so that the scanned point cloud data is close to the model point cloud position. In step 6, the ICP algorithm is used for fine registration, which includes:

[0023] Given the scanned point cloud P = {p1, p2, … p n}, the model point cloud Q = {q1, q2, … q n}, find the optimal rigid transformation such that the Euclidean distance f(R, T) between the corresponding points of the two point clouds is minimized, i.e. satisfy the following equation (1):

[0024]

[0025] Where R is the rotation matrix and t is the translation vector.

[0026] Further, in step 7, the flexible electrode strip is scanned based on the structured light scanner to obtain a point cloud of scanning points. Due to the blue markers attached to the electrodes, a color threshold method is used to extract the point cloud meeting the color standard; the number of clusters is defined according to the number of electrodes in the flexible electrode strip and the electrode spacing, the Euclidean distance between the point clouds meeting the color standard is calculated and clustering is performed, the number of points of the clustered clusters is sorted, and the clusters exceeding the number of electrodes are discarded; the RANSAC algorithm is used for fitting for each cluster according to the size of the electrode radius, and the centroid of the fitted point cloud is calculated as the electrode position.

[0027] The present application has the following advantages compared with the prior art:

[0028] (1) The present application combines magnetocardiogram, electrocardiogram and CT device, realizes the joint registration of MCG, ECG and CT by means of optical scanning device through the design of CT registration module. Compared with the reconstruction of cardiac electrophysiological activity using MCG or ECG alone, the two kinds of information of MCG and ECG are complementary, which can provide more accurate and complete cardiac electrophysiological activity mechanism, and promote the clinical research and application of cardiac function imaging.

[0029] (2) In the present application, the MCG and ECG signals of the subject are collected at the same time, and then the CT registration module is arranged for CT scanning, which simplifies the registration process under the premise of ensuring the registration accuracy, and promotes the joint imaging research of ECG and MCG. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a structural schematic diagram of a joint registration system based on magnetocardiogram and electrocardiogram of the present application;

[0031] Figure 2 is a structural schematic diagram of a CT registration module of a joint registration system based on magnetocardiogram and electrocardiogram of the present application;

[0032] Figure 3 is a flow chart of a joint registration method based on magnetocardiogram and electrocardiogram of the present application.

[0033] Figure legend: 1-magnetocardiogram acquisition module; 2-electrocardiogram acquisition module; 3-optical scanning device; 4-CT registration module. DETAILED DESCRIPTION

[0034] The present application will be described in more detail below with reference to the accompanying drawings and examples, and the specific examples described herein are only used to explain the present application, not to limit the present application.

[0035] Example 1

[0036] A joint registration system based on magnetocardiogram and electrocardiogram of the present application, as shown in Figure 1As shown, 1 is a cardiac magnetic acquisition module, 2 is a cardiac electrical acquisition module, 3 is an optical scanning device, and 4 is a CT registration module.

[0037] The CT registration module 4 is a non-magnetic photosensitive resin material, such as Figure 2 As shown, the arc-shaped marker has four groups of spherical marker points with different faces; the arc-shaped marker has an arc length of 100 mm, a width of 60 mm, and a thickness of 2 mm, and the spherical marker points have a large radius of 10 mm and a small radius of 3 mm, which are suitable for all subjects. The CT registration module 4 has a density different from that of human tissues, and the structure information of the CT registration module 4 can be collected during CT scanning, which is convenient for extraction after CT image segmentation and three-dimensional reconstruction.

[0038] The cardiac magnetic acquisition module 1 includes a magnetic shielding barrel, a cardiac magnetic atomic magnetometer array panel, an atomic magnetometer, a mechanical arm, and a cardiac magnetic data acquisition system. The cardiac magnetic atomic magnetometer array panel can insert 36 or more atomic magnetometers, arranged in a 6*6 sensor array, and the height and angle between the panel and the human body can be arbitrarily adjusted by the mechanical arm. In the magnetic shielding barrel, the cardiac magnetic signals detected by the atomic magnetometer are output to the cardiac magnetic data acquisition system outside the barrel and recorded.

[0039] The cardiac electrical acquisition module 2 includes a plurality of flexible electrode strips integrated with cardiac electrical electrodes and preamplifiers, and a cardiac electrical data acquisition system. The cardiac electrical signals collected by the cardiac electrical electrodes are amplified by the built-in preamplifiers, output to the cardiac electrical data acquisition system outside the magnetic shielding barrel through optical fiber connection, and recorded. The preamplifier shell is magnetically shielded to prevent magnetic noise from the internal circuit. The cardiac electrical electrodes are made of non-magnetic silver chloride material, and each electrode strip has 9 or more electrodes arranged in a 4*9 array, totaling 36 electrodes. Each electrode is marked with a blue marker to indicate the electrode position.

[0040] The optical scanning device 3 includes a structured light scanner and a host computer system. Before collecting the magnetocardiogram and electrocardiogram, the CT registration module is fixed on one side of the subject's torso to closely fit the torso, and the flexible electrode strip is arranged on the surface of the subject's torso. The CT registration module and the flexible electrode strip are scanned by the structured light scanner, and the point cloud data file is recorded and saved by the host computer system.

[0041] The joint registration system based on the magnetocardiogram and the electrocardiogram includes a CT scanning device, which collects the structure information of the CT registration module during CT scanning.

[0042] Example Two

[0043] As shown in Figure 3As shown, a joint registration method flow chart based on magnetocardiogram and electrocardiogram is given, and the registration is performed based on the joint registration system of magnetocardiogram and electrocardiogram, and the specific implementation includes the following steps:

[0044] Step 1: The subject lies down, a flexible electrode strip is arranged on the surface of the trunk, and a blue marker is placed at the electrode, and the subject is connected to an electrocardiogram data acquisition system. Then the magnetometer array panel is adjusted to a suitable position and height with the trunk using a mechanical arm;

[0045] Step 2: The subject is sent into the magnetic shielding barrel, and the electrocardiogram and magnetocardiogram signals are collected synchronously by the electrocardiogram acquisition module 2 and the magnetocardiogram acquisition module 1;

[0046] Step 3: After the electrocardiogram and magnetocardiogram signal collection is completed, the subject is pushed out of the magnetic shielding barrel. According to the waist circumference of the subject, the CT registration module is fixed at a suitable position on one side of the body, so that it is completely attached to the surface of the trunk;

[0047] Step 4: Start the structured light scanner, pause after quickly completing the scanning of the magnetometer array panel. Use the mechanical arm to move away the panel, and start the structured light scanner again to complete the subsequent scanning of the flexible electrode strip and the CT registration module 4. Process and package the scanning data through the host computer system, and save it as a point cloud data file;

[0048] Step 5: The subject wears the CT registration module 4 to perform CT scanning, completes the trunk CT image acquisition, saves it as a DICOM format, and performs segmentation and three-dimensional reconstruction on the CT image;

[0049] Step 6: Extract the magnetometer array panel point cloud from the point cloud data file obtained by scanning, and perform coarse registration on the panel point cloud and the panel model by manually selecting points, and perform fine registration by using the iterative closest point method ICP to obtain the conversion matrix T1. Specifically:

[0050] Step 6.1: Manually extract the magnetometer array panel point cloud and the CT registration module point cloud, respectively;

[0051] Step 6.2: Downsample the scanned array panel point cloud data;

[0052] Step 6.3: Manually select four groups of non-coplanar marker points on the scanned panel point cloud and the model panel point cloud to solve the rigid transformation matrix, so that the scanned panel point cloud is close to the model panel point cloud in position;

[0053] Step 6.4: Calculate the scanned panel point cloud P = {p1, p2, … p n} and the model panel point cloud Q = {q1, q2, … q nThe Euclidean distance f(R, T) between the corresponding points is found, and the optimal rotation matrix R and translation vector t are found to minimize it;

[0054]

[0055] Step 7: Similar to step 6, according to the marker points of the CT registration module 4, the CT registration module point cloud is coarsely registered with the CT image by manual point selection, and the ICP algorithm is used for fine registration to obtain the conversion matrix T2;

[0056] Step 8: According to the color of the electrode marker, set the color threshold to extract the point cloud corresponding to the electrode position. For all point clouds meeting the color standard, use the preset electrode spacing to perform Euclidean clustering. Then, according to the electrode shape, use the random sample consensus (RANSAC) algorithm to fit the clustered electrode point cloud. Apply the conversion matrix T1 to the fitted electrode point cloud to align the electrode position and direction to the CT coordinate system;

[0057] Step 8.1: Manually extract the flexible electrode strip point cloud;

[0058] Step 8.2: According to the blue marker attached to the electrocardiogram electrode, use color threshold to extract point clouds meeting the color standard. Color is represented using the HSV color model, which can distinguish color and intensity information to better identify color;

[0059] Step 8.3: According to the number of electrodes in the flexible electrode strip and the electrode spacing, define clusters, calculate the Euclidean distance between point clouds meeting the color standard and perform clustering, sort the number of points in the clustered clusters, and discard clusters exceeding the number of electrodes;

[0060] Step 8.4: Since the electrocardiogram electrode is circular, use the RANSAC algorithm to fit the electrode shape according to the size of the electrode radius for each cluster, and calculate the centroid of the fitted electrode as the electrode position. The specific steps are as follows:

[0061] (1) Randomly select three points from the point cloud set to calculate the center and radius of the circle;

[0062] (2) Set the threshold to the electrode radius, calculate the distance of each point to the circle, if the distance is less than the given threshold, record it as an inner point, otherwise record it as an outer point;

[0063] (3) Count the number of inner points, record the parameters of the estimated circle and the number of inner points each time;

[0064] (4) Set the number of iterations, repeat the above (1)-(3), and finally select the model with the most inner points as the fitting result.

[0065] Step 8.5: Apply the conversion matrix T1 to the fitted electrode point cloud;

[0066] Step 9: Apply the transformation matrix T = T2 x T1 to the array panel model of the magnetometer-on-a- chip, respectively, to align the atomic magnetometer positions and orientations to the CT coordinate system.

[0067] The contents not described in detail in the specification of the present application belong to the prior art known to those skilled in the art.

[0068] The above examples are provided only for describing the present application and are not intended to limit the scope of the present application. The scope of the present application is defined by the appended claims. The present application should also include modifications and variations of the present application if the modifications and variations belong to the scope of the claims of the present application and equivalents thereof.

Claims

1. A joint registration system based on magnetocardiography and electrocardiography, characterized in that, The system includes: a magnetocardiogram acquisition module, an electrocardiogram acquisition module, an optical scanning device, a CT scanning device, and a CT registration module; The optical scanning device is used to scan the magnetic field atomic magnetometer array panel in the magnetic field acquisition module, the flexible electrode strip of the electrocardiogram acquisition module, and the CT registration module worn by the subject to obtain scanning data and save it as point cloud data. The CT scanning device is used to perform CT scans on subjects wearing the CT registration module to obtain CT images; The system registers point cloud data obtained from an optical scanning device and CT images obtained from a CT scanning device; the system is configured to perform the following registration process: Point clouds of the magnetocardiogram atomic magnetometer array panel and CT registration module were extracted from the point cloud data file obtained from the scan. Based on the panel marker points, the panel point cloud and the panel model were coarsely registered by manually selecting non-coplanar marker points, and finely registered using the ICP algorithm to obtain the transformation matrix. Similarly, based on the marker points of the CT registration module, the point cloud of the CT registration module and the CT image are coarsely registered by manually selecting non-coplanar marker points, and finely registered using the ICP algorithm to obtain the transformation matrix. ; Based on the color of the electrode markers, a color threshold is set to extract the point cloud corresponding to the electrode position. For all point clouds that meet the color standard, Euclidean clustering is performed using a preset electrode spacing. Then, based on the electrode shape, the clustered electrode point cloud is fitted using the Random Sample Consensus (RANSAC) algorithm to obtain the position of the electrode in the CT coordinate system. The magnetic field atomic magnetometer array panel is a custom-made device. The position and orientation of the atomic magnetometers are known, and a transformation matrix is ​​applied. Align the atomic magnetometer to the CT coordinate system.

2. The joint registration system based on magnetocardiography and electrocardiography according to claim 1, characterized in that, The magnetocardiogram acquisition module includes a magnetic shielding barrel, a magnetocardiogram atomic magnetometer array panel, atomic magnetometers, a robotic arm, and a magnetocardiogram data acquisition system; the magnetocardiogram atomic magnetometer array panel holds multiple atomic magnetometers arranged in an array; the robotic arm is used to adjust the height and angle of the magnetocardiogram atomic magnetometer array panel relative to the human body; The magnetic field signals detected by the atomic magnetometer are recorded by the magnetic field data acquisition system.

3. The joint registration system based on magnetocardiography and electrocardiography according to claim 1, characterized in that, The ECG acquisition module includes multiple flexible electrode strips integrating ECG electrodes and preamplifiers, as well as an ECG data acquisition system. Each flexible electrode strip containing the integrated ECG electrodes and preamplifiers holds multiple electrodes, which are marked with blue markers of the same size as the electrodes. The preamplifier housing is magnetically shielded to prevent magnetic noise from being generated in its internal circuitry. The ECG signals acquired by the ECG electrodes are amplified by the preamplifier and recorded by the ECG data acquisition system.

4. The joint registration system based on magnetocardiography and electrocardiography according to claim 1, characterized in that, The optical scanning equipment includes a structured light scanner and a host computer system. Before acquiring magnetocardiogram and electrocardiogram, a CT registration module is fixed on one side of the subject's torso, and an electrocardiogram acquisition module is placed on the surface of the subject's torso. The flexible electrode strips in the CT registration module and the electrocardiogram acquisition module are manually scanned using the structured light scanner, and the data is recorded through the host computer system.

5. The joint registration system based on magnetocardiography and electrocardiography according to claim 1, characterized in that, The CT registration module is made of non-magnetic photosensitive resin material; the CT registration module includes an arc-shaped marker and a spherical marker point located above the arc-shaped marker; the CT registration module performs registration by manually selecting a reference point and using the iterative nearest neighbor method (ICP).

6. A joint registration method based on magnetocardiography and electrocardiography, applied to the joint registration system according to any one of claims 1-5, characterized in that, Includes the following steps: Step 1: The subject lies flat, flexible electrode strips are placed on the surface of the torso and blue markers are placed at the electrodes, the subject is connected to the electrocardiogram data acquisition system, and then the magnetic atom magnetometer array panel is adjusted to a suitable position and height with the torso using a robotic arm; Step 2: The subject is placed in the magnetically shielded container, and ECG and magnetocardiogram signals are collected simultaneously through the ECG acquisition module and the magnetocardiogram acquisition module; Step 3: After the ECG and magnetocardiogram signals are collected, the subject is pushed out of the magnetic shielding container. Based on the subject's waist circumference, the CT registration module is fixed at a suitable position on one side of the body to ensure that it is completely in contact with the surface of the torso. Step 4: Start the structured light scanner, quickly complete the scanning of the cardiac magnetometer array panel, pause, use the robotic arm to remove the panel, restart the structured light scanner to complete the subsequent scanning of the flexible electrode strip and CT registration module, process and encapsulate the scan data through the host computer system, and save it as a point cloud data file; Step 5: The subject wears a CT registration module to undergo a CT scan, completes the acquisition of trunk CT images, saves them in DICOM format, and performs segmentation and 3D reconstruction of the CT images; Step 6: Extract the point cloud of the magnetocardiogram atomic magnetometer array panel and the point cloud of the CT registration module from the point cloud data file obtained from the scan. Based on the panel marker points, perform coarse registration between the panel point cloud and the panel model by manually selecting points, and then perform fine registration using the ICP algorithm to obtain the transformation matrix. Similarly, based on the marker points of the CT registration module, the point cloud of the CT registration module and the CT image are coarsely registered by manually selecting points, and finely registered using the ICP algorithm to obtain the transformation matrix. ; Step 7: Based on the color of the electrode markers, set a color threshold to extract the point cloud corresponding to the electrode position. For all point clouds that meet the color standard, perform Euclidean clustering using the preset electrode spacing. Then, based on the electrode shape, use the Random Sample Consensus (RANSAC) algorithm to fit the clustered electrode point cloud to obtain the position of the electrode in the CT coordinate system. Step 8: The magnetic field atomic magnetometer array panel is a custom-made device. The position and orientation of the atomic magnetometers are known. A transformation matrix is ​​applied... Align the atomic magnetometer to the CT coordinate system.

7. The joint registration method based on magnetocardiography and electrocardiography according to claim 6, characterized in that, Step 6 includes downsampling the scanned point cloud data, manually selecting four sets of non-coplanar points to solve for the rigid transformation matrix, so that the scanned point cloud data is close to the model point cloud position; the fine registration using the ICP algorithm specifically includes: The point cloud obtained from the known scan Model point cloud Find the optimal rigid body transformation that makes the point cloud and Euclidean distance between corresponding points The minimum is that which satisfies the following equation (1): in, Let be a rotation matrix. It is a translation vector.

8. The joint registration method based on magnetocardiography and electrocardiography according to claim 6, characterized in that, In step 7, the flexible electrode strip is scanned using a structured light scanner to obtain a scanned point cloud. Based on the blue markers attached to the electrodes, a color threshold is used to extract point clouds that meet the color standard. Clusters are defined based on the number of electrodes and the spacing between electrodes in the flexible electrode strip. The Euclidean distance between point clouds that meet the color standard is calculated and clustered. The number of points in the clustered clusters is sorted, and clusters exceeding the number of electrodes are discarded. The RANSAC algorithm is used to fit each cluster according to the size of the electrode radius, and the centroid of the fitted point cloud is calculated as the electrode position.

Images