Transcranial magnetic stimulation hand hotspot automatic search system based on optical navigation

Abstract

The application discloses a transcranial magnetic stimulation hand hotspot automatic search system based on optical navigation, which comprises a brain region positioning module, a grid point array setting module, an optical navigation module and a hotspot judgment module.

Description

Technical Field

[0001] This invention belongs to the field of transcranial magnetic stimulation technology, specifically relating to an automatic search system for transcranial magnetic stimulation hand hotspots based on optical navigation. Background Technology

[0002] Transcranial magnetic stimulation (TMS) is a painless, non-invasive method of brain stimulation that has been widely used in clinical and research fields in recent years. The basic principle of TMS is to induce a time-varying induced electric field in the cerebral cortex through a time-varying magnetic field applied extracranially, thereby generating an induced current in the brain tissue. When the induced current exceeds the excitation threshold of the nerve tissue, an effect similar to direct electrical stimulation is produced, effectively stimulating the corresponding brain tissue. During TMS, the transcranial magnetic stimulator is connected to a stimulation coil, which is placed on the area of ​​the subject's head that needs stimulation. Based on the principle of electromagnetic induction, the pulsed magnetic field generated by the stimulation coil induces a current in the cerebral cortex, thereby stimulating the cortical nerves and producing a series of physiological and biochemical reactions.

[0003] During TMS administration, the coil needs to be positioned at a specific point on the patient's scalp; this point is called the target point. A hot spot is one type of target point. The criteria for determining whether a target point is a hot spot is that when TMS is applied to this point, the corresponding muscle will produce a positive MEP (motor evoked potential, which refers to the muscle motor complex potential recorded by stimulating the motor cortex in the contralateral target muscle).

[0004] The so-called TMS automatic hotspot search for hand movements involves using a robotic arm to automatically control a stimulation coil to reach the stimulation target point, which is located on the scalp corresponding to the area in the cerebral cortex that controls hand movements. TMS is applied to the stimulation target point, and electrodes are attached to the muscles of the contralateral hand to detect the motor epithelial projection (MEP) and determine if a positive result is obtained. When locating the area in the cerebral cortex that controls hand movements, the primary motor cortex must first be located. The accuracy of the primary motor cortex location is crucial for finding hotspots, and its precision directly affects the accuracy of hotspot location.

[0005] Existing technologies offer several methods for locating the primary motor cortex of the brain. For example, Patent Document 1 (CN113769275A) proposes physically locating brain functional areas by wearing a transcranial magnetic stimulation (TMS) helmet. The TMS helmet is combined with a TMS stimulation coil, and the location of brain functional areas is achieved through fixed marker points. However, this method is inconvenient to use, and some patients are not physically able to wear the helmet, making this approach unsuitable. Furthermore, this method does not provide a visually intuitive display of brain region location.

[0006] For example, Patent Document 2 (CN114305730A) proposes a brain region localization method to register a brain grid atlas onto a patient's medical imaging data, where the medical imaging data refers to head MRI data. This allows for the identification of the primary motor cortex, and the division of an N×N localization target matrix within the primary motor cortex for applying TMS stimulation. However, the localization accuracy of this method depends on the registration algorithm and the segmentation precision of the brain grid atlas. Even with high-precision algorithms, the overall localization accuracy remains low, and the method requires the patient to have MRI image data, making it a demanding condition.

[0007] For example, Patent Document 3 (CN115035124A) proposes a method for creating the DLPFC target region using image processing algorithms and structural images, and calculating the position of the primary motion region in real time. However, this method takes too long to output the target point, has poor real-time performance, and low accuracy.

[0008] In summary, most existing solutions require the support of individual patient MRI data, while solutions that do not use individual patient MRI data either have low positioning accuracy or are limited in application conditions and have poor applicability. Summary of the Invention

[0009] To address the aforementioned problems in the existing technology, this invention provides an automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation. The technical problem to be solved by this invention is achieved through the following technical solution:

[0010] An automatic hotspot search system for transcranial magnetic stimulation hand based on optical navigation, comprising:

[0011] The brain region localization module is used to locate brain regions based on point cloud registration methods in the absence of MRI data of the individual to be tested, so as to obtain the primary motor area of ​​the cerebral cortex of the individual to be tested.

[0012] The grid matrix setting module is used to divide the primary motion area into grids to obtain a grid matrix;

[0013] An optical navigation module is used to locate the grid matrix to determine the position of each stimulation point on the head of the individual to be tested, and to apply stimulation to each stimulation point in sequence.

[0014] The hotspot detection module is used to collect and analyze the physiological signals generated after stimulation of each stimulation site in order to determine the hotspots of hand movements.

[0015] The beneficial effects of this invention are:

[0016] 1. This invention proposes an automatic hand hotspot search system based on optical navigation for transcranial magnetic stimulation (TMS) when individual patient MRI data is unavailable. This system utilizes point cloud registration to locate the patient's primary motor cortex, thus achieving more accurate automatic hand hotspot search. This system does not rely on individual patient MRI data, has wide applicability, is easy to use, and offers higher positioning accuracy.

[0017] 2. The transcranial magnetic stimulation hand hotspot automatic search system based on optical navigation provided by this invention can also combine the 10-20 system with the point cloud registration method to realize the localization of the primary motor area, which can more intuitively reflect the localization process of the motor area and facilitate operation.

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] Figure 1 This is a structural block diagram of an automatic search system for hand hotspots based on optical navigation for transcranial magnetic stimulation provided in an embodiment of the present invention;

[0020] Figure 2 This is a structural block diagram of a brain region localization module provided in an embodiment of the present invention;

[0021] Figure 3 This is a flowchart illustrating the workflow of a brain region localization module provided in an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the positioning bracket being worn according to an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of the head support provided in the embodiment of the present invention in the camera coordinate system;

[0024] Figure 6 This is a schematic diagram of the head tracking coordinate system provided in an embodiment of the present invention;

[0025] Figure 7 This is a schematic diagram of the positioning probe provided in an embodiment of the present invention;

[0026] Figure 8 This is a schematic diagram of the probe coordinate system provided in an embodiment of the present invention;

[0027] Figure 9 This is a structural block diagram of a grid dot matrix setting module provided in an embodiment of the present invention;

[0028] Figure 10 This is a structural block diagram of another brain region localization module provided in an embodiment of the present invention;

[0029] Figure 11This is a virtual coordinate system and a 10-20 point display diagram provided in the embodiments of the present invention;

[0030] Figure 12 This is a structural block diagram of another grid dot matrix setting module provided in an embodiment of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Example 1

[0033] Please see Figure 1 , Figure 1 This is a structural block diagram of an automatic hand hotspot search system based on optical navigation provided in an embodiment of the present invention, which includes:

[0034] The brain region localization module is used to locate brain regions based on point cloud registration methods in the absence of MRI data of the individual to be tested, so as to obtain the primary motor area of ​​the cerebral cortex of the individual to be tested.

[0035] The grid matrix setting module is used to divide the primary motion area into grids to obtain a grid matrix;

[0036] The optical navigation module is used to locate the grid matrix to determine the position of each stimulation point on the head of the individual being tested, and to apply stimulation to each stimulation point in sequence.

[0037] The hotspot detection module is used to collect and analyze the physiological signals generated after stimulation of each stimulation site in order to determine the hotspots of hand movements.

[0038] This embodiment proposes an automatic hand hotspot search system based on optical navigation for transcranial magnetic stimulation (TMS) when individual patient MRI data is unavailable. The system first locates the primary motor cortex using point cloud registration, then performs gridding over the primary motor cortex, and finally uses optical navigation to locate the gridded target points within the primary motor cortex. Stimulation is then applied to these gridded target points to determine the location of the hand's motor hotspots. This system does not rely on individual patient MRI data, has wide applicability, is easy to use, and offers higher positioning accuracy.

[0039] The following provides a detailed explanation of each module.

[0040] For the brain region localization module, as an optional implementation method, please refer to [link / reference needed]. Figure 2 and Figure 3 , Figure 2 This is a structural block diagram of a brain region localization module provided in an embodiment of the present invention. Figure 3 This is a flowchart illustrating the operation of a brain region localization module provided in this embodiment of the invention. The brain region localization module provided in this embodiment mainly includes:

[0041] The first three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on standard NMR data to obtain a standard head model and establish a virtual coordinate system for the standard head model;

[0042] The point cloud registration unit is used to track the head of the individual under test using a camera and extract several key points of the head in conjunction with a positioning probe. At the same time, by establishing the connection between the head tracking coordinate system, the camera coordinate system, the probe coordinate system and the virtual coordinate system, the individual space is registered with the standard head model. Among them, the head tracking coordinate system is the coordinate system corresponding to the actual head of the individual under test, the camera coordinate system is the coordinate system corresponding to the camera, and the probe coordinate system is the coordinate system corresponding to the positioning probe.

[0043] The segmentation unit is used to segment the standard brain atlas (AAL) to obtain different brain regions in the virtual coordinate system, and to register the brain regions to the individual space according to the connection between the head tracking coordinate system and the virtual coordinate system; at the same time, it transforms the standard head model to the individual space according to the connection between the head tracking coordinate system and the virtual coordinate system.

[0044] The first mapping unit is used to perform three-dimensional visualization and cortical mapping of standard head models and brain regions in individual space to obtain the primary motor area.

[0045] Specifically, for the first 3D reconstruction unit, the process of 3D reconstruction using standard NMR data is as follows:

[0046] First, the standard MRI data was segmented to obtain gray matter, cerebrospinal fluid, white matter, skull, and scalp. Among these,

[0047] The segmentation of gray matter, cerebrospinal fluid, and white matter can be achieved using existing region growing algorithms. This involves selecting a starting seed point in the region of interest or target region of the image, then searching for points with similar pixels within the neighborhood of the seed point, and finally connecting the searched points to form the final target region. Region growing algorithms include isolated connectivity, confidence connectivity, and connectivity thresholding.

[0048] The scalp and skull can be segmented using an isosurface extraction algorithm. By setting an extraction value, the input data values ​​are extracted, and the portion of the data that equals the extraction value is extracted to obtain isosurface data, which is the corresponding image of the scalp and skull.

[0049] Then, 3D reconstruction is performed to obtain 3D reconstructed images of the brain. Since the surface of the reconstructed image is not smooth, the Laplacian smoothing algorithm is used for mesh smoothing. After multiple iterations, a smoothed image can be obtained. At this point, the 3D reconstruction of the standard head model is complete.

[0050] It should be noted that after completing the segmentation and 3D reconstruction of the standard head model, it is also necessary to complete the 3D visualization of the standard head model and create a virtual coordinate system, and refer to the patient's actual head as the individual space.

[0051] Furthermore, the main task of the point cloud registration unit is to establish a connection between the patient's actual head and the standard head model. This connection is between coordinate systems; the coordinate system established in the individual space is the head tracking coordinate system. In other words, this process is the process of establishing a connection between the head tracking coordinate system and the virtual coordinate system. This process is described in detail below:

[0052] 1. Use a camera to track an infrared reflective ball mounted on the head support of the individual under test, extract the three-dimensional coordinates of the infrared reflective ball in the camera coordinate system, and calculate the transformation matrix B from the head tracking coordinate system to the camera coordinate system.

[0053] First, this process requires the patient to wear a tracking and positioning brace, such as Figure 4 As shown, the positioning bracket is divided into four branches: left branch, right branch, upper branch, and lower branch. There are three infrared reflective positioning balls on the positioning bracket, located in the left, right, and upper branches of the bracket, arranged in a triangular pattern.

[0054] Then, a binocular camera is used to track the infrared reflective ball mounted on the head support of the individual under test, so as to realize the target tracking of the individual's head and extract the three-dimensional coordinates of the infrared reflective ball in the camera coordinate system.

[0055] Understandably, the infrared light source is placed at the camera, and an infrared filter is added to the camera to filter out interference from visible light information. Then, each acquired frame needs to be processed as follows to achieve the target tracking process.

[0056] (a) Proportional compression to improve processing speed;

[0057] (b) The image is subjected to Gaussian filtering to remove noise, the image is binarized, the image is opened and closed, and small contours are removed.

[0058] (c) Use a contour extraction algorithm to extract all contours and redraw the results on a black layer;

[0059] (d) Use the connected component analysis algorithm to analyze all connected component information on the redrawn image, obtain information such as the center point and area of ​​all connected components, exclude the contours of connected components with excessively large areas, calculate the roundness of connected components, and exclude non-circular connected components.

[0060] (e) Sort by the area of ​​the connected components, take out the three largest ones, and you can get the coordinates of the ball to be positioned on the camera imaging plane. Then enlarge the obtained coordinates according to the compression ratio.

[0061] The obtained planar coordinates of the head support are used for 3D reconstruction to obtain its 3D coordinate values ​​P1(Xa,Ya,Za), P2(Xb,Yb,Zb), and P3(Xc,Yc,Zc) in the camera coordinate system, as shown below. Figure 5 As shown, the camera coordinate system is set at the optical center of the left camera of the binocular camera. The Z-axis coincides with the optical axis and points to the observed object. The Y-axis is vertically upward, and the X-axis is perpendicular to both the Z-axis and the Y-axis, pointing towards the right camera.

[0062] Understandably, before performing the above-mentioned 3D reconstruction process, it is necessary to obtain the camera's intrinsic and extrinsic parameter matrices. The method to obtain these matrices is to perform camera calibration. Camera calibration is affected by a variety of factors, including the distance between the calibration board and the camera, the accuracy of the calibration board itself, the camera's exposure, the calibration board's pose, and the camera's focal length.

[0063] The above process completes the extraction of the three-dimensional coordinates of the infrared reflective ball in the camera system. Next, it is necessary to associate the coordinates in the virtual coordinate system constructed by the first 3D reconstruction unit with the coordinates in the camera coordinate system. Since the two are not directly related, this embodiment uses a head-mounted positioning bracket to construct a head-tracking coordinate system as a bridge between the two. Furthermore, addressing the problems of long computation time and poor accuracy in existing image processing methods, a method with shorter computation time and higher accuracy is proposed. This method mainly consists of the association process from the head-tracking coordinate system to the camera system and the association process from the virtual system to the head-tracking coordinate system.

[0064] Specifically, the spatial distances D1, D2, and D3 between the reflective spheres are first calculated using the spatial distance formula, as follows:

[0065]

[0066]

[0067]

[0068] Sort the values ​​of D1, D2, and D3 in ascending order, and it is obvious that the maximum value d can be obtained. max =D3, minimum value d min1 =D1 and the intermediate value d mid=D2, use the sorted results to label the three reflective balls, ensuring the labels are always P1(Xa,Ya,Za), P2(Xb,Yb,Zb), P3(Xc,Yc,Zc). Then, a head-tracking coordinate system needs to be constructed, with the origin at the center of the support frame. tl Indicates, such as Figure 6 As shown.

[0069] Since the head tracking coordinate system is constructed using the camera, there is a natural mathematical relationship between the head tracking coordinate system and the camera coordinate system. The transformation matrix B from the head tracking coordinate system to the camera coordinate system can be calculated, and the structure of matrix B is as follows:

[0070]

[0071] Where T1 is the translation matrix for transforming the head tracking system to the camera system, which is a 3x1 matrix, and R1 is the rotation matrix for transforming the head tracking system to the camera system, which is a 3x3 matrix.

[0072] like Figure 6 As shown, let θ3 be the angle formed from P1 through P2 to P3.

[0073] T1=(Xb+d min1 *cos∠θ3,Yb+d min1 *cos∠θ3,Zb+d min1 *cos∠θ3) T

[0074] Construct the unit vector α for each axis of the head tracking coordinate system. x α y α z :

[0075]

[0076]

[0077] α z =α x ×α y

[0078] In the above formula, the × sign represents the cross product of vectors, the "||" symbol represents the magnitude of the vectors, P1, P2, and P3 refer to the coordinate values, and α... x α y α z The above calculations result in row vectors, so R1 = [α] y T ,α x T ,α z T ].

[0079] From this, we can derive the transformation matrix B from the head-tracking system to the camera system.

[0080] 2. Use a camera to track and locate the reflective ball on the probe and construct the probe coordinate system accordingly. Calculate the transformation matrix D from the probe coordinate system to the camera coordinate system.

[0081] Please see Figure 7 , Figure 7 This is a schematic diagram of the positioning probe provided in an embodiment of the present invention. Since the positioning probe can indicate several key points on the head of the individual being tested, such as the root of the nose, and the left and right ear roots, the coordinates of these key points in the head tracking coordinate system are the same as the probe coordinates. The acquisition process is as follows:

[0082] (a) Connect the site acquisition triggering device to the system. The device consists of an acquisition button and a signal processor. The acquisition button has a self-reset function, and the signal processor is a device controlled by a microcontroller that has functions such as AD signal acquisition and conversion, button trigger judgment, and serial communication.

[0083] (b) Hold the positioning probe and point it at the patient's nasal root position. After confirming that the positioning probe marker is completely captured by the camera, press the trigger button to complete the nasal root position acquisition. Then the program will calculate the coordinates of the nasal root position in the head tracking coordinate system in real time based on the positional relationship of the marker on the positioning probe.

[0084] (c) Then use the positioning probe to point to the right ear root and left ear root positions in turn, and repeat the above (b) process to obtain the coordinates of the three points under the head tracking system.

[0085] Then, the camera is used to track all the reflective spheres on the probe, and the coordinates of the top three spheres in the camera coordinate system are obtained as P4(Xd,Yd,Zd), P5(Xe,Ye,Ze) and P6(Xf,Yf,Zf), which are used to construct the probe coordinate system.

[0086] Calculate the spatial distances D4, D5, and D6 between the three balls:

[0087]

[0088]

[0089]

[0090] Since the physical structure of the positioning probe is basically the same as that of the positioning bracket worn on the head, the method for constructing the probe coordinate system is similar to that for the head tracking coordinate system, such as... Figure 8 As shown.

[0091] The transformation matrix D from the probe coordinate system to the camera coordinate system is expressed as:

[0092]

[0093] Where T2 is the translation matrix for transforming the probe coordinate system to the camera coordinate system, which is a 3x1 matrix, and R2 is the rotation matrix for transforming the probe coordinate system to the camera coordinate system, which is a 3x3 matrix.

[0094] T2=(Xe+d min2 *cos∠θ2,Ye+d min2 *cos∠θ2,Ze+d min2 *cos∠θ2) T

[0095] d min2 Let D4 be the minimum distance between the three reflective spheres P4, P5, and P6 on the positioning probe. Take a unit vector β for each coordinate axis. x β y β z P4, P5, and P6 refer to coordinate values.

[0096]

[0097]

[0098] β z =β x ×β y

[0099] Then we have R2 = [β] y T ,β x T ,β z T ].

[0100] 3. Based on transformation matrix B and transformation matrix D, obtain the coordinate values ​​of several key points on the head of the individual under test in the head tracking coordinate system.

[0101] Obtain the coordinates (Xp, Yp, Zp) of the probe tip in the probe coordinate system, where Xp = (d1 + d2) * cos(180 - θ1), Yp = 0, and Zp = -(d2 * sin(180 - θ1)).

[0102] The matrix probe_point is constructed as follows:

[0103] probe_point = [Xp, Yp, Zp, 1] T .

[0104] Combining transformation matrix B and transformation matrix D, the coordinates (Xh, Yh, Zh) of the probe tip in the head tracking coordinate system can be obtained as follows:

[0105] (Xh,Yh,Zh,1) T =B -1 *D*probe_point

[0106] Since the probe can indicate several key points on the head of the individual being tested, such as the root of the nose, the left and right ear roots, the coordinates of the key points in the head tracking coordinate system are the same as the coordinates of the probe tip in the head tracking coordinate system.

[0107] 4. Based on the coordinates of several key points on the head of the individual to be tested in the head tracking coordinate system and the coordinates of these key points in the virtual coordinate system, the point cloud registration function is used to obtain the transformation matrix A from the head tracking coordinate system to the virtual coordinate system.

[0108] First, obtain the 3D coordinates of the three key points—left ear root, nose root, and right ear root—in the virtual coordinate system from the first step. Since the standard head model constructed in the first step has already undergone point cloud processing, the coordinates can be directly output here. Then, use the standard point cloud registration function to connect the coordinates of the three points—left ear root, nose root, and right ear root—in the head tracking coordinate system with their coordinates in the virtual coordinate system, obtaining the transformation matrix between the two coordinate systems. The transformation matrix from the head tracking coordinate system to the virtual coordinate system is called A.

[0109] 5. Invert the transformation matrix A to obtain the transformation matrix from the standard head model to the actual head model, so as to achieve the registration between the individual space and the standard head model.

[0110] Since what is needed in practice is the coordinates from the virtual coordinate system to the head tracking coordinate system, that is, the transformation matrix from the standard head model to the actual head model, it is necessary to take the inverse matrix of A.

[0111] Thus, we have obtained the transformation matrix A from the head tracking coordinate system to the virtual coordinate system and the transformation matrix B from the head tracking coordinate system to the camera coordinate system. The relationship between the virtual coordinate system and the camera coordinate system is established by these two matrices.

[0112] Furthermore, the segmentation unit uses a threshold segmentation method to segment different brain regions from the standard brain atlas AAL. The resulting brain region images are in a virtual coordinate system. Then, the matrix A obtained in the previous step and the threshold-segmented brain regions need to be registered to the actual patient's head, which is called the brain region in individual space.

[0113] In addition, the standard header template needs to be transformed using matrix A in the same way to convert it to the individual space.

[0114] Finally, the first mapping unit performs three-dimensional visualization of the standard head model image and brain regions from the previous step. The brain region images are used to locate the cortical regions of interest through cortical mapping, ultimately determining the primary motor area.

[0115] The brain region localization module provided in this embodiment uses point cloud registration to locate the primary motor area of ​​a patient's head without the patient's individual MRI data. Compared with existing physical localization methods such as wearing a transcranial magnetic resonance imaging (TMI) helmet, it is more convenient and has higher applicability.

[0116] For the grid dot matrix setting module, as one optional implementation method, please refer to [link / reference]. Figure 9 , Figure 9 This is a structural block diagram of a grid dot matrix setting module provided in an embodiment of the present invention, which includes:

[0117] The first dividing unit is used to construct a square grid matrix on the standard head model. Specifically, a point with a curve distance S from the midpoint of the curve is selected on one side of the curve as the unique intersection point, and a plane intersecting with the scalp layer is determined. With the selected point as the center, an N*N grid covering the primary motor area is constructed radiating outwards in all directions at a determined interval on the plane, and mapped onto the standard head model. The curve is the curve formed by the intersection of the plane formed by the center point of the primary motor area and two key points of the individual's head with the scalp layer.

[0118] The first transformation unit is used to transform the mesh matrix on the standard head model to the individual space according to the transformation matrix A.

[0119] Specifically, the process of constructing an N*N mesh matrix on the standard head model for the first partitioning unit is as follows:

[0120] In point cloud registration methods, the center position of the primary motor cortex and the positions of the left and right ear roots can be calculated. These three points define a plane, which must intersect the scalp with a curve. A point is selected at a distance S from the curve to the right or left of the center point of the primary motor cortex. Construction method: A plane intersecting the scalp is determined using the selected point as the unique intersection point. Then, using this point as the center, an N*N grid of points is constructed radially in all directions at a defined interval on the plane. Here, the value of N is sufficient to cover the hand movement area. Since the points in the grid are relatively close to the scalp, a minimum distance search is used to map them onto the patient's individual head model.

[0121] The first transformation unit transforms the mesh matrix on the standard head model to the individual space as follows:

[0122] Let the coordinates of any point on the grid be (X1, Y1, Z1). These coordinates are in the virtual coordinate system of the standard template, and we can construct the matrix target_point:

[0123] target_point = (X1, Y1, Z1, 1) T

[0124] Take the coordinates (X2, Y2, Z2) of any point on the primary motion cortex mesh in the head tracking coordinate system, and construct the matrix head_point based on its coordinates.

[0125] head_point = (X2, Y2, Z2, 1) T

[0126] The superscript T indicates the matrix transpose, meaning head_point is a 4x1 column vector.

[0127] At this point, the relationship between the two matrices, head_point and target_point, can be represented:

[0128] head_point = A -1 *target_point

[0129] The superscript -1 indicates that matrix A is inverted. A is obtained when locating brain regions without MRI data. The above process is repeated until all points are transformed.

[0130] After dividing the grid into points, the stimulation coils need to be precisely positioned on each grid point. This process is completed by the optical navigation module, which includes:

[0131] First, the grid points of the primary motion area in the camera coordinate system are transformed to the robot arm coordinate system.

[0132] Specifically, the coordinates (x3, y3, z3) of any point on the primary motion cortex mesh in the camera coordinate system are taken, and the matrix cam_point is constructed.

[0133] cam_point = (x3, y3, z3, 1) T

[0134] After point cloud registration, the camera system and virtual system points are linked together, resulting in the following relationship matrix:

[0135] cam_point=B*head_point=B*A -1 *target_point

[0136] Next, the points transformed to the camera coordinate system need to be further transformed to the robotic arm coordinate system. The origin of the robotic arm coordinate system is set at the center of the robotic arm base, with the Z-axis pointing vertically upwards, conforming to the left-hand rule. The transformation relationship between the robotic arm coordinate system and the camera coordinate system needs to be determined by hand-eye calibration. The accuracy of the hand-eye calibration directly affects the transformation accuracy. The hand-eye calibration process ultimately yields the transformation matrix C from the camera coordinate system to the robotic arm coordinate system. C is a 4x4 matrix.

[0137] Take the coordinates of any point on the primary motion cortex in the robot arm coordinate system as (x4, y4, z4), and construct the matrix robot_point;

[0138] robot_point = (x4, y4, z4, 1) T

[0139] The following relationship holds:

[0140] robot_point=C*cam_point=C*B*head_point=C*B*A -1 *target_point

[0141] Through the above process, the coordinates of any point on the primary kinematic cortex in the robot arm coordinate system can be obtained.

[0142] Finally, based on the robotic arm coordinate system, the program controls the robotic arm to drive the stimulation coil to the designated point, and that point can then be stimulated.

[0143] The hotspot detection module is specifically used for:

[0144] Collect composite points at the corresponding hand muscles, and when a MEP positive result is obtained, identify the corresponding grid points as hotspots and record them;

[0145] Stimulation of all grid points on the primary motor cortex and MEP positivity are achieved through a specific path, and all hotspots are aggregated to obtain hotspot regions.

[0146] Specifically, when identifying hand movement hotspots, if all grid points are located on the right side of the brain, the electrode potentials of the left hand should be recorded; conversely, if all grid points are located on the left side of the brain, the electrode potentials of the right hand should be recorded. Electrode potential recording requires the use of electrodes attached to the hand muscles. The collected data is imported into an analysis device for analysis, and the results are displayed as MEP waveforms on a computer. The robotic arm controls the stimulation coil to move to a point on the grid. Once in position, the TMS device stimulation intensity is set to 50% MSO for men and 45% MSO for women. MSO is the maximum intensity of the TMS stimulation device's output. The TMS device is then triggered to apply stimulation, and the composite potentials of the hand muscles are recorded using electrodes attached to the contralateral hand muscles. The result is then determined whether a positive MEP is generated. If the TMS induces an MEP amplitude ≥50μV at or below the MSI (where MSI is the device's output intensity), it is defined as a positive MEP, and the corresponding grid point is identified as a hotspot and recorded. Stimulation of all grid points on the primary motor cortex and MEP positivity are achieved through a specific path, and the area where all hotspots converge is the hotspot region.

[0147] This completes the automatic search for hotspots in the hand caused by transcranial magnetic stimulation.

[0148] This invention proposes an automatic hand hotspot search system based on optical navigation for transcranial magnetic stimulation (TMS) when individual patient MRI data is unavailable. This system utilizes point cloud registration to locate the patient's primary motor cortex, thus achieving more accurate automatic hotspot search. This system does not rely on individual patient MRI data, has wide applicability, is easy to use, and offers higher positioning accuracy.

[0149] Example 2

[0150] Based on the above embodiment one, this embodiment also provides an automatic search system for transcranial magnetic stimulation hand hotspots based on optical navigation, which includes:

[0151] The brain region localization module is used to locate brain regions based on point cloud registration methods in the absence of MRI data of the individual to be tested, so as to obtain the primary motor area of ​​the cerebral cortex of the individual to be tested.

[0152] The grid matrix setting module is used to divide the primary motion area into grids to obtain a grid matrix;

[0153] The optical navigation module is used to locate the grid matrix to determine the position of each stimulation point on the head of the individual being tested, and to apply stimulation to each stimulation point in sequence.

[0154] The hotspot detection module is used to collect and analyze the physiological signals generated after stimulation of each stimulation site in order to determine the hotspots of hand movements.

[0155] Alternatively, as one implementation method, such as Figure 10 As shown, the brain region localization module in this embodiment includes:

[0156] The second 3D reconstruction unit is used to perform 3D reconstruction on standard NMR data to obtain a standard head model and establish a virtual coordinate system for the standard head model;

[0157] The second mapping unit is used to map the 10-20 system onto the standard head model to determine the positions of points C3 and C4, thereby obtaining the primary motor areas of the cerebral cortex of the individual being tested; where point C3 corresponds to the left primary motor area and point C4 corresponds to the right primary motor area.

[0158] In this embodiment, for cases where there is no MRI data of the individual to be tested, a 10-20 system is introduced and a point cloud registration method is used to locate the primary motion zone.

[0159] Specifically, firstly, the second 3D reconstruction unit needs to establish a standard head model and a virtual coordinate system, which can be referred to in the above embodiment one.

[0160] Then, the second mapping unit needs to map the 10-20 system onto the visualized standard header model. The mapping steps are as follows:

[0161] The primary motor cortex (M1) is located in the precentral gyrus, anterior to the central sulcus, and occupies most of the precentral gyrus. According to the 10-20 system distribution, electrode sites beginning with "C" indicate the central area, which is related to movement. Comparing these sites with general brain region distribution maps, it can be seen that they are contained within the primary motor cortex. Therefore, locating these sites can indicate the position of the motor cortex. Specifically, C3 is generally considered to be the left M1 area, and C4 is considered to be the right M1 area.

[0162] First, a 3D reconstruction needs to be performed on the imported standard header MRI data, and as follows: Figure 11 As shown, a virtual coordinate system is constructed at a fixed position in the upper left of the reconstructed head model. Then, the coordinates of four points—the root of the nose (Nz), the root of the left ear (AL), the root of the right ear (AR), and the tip of the occipital protuberance (Iz)—in this virtual system need to be determined. These four points serve as reference points for calculating the length of the head surface. The reconstructed head model is actually composed of point clouds; all point clouds with Z values ​​greater than the reference points are excluded.

[0163] First, we need to determine the location of point Cz. We select a point whose straight-line distance to the four reference points is the same: distance a to Nz, distance b to Iz, distance c to AL, and distance d to AR. There will definitely be many points that satisfy the conditions. Here, we select the one that minimizes the value of (|ab|+|cd|), where || represents the absolute value on the left side of the equation. We initially designate this point as the first point Cz.

[0164] We define a plane using the first Cz point and the three points Nz and Iz. This plane intersects with a number of points in the point cloud of the head model, and we can roughly draw a curve that encompasses all these points. This curve is the scalp curve (hereinafter referred to as the curve) that we need. Here, we set the second Cz point and define it at the center of the curve.

[0165] Next, using the determined second Cz point and points AR and AL, a plane is defined. A curve intersecting the scalp layer can be drawn on this plane, and the third Cz point is placed at its center. Then, based on the X value of the third Cz and the Y and Z values ​​of the second Cz, a fourth Cz is defined. This fourth Cz will be offset from the scalp layer. A minimum distance search algorithm is used to transfer it onto the scalp layer; this is the final determined position of the Cz value.

[0166] Based on the above method, a curve can be defined using Cz, Nz, and Iz to locate the points Fpz, Fz, Pz, and Oz, respectively, at 10%, 30%, 70%, and 90% of the curve starting from Nz. Similarly, a curve can be defined using Cz, AR, and AL, and the points T3, C3, C4, and T4 can be set at 10%, 30%, 70%, and 90% of the curve starting from AL.

[0167] Then, using the first half of the curve determined by T3, Fpz, and T4, starting from T3, set F7, Fp1, Fp2, and F8 at 20%, 40%, 60%, and 80% of the total length, respectively.

[0168] Then, using the latter half of the curve determined by T3, Oz, and T4, starting from T3, set T5, O1, O2, and T6 points respectively at 20%, 40%, 60%, and 80% of the total length.

[0169] After mapping, the 10-20 system points were quantized, and the positions of the points we needed were obtained. At this point, the left and right primary motion zones can be determined by locating points C3 and C4. The virtual coordinate system and the 10-20 points are displayed as follows. Figure 11 As shown.

[0170] Next, point cloud sampling and registration operations need to be performed on the standard head model and the actual patient's head. Refer to the point cloud selection and registration process in Example 1 above to obtain the transformation matrix from the standard head model to the actual patient's head. All points 10-20 and the standard head model can be transformed to the actual patient's head, thereby determining the position of the primary motion zone of the actual patient's head, that is, the position of points C3 and C4.

[0171] Furthermore, such as Figure 12 As shown, the grid dot matrix setting module in this embodiment includes a second partitioning unit and a second conversion unit, wherein,

[0172] The second partitioning unit is used to construct a square grid matrix on the standard head model based on the 10-20 system; specifically, it includes:

[0173] Obtain the positions of C3 / C4 points marked by the 10-20 system, select C3 / C4 points as the selected points, and use the selected points as the unique intersection points to determine a plane that intersects with the scalp layer; construct an N*N grid covering the primary motor zone by radiating outwards in all directions at a defined interval on the plane with the selected points as the center, and map it onto the standard head model.

[0174] The second transformation unit is used to transform the mesh matrix on the standard head model to the individual space based on the transformation matrix A.

[0175] Specifically, after determining the location of the patient's primary motor cortex, it is necessary to search for hotspots within the primary motor cortex. This requires a rough estimation of the area controlling hand movement within the primary motor cortex. The area controlling hand movement is roughly located in the anterior part of the precentral gyrus and paracentral lobule, near points C3 and C4 of the 10-20 system. The curvilinear distance from points C3 and C4 to the center of the primary motor cortex is S, and the value of S can be determined. Then, a square grid of points (N*N, where N is a positive integer) needs to be automatically generated.

[0176] This embodiment proposes a method for setting up a grid of dots in the primary motor area without individual patient MRI data. This method can quickly set up a grid of dots without MRI.

[0177] Specifically, using C3 / C4 as the selected points, a grid matrix is ​​constructed according to the process of constructing an N*N grid matrix on the standard template in Example 1; then, the already obtained registration matrix is ​​used to transform it into the patient's individual space, and the transformation method is the same as in Example 1 above.

[0178] This completes the grid division of the primary motion area.

[0179] Next, the optical navigation module and hotspot detection module are used to achieve hotspot location and search. The optical navigation module and hotspot detection module in this embodiment are the same as those in Embodiment 1 above, and will not be described again here.

[0180] The transcranial magnetic stimulation hand hotspot automatic search system based on optical navigation provided in this embodiment combines the 10-20 system with the point cloud registration method to achieve the localization of the primary motor zone, which can more intuitively reflect the localization process of the motor zone and is easy to operate.

[0181] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. An automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation, characterized in that, include: The brain region localization module is used to locate brain regions based on point cloud registration methods in the absence of MRI data of the individual to be tested, so as to obtain the primary motor area of ​​the cerebral cortex of the individual to be tested. The grid matrix setting module is used to divide the primary motion area into grids to obtain a grid matrix; An optical navigation module is used to locate the grid matrix to determine the position of each stimulation point on the head of the individual to be tested, and to apply stimulation to each stimulation point in sequence. The hotspot detection module is used to collect and analyze the physiological signals generated after stimulation of each stimulation site in order to determine the hotspots of hand movements. The brain region localization module includes: The first three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on standard NMR data to obtain a standard head model, and to establish a virtual coordinate system for the standard head model; The point cloud registration unit is used to track the head of the individual under test using a camera and extract several key points of the head in conjunction with a positioning probe. At the same time, by establishing the connection between the head tracking coordinate system, the camera coordinate system, the probe coordinate system and the virtual coordinate system, the individual space is registered with the standard head model. The head tracking coordinate system is the coordinate system corresponding to the actual head of the individual under test, the camera coordinate system is the coordinate system corresponding to the camera, and the probe coordinate system is the coordinate system corresponding to the positioning probe. The segmentation unit is used to segment the standard brain atlas (AAL) to obtain different brain regions in the virtual coordinate system, and to register the brain regions to the individual space according to the connection between the head tracking coordinate system and the virtual coordinate system; at the same time, it transforms the standard head model to the individual space according to the connection between the head tracking coordinate system and the virtual coordinate system. The first mapping unit is used to perform three-dimensional visualization and cortical mapping of standard head models and brain regions in individual space to obtain the primary motor area.

2. The automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation according to claim 1, characterized in that, The point cloud registration unit uses a camera to track the head of the individual under test and a positioning probe to extract several key points of the head. Simultaneously, by establishing connections between the head tracking coordinate system, camera coordinate system, probe coordinate system, and the virtual coordinate system, it achieves registration between the individual space and the standard head model, including: An infrared reflective ball mounted on the head support of the individual under test is tracked by a camera. The three-dimensional coordinates of the infrared reflective ball in the camera coordinate system are extracted, and the transformation matrix from the head tracking coordinate system to the camera coordinate system is calculated accordingly. B ; The camera tracks and positions the reflective sphere on the probe, and a probe coordinate system is constructed accordingly. The transformation matrix from the probe coordinate system to the camera coordinate system is then calculated. D ; Based on the transformation matrix B and transformation matrix D Obtain the coordinates of several key points on the head of the individual under test in the head tracking coordinate system; Based on the coordinates of several key points on the head of the individual under test in the head tracking coordinate system and the coordinates of these key points in the virtual coordinate system, a point cloud registration function is used to obtain the transformation matrix from the head tracking coordinate system to the virtual coordinate system. A ; For the transformation matrix A Inverting the transformation matrix yields the transformation matrix from the standard head model to the actual head model, enabling registration between the individual space and the standard head model.

3. The automatic search system for transcranial magnetic stimulation hand hotspots based on optical navigation according to claim 2, characterized in that, The method involves using a camera to track an infrared reflective ball mounted on the head support of the individual under test, extracting the three-dimensional coordinates of the infrared reflective ball in the camera coordinate system, and calculating the transformation matrix from the head tracking coordinate system to the camera coordinate system based on this. B ,include: A binocular camera was used to track the infrared reflective spheres mounted on the head support of the individual under test, so as to extract the three-dimensional coordinates of the infrared reflective spheres in the camera coordinate system. The coordinates of the three external reflective spheres P1, P2, and P3 on the head support in the camera coordinate system were obtained and denoted as: P1(Xa,Ya,Za), P2(Xb,Yb,Zb), P3(Xc,Yc,Zc). Calculate the transformation matrix from the head tracking coordinate system to the camera coordinate system based on the coordinates of the reflective ball in the camera coordinate system. B Its expression is: ； in, R 1 is the rotation matrix for transforming the head tracking coordinate system to the camera coordinate system, expressed as: R 1=[α y T ,α x T ,α z T ]； α y α x α z is the unit vector of each axis of the head tracking coordinate system, and T is the transpose operation; T 1 is the translation matrix from the head tracking coordinate system to the camera coordinate system, expressed as: T 1=(Xb+d min1 *cos∠θ3, Yb+d min1 *cos∠θ3, Zb+d min1 *cos∠θ3) T d min1 Let θ be the minimum distance between the three reflective balls P1, P2, and P3 on the head support, and let θ3 be the angle formed from P1 through P2 to P3.

4. The automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation according to claim 2, characterized in that, The camera tracks and positions the reflective sphere on the probe, and a probe coordinate system is constructed accordingly. The transformation matrix from the probe coordinate system to the camera coordinate system is then calculated. D ,include: Use the camera to track and locate all the reflective balls on the probe, and obtain the coordinates of the top three balls in the camera coordinate system, denoted as P4(Xd,Yd,Zd), P5(Xe,Ye,Ze), and P6(Xf,Yf,Zf). Construct a probe coordinate system and calculate the transformation matrix from the probe coordinate system to the camera coordinate system. D Its expression is: ； in, R 2 is the rotation matrix for transforming the probe coordinate system to the camera coordinate system, expressed as: R 2=[β y T ,β x T ,β z T ]； β x β y β z Let T be the unit vector of each coordinate axis of the probe coordinate system, and T be the transpose operation; T 2 is the translation matrix for transforming the probe coordinate system to the camera coordinate system, expressed as: T 2 = (Car + d) min2 *cos∠θ2,Ye+ d min2 *cos∠θ2,Ze+ d min2 *cos∠θ2) T d min2 θ2 is the minimum distance between the three reflective balls P4, P5, and P6 on the positioning probe, and θ2 is the angle formed from P4 through P5 to P6.

5. The automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation according to claim 2, characterized in that, The grid dot matrix setting module includes: The first partitioning unit is used to construct a square grid matrix on the standard head model; specifically, it includes: On one side of the curve, a point S is selected at a distance S from the midpoint of the curve as the unique intersection point, and a plane intersecting with the scalp layer is determined. With the selected point as the center, an N*N grid covering the primary motor zone is constructed radiating outwards in all directions at a determined interval on the plane, and mapped onto the standard head model. The curve is the curve formed by the intersection of the plane formed by the center point of the primary motor zone and two key points of the head of the individual being tested with the scalp layer. The first transformation unit is used to perform transformation based on the transformation matrix. A Transform the mesh matrix on the standard head model to the individual space.

6. The automatic transcranial magnetic stimulation hand hotspot search system based on optical navigation according to claim 2, characterized in that, The brain region localization module includes: The second three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on standard NMR data to obtain a standard head model, and to establish a virtual coordinate system for the standard head model; The second mapping unit is used to map the 10-20 system onto the standard head model to determine the positions of points C3 and C4, thereby obtaining the primary motor areas of the cerebral cortex of the individual under test; wherein point C3 corresponds to the center of the left primary motor area and point C4 corresponds to the center of the right primary motor area.

7. The automatic transcranial magnetic stimulation hand hotspot search system based on optical navigation according to claim 6, characterized in that, The grid dot matrix setting module includes: The second partitioning unit is used to construct a square grid matrix on the standard head model based on the 10-20 system; specifically, it includes: Obtain the position of C3 / C4 points marked by the 10-20 system on the standard head model. Select C3 / C4 points as the selected points and use the selected points as the unique intersection points to determine a plane that intersects with the scalp layer. Construct an N*N grid of points covering the primary motor zone by radiating outwards in all directions at a defined interval from the selected points on the plane, and map it onto the standard head model. The second conversion unit is used to convert the mesh matrix on the standard head model to the individual space based on the conversion matrix A.

8. The automatic transcranial magnetic stimulation hand hotspot search system based on optical navigation according to claim 1, characterized in that, The optical navigation module is specifically used for: Transform the grid points of the primary motion area in the camera coordinate system to the robot arm coordinate system; Based on the robotic arm coordinate system, the program controls the robotic arm to drive the stimulation coil to a designated point for stimulation.

9. The automatic hotspot search system for transcranial magnetic stimulation of the hand based on optical navigation according to claim 1, characterized in that, The hotspot detection module is specifically used for: Collect composite points at the corresponding hand muscles, and when a MEP positive result is obtained, identify the corresponding grid points as hotspots and record them; Stimulation and MEP positivity determination of all grid points on the primary motor zone are achieved through a certain path, and all hotspots are aggregated to obtain the hotspot region.

Images