CN106420055A - Brain tissue deformation correction system based on wireless transmission - Google Patents

Abstract

The invention belongs to the field of medical image processing and application, and relates to a brain tissue deformation correction system based on wireless transmission, in particular to a brain tissue deformation correction system in neurosurgery. The system includes a brain tissue deformation workbench and a 3D laser scanner workbench. The brain tissue deformation correction system workbench is equipped with a brain tissue deformation correction software system, which can communicate with the neurosurgery navigation system through a wireless local area network. The brain tissue deformation correction software system includes a three-dimensional visualization module, a calibration module, a brain tissue extraction module, a meshing module, a boundary condition acquisition module, a finite element calculation module, a preoperative image update module and a communication module. The accuracy of this system is reliable, and it can be integrated into the existing neurosurgery navigation system to help realize the correction of intraoperative soft tissue deformation, thereby greatly improving the accuracy of the navigation system and contributing to clinical application.

Description

A Brain Tissue Deformation Correction System Based on Wireless Transmission

technical field

The invention belongs to the field of medical image processing and application, and relates to a brain tissue deformation correction system based on wireless transmission, in particular to a brain tissue deformation correction system in neurosurgery. The system includes a brain tissue deformation workbench and a three-dimensional laser scanner workbench, which can be integrated into the existing neurosurgery navigation system to help achieve intraoperative soft tissue deformation correction, thereby greatly improving the accuracy of the navigation system and contributing to clinical applications.

Background technique

The prior art discloses that a neurosurgery navigation system (Image guided Neurosurgery System, IGNS) is helpful to improve the quality of surgical operation and reduce brain injury during operation. However, during the operation, the brain tissue will be deformed due to the influence of gravity, traction or resection, resulting in a decrease in the accuracy of the neurosurgical navigation system based on preoperative images.

Used in clinical practice, for example, intraoperative magnetic resonance imaging can provide real-time images of intraoperative brain tissue, and is considered an effective method for intraoperative management of brain displacement, but intraoperative magnetic resonance technology is expensive, which limits its use. Clinical application; biomechanical model, using the biomechanical properties of soft tissue to constrain the motion characteristics of soft tissue, and estimating the deformation of the entire brain tissue with the help of finite element equations and intraoperative sparse data, can overcome the shortcomings of intraoperative imaging, Miga[1,2] used the consolidation theoretical model to simulate and correct the deformation of brain tissue after craniotomy, and proved through animal experiments that: the consolidation theoretical model can correct about 75% of the brain tissue deformation error; another 4 clinical cases proved that: after the improvement, the consolidation The average correction rate of the theoretical model can reach 79%. The disadvantage of this method is that it is very difficult to obtain the boundary conditions. This is because the loss of cerebrospinal fluid during the operation is an important parameter for solving the theoretical model of consolidation. That is, to obtain the boundary conditions, it is necessary to It is necessary to know the changes of cerebrospinal fluid after craniotomy, which is also a difficult problem that has been wanted to be solved in clinical practice; Ferrant[3] et al. use linear elastic biomechanical model to model, and use intraoperative magnetic resonance to obtain boundary conditions, adopt this method The biggest defect of this method is that it does not get rid of the dependence on intraoperative magnetic resonance to reduce the displacement error of the surface of the cerebral cortex to about 1mm. Compared with the consolidation theory model, the linear elastic model uses surface force or surface displacement as the boundary condition. If cheap intraoperative auxiliary equipment such as a 3D scanner is used, combined with the surface tracking algorithm, the boundary condition can be easily obtained, which is convenient for clinical practice. application. Based on the current state of the art, the inventors of the present application intend to provide a system for correcting brain tissue deformation based on wireless transmission, especially a system for correcting brain tissue deformation during neurosurgery, so as to help realize intraoperative soft tissue deformation correction, thereby The accuracy of the navigation system is greatly improved, which is helpful for clinical application.

References relevant to the present invention are:

[1] M.I.Miga, K.D.Paulsen, P.J.Hoopes, F.E.Kennedy, Jr., A.Hartov, and D.W.Roberts,"In vivo quantification of a homogeneous brain deformation model for updating preoperative images during surgery,"IEEE Trans Biomed Eng, vol.47,pp.266-73,2000.

[2] M.I.Miga, K.D.Paulsen, J.M.Lemery, S.D.Eisner, A.Hartov, F.E.Kennedy, and D.W.Roberts, "Model-updated image guidance: initial clinical experience with gravity-induced brain deformation," IEEE Trans Med Imaging, vol. 18, pp.866-74, 1999.

[3] M.Ferrant, A.Nabavi, B.Macq, F.A.Jolesz, R.Kikinis, and S.K.Warfield,"Registration of 3-D intraoperative MR images of the brain using a finite-element biomechanical model,"IEEE Trans Med Imaging, vol.20, pp.1384-97, 2001.

Contents of the invention

The purpose of the present invention is to overcome the defects of the prior art and provide a system for correcting brain tissue deformation based on wireless transmission, especially a system for correcting brain tissue deformation in neurosurgery. The system consists of a brain tissue deformation workbench and a three-dimensional laser scanner workbench. The brain tissue deformation workbench exchanges data with the neurosurgery navigation system through wireless communication. The data includes preoperative images, transformation matrices, and corrected and updated images; The system can be integrated into the existing neurosurgery navigation system to help achieve intraoperative soft tissue deformation correction, thereby greatly improving the accuracy of the navigation system and contributing to clinical applications.

The brain tissue deformation correction system based on wireless transmission of the present invention is composed of a brain tissue deformation workbench and a three-dimensional laser scanner; the brain tissue deformation workbench is equipped with a brain tissue deformation correction software system; the scanning resolution of the three-dimensional laser scanner is 500 microns Above, the accuracy is above 1mm.

In the present invention, the brain tissue deformation correction software system includes a three-dimensional visualization module, a calibration module, a brain tissue extraction module, a meshing module, a boundary condition acquisition module, a finite element calculation module, a preoperative image update module, and a communication module; wherein, the three-dimensional The visualization module is used for 3D visual display of the image sequence, the calibration module is used to obtain the coordinate transformation relationship between the space of the 3D laser scanner and the space of the tracking tool fixed on the surface of the 3D laser scanner, and the brain tissue extraction module is used to extract the brain tissue Extracted from the preoperative image to obtain the target tissue, the meshing module is used to discretize the target tissue into small grid units, the boundary condition acquisition module is used to obtain the boundary conditions of the finite element model, and the finite element calculation module is based on finite element The method calculates the displacement of all grid nodes. The preoperative image update module adopts the interpolation algorithm to obtain the deformed preoperative image based on the grid node displacement and preoperative image calculation. The communication module is used to realize the brain tissue deformation correction system and neurosurgery. Data exchange for navigation systems.

Specifically, in the present invention, the brain tissue deformation correction software system has a workflow as follows:

(1) Calibrate the 3D laser scanner before operation to obtain the spatial coordinate transformation relationship between the 3D laser scanner space and the tracking tool space fixed on the surface of the 3D laser scanner, and store this spatial coordinate transformation relationship in the brain tissue deformation work on stage;

(2) Transfer preoperative images from the neurosurgery navigation system to the brain tissue deformation workbench;

(3) Using the automatic segmentation algorithm combined with the manual segmentation method to extract the brain tissue from the preoperative image to obtain the target tissue;

(4) Combine the octree-like and MT (Marching Tetrahedron) algorithm to grid the above targets;

(5) Establish a linear elastic physical model of the target tissue, and assign corresponding biomechanical properties to each grid unit;

(6) Obtain boundary conditions with the help of 3D laser scanner and sub-block surface tracking algorithm;

(7) Using the finite element method combined with boundary conditions to solve the linear elastic physical model, and obtain the deformation of any position of the target tissue;

(8) The reverse interpolation algorithm is used to update the preoperative three-dimensional image data based on the obtained whole-brain deformation field, that is, the deformed grid, and perform three-dimensional visual display.

(9) The predicted updated image is transmitted back to the neurosurgical navigation system through the communication module.

Among them, the calibration process of the 3D laser scanner is realized through the calibration module, including the following steps:

(1) Start the 3D laser scanner to scan the calibration target, and save the scanned point cloud file into the brain tissue deformation workbench;

(2) Open the point cloud file and select 5-7 feature points from the calibration target point cloud image;

(3) Select the corresponding feature points in the point cloud image on the calibration target by means of the tracking device and the probe that can be tracked by the tracking device;

(4) Use the point registration algorithm to register the two to obtain the coordinate transformation relationship between the 3D laser scanner space and the tracking tool space.

Among them, the acquisition of preoperative images is realized through the communication module; the surgical navigation system is used as the server, and the brain tissue deformation correction system is used as the client. The client first sends a request, and the server responds according to the handshake command; Transmit data, and finally end the transmission with a command; all commands are represented by character strings, starting with CMD, and the handshake command is: CMDASKIMG (request to transmit preoperative images, sent by the client); CMDREQACK (agreed to transfer, sent by the server); end command It is CMDSENDFIN (end of transmission, issued by the server).

Among them, the extraction of the target brain tissue is realized by the brain tissue extraction module, which includes two steps: first, the automatic segmentation algorithm is used to automatically extract the brain tissue; and then the manual segmentation method is used to accurately segment the extracted brain tissue. In the automatic segmentation algorithm, first obtain the upper and lower limit gray values of the image and the rough threshold t of the tissue and the background through the gray histogram; then use the threshold t to estimate the general position of the tissue center of gravity in the image, and estimate The approximate size of the tissue (represented as a sphere); the tissue surface is then modeled as a discrete triangular mesh surface. The initial model is a discrete triangular mesh sphere, the center of which is located at the tissue center of gravity, and the radius is 1/2 of the estimated tissue radius; finally, the initial spherical discrete triangular mesh is deformed slowly, one vertex at a time. When the vertices are deformed to the boundaries of the brain tissue, the deformation force needs to be respected to ensure the smooth surface of the tissue.

Among them, the meshing processing of the target brain tissue is realized through the meshing module. First, the three-dimensional image space is divided into grids of hexahedral units by using the octree-like algorithm, and each hexahedral unit is divided into five tetrahedral units. ; Then, use the MT-like algorithm to cut the grid, remove the background, and finally obtain the 3D grid unit that only contains brain tissue. Finally, based on the FEM deformable mesh method, the shape of the generated 3D tetrahedral mesh is adjusted so that the nodes on the mesh boundary are aligned with the boundary of the corresponding 3D binary image.

Among them, the physical model of the target tissue is established, and corresponding biomechanical properties are assigned to each grid unit, which is realized in the finite element module. Since the deformation process of the brain tissue is slow and the deformation is small, and the strain and the stress have a linear relationship, the present invention simulates it as a uniform elastic body based on the theory of linear elasticity, and the physical model is expressed as a series of partial differential equations:

Among them, u is the displacement vector, F is the force vector, μ=E/2(1+v), ν is Poisson's ratio, and E is the modulus of elasticity. The mechanical properties of the linear elastic model are assigned to each grid unit: Poisson's ratio ν and elastic modulus E.

Among them, the acquisition of boundary conditions with the help of 3D laser scanner and sub-block surface tracking algorithm is completed in the boundary condition acquisition module. First, the surface grid nodes are extracted as the initial surface of soft tissue; second, the deformed soft tissue surface points are acquired by 3D laser scanner set, as the deformed soft tissue surface; thirdly, use the rigid registration method to initially register the deformed soft tissue surface and the initial soft tissue surface; fourthly, based on the sub-block energy function minimum non-rigid registration algorithm to obtain two The mapping relationship between points in the point set.

In the present invention, after the initial soft tissue surface and the deformed soft tissue surface are obtained, the rigid registration method is used to initially register the deformed soft tissue surface and the initial soft tissue surface; in one embodiment of the present invention, the coordinate system conversion of different spaces is used Realize the rigid registration of the initial soft tissue surface and the deformed soft tissue surface, including the coordinate transformation of the following four spaces: (1) Transform the 3D laser scanner space to the tracking tool space; (2) Transform the tracking tool space to the tracking device space ; (3) Transform the space of the tracking device into the space of the reference frame; (4) Transform the space of the reference frame into the image space, the image space is the space where the soft tissue is before deformation, wherein the tracking tool is fixed on the 3D laser scanner, and the tracking tool can be Track device tracking. The reference frame is fixed on the patient's body, and the reference frame can also be tracked by the tracking device. The first spatial coordinate transformation is realized in the calibration module through preoperative calibration; the second and third spatial coordinate transformations are achieved through the intraoperative tracking device. Tracking is realized; the fourth spatial coordinate transformation is realized through the rigid registration of the preoperative image and the marker point in the patient space, and the last three spatial coordinate transformations are obtained in the neurosurgical navigation system and transmitted to the brain tissue deformation correction system through the communication module.

In the present invention, after initial registration between the deformed soft tissue surface and the initial soft tissue surface, the mapping relationship between the two point-concentrated points is obtained based on the sub-block type energy function minimum non-rigid registration algorithm, and the sub-block type The surface tracking algorithm consists of steps:

(1) Divide the initial surface grid node X={x,i=1,2,...M} of the soft tissue to be tracked into several small blocks: P={p,k=1,2,...L}, Except for the small block p _L , each other small block p _k contains l nodes, and the small block p _L contains M-(L-1)*l nodes;

(2) Construct the energy minimum equation for each point set small block p _k to obtain the surface displacement of the point:

Where p _k ={p,o=1,2,...l} is a point set small piece, Y={y,j=1,2,...N} is a point set of a point cloud. c _jk is a correlation matrix defining the corresponding probability between two point sets, and λ is a weighting factor. Use thin-plate spline (Thin-plateSpline) non-rigid registration algorithm to calculate the mapping function f between the deformed cortex (Y) and the undeformed cortex (p _k ), so as to obtain the point set small block surface displacement Disp _pk ;

(3) Gather the displacement Disp _pk of each small block of point set, and obtain the displacement Disp _x of all soft tissue surface grid nodes that need to be tracked.

Among them, the finite element method combined with boundary conditions is used to solve the linear elastic physical model, and the deformation of any position of the target tissue is obtained in the finite element calculation module. The linear elastic physical model consists of a series of partial differential equations, which can be written as follows: Ka=P, where K is the rigid matrix, a is the node displacement, and P is the node load. K is determined by two parameters, Young's modulus of elasticity E and Poisson's ratio ν. We use a linear solver based on PETSc (Portable, Extensible Toolkit for Scientific Computation) to realize multi-processor parallel computing to speed up the computing speed.

Wherein, updating the preoperative three-dimensional image is completed in the updating module. Use the reverse interpolation method to update the preoperative image: start from the deformed grid unit, find the integer coordinate point in the unit, use the shape function to obtain the position of the point before the deformation, and then use trilinear interpolation to obtain the point The gray value of , and then update the three-dimensional data field of the preoperative image.

Among them, the 3D visualization display is completed in the 3D visualization module. After the preoperative 3D image data is updated based on the obtained whole-brain deformation field using the reverse interpolation algorithm, the 3D data field is displayed using the ray projection algorithm. In the 3D visualization module, through setting The volume rendering function of the 3D data field can be realized by setting parameters such as light and shadow, sharpness, color mapping, opacity mapping and reconstruction speed.

Wherein, the transmission of the predicted updated image back to the neurosurgery navigation system is implemented in the communication module. The surgical navigation system serves as the server, and the brain tissue deformation correction system serves as the client. The client sends out the request first, and the server responds according to the handshake command. The handshake is first performed by command, then the data is transmitted, and finally the transmission is ended by command. The image data is placed in a one-dimensional array, and the handshake commands of the application layer are as follows: CMDSENDREQ (request for image transmission, sent by the client); CMDREQACK (agreed to receive image update data, sent by the client); CMDSENDFIN end of transmission, sent by the client) .

The present invention has reliable precision, can be conveniently integrated into the existing navigation system of neurosurgery, helps to realize soft tissue deformation correction during operation, thus greatly improves the precision of the navigation system, and is helpful for clinical application.

Description of drawings

Fig. 1 is a block diagram of the software system of the brain tissue deformation correction system.

Fig. 2 is a working flow diagram of the brain tissue deformation correction system.

Fig. 3 is the software system interface of the brain tissue deformation correction system and the calculation results of the three-dimensional visualization module.

Figure 4 is the determination of the boundary conditions, where 1 is the point cloud data of the cortical surface acquired by the 3D laser scanner, 2 is the result of initial registration between the deformed soft tissue surface and the initial soft tissue surface, 3 is the result before surface tracking, and 4 is Get the result of the mapping relationship between points in two point sets.

detailed description

Embodiment 1 clinical trial

1. Install the tracking tool on the 3D laser scanner, and use the calibration module to calibrate to obtain the coordinate transformation relationship from the space transformation of the 3D laser scanner to the space of the tracking tool;

2. The brain tissue deformation correction workbench communicates with the neurosurgical navigation system to request the transmission of preoperative image data. The brain tissue deformation correction workbench is the client, and the neurosurgical navigation system is the server. The client first sends the command code CMDASKIMG; the server confirms that the command code is correct, and sends CMDREQACK to confirm; the server sends the preoperative image data to the client, and stores the 240×240×197 three-dimensional MRI brain tissue data in a one-dimensional array; the server Send CMDSENDFIN to stop the transmission; after receiving CMDSENDFIN, the client stops receiving and saves the preoperative image data;

3. For the 240×240×197 three-dimensional MRI brain tissue data field, the brain tissue is extracted by using an algorithm combining automatic segmentation and manual segmentation, and the score gray threshold is set to 0.5;

4. Discrete the extracted brain tissue into 162,650 grid units with a tetrahedron grid, the number of nodes is 30,150, and the maximum tetrahedron size is 5×5×5mm ³ , and the grid surface nodes are extracted;

5. Set the biomechanical property parameters of the brain tissue for each unit of the online elasticity model. Young's modulus=3Kpa, Poisson's ratio=0.45;

6. Fix the reference frame on the patient's head, and the reference frame can be tracked by the tracking device. Install the 3D laser scanner on a tripod with standard degrees of freedom, adjust the tripod so that the acquisition lens is located at 30 to 60 cm in the normal direction of the surface of the craniotomy cortex, and scan to obtain the point cloud of the deformed soft tissue surface in the craniotomy area;

7. At the same time of scanning, the coordinate transformation relationship from the tracking tool space to the tracking equipment space and the coordinate transformation relationship from the tracking equipment space to the reference frame space are obtained by means of the optical tracking equipment in the neurosurgical navigation system. The coordinate transformation relationship from the reference frame space to the image space is obtained through the rigid registration of the preoperative image and the marker points in the patient space;

8. With the help of the communication module, the three spatial transformation relationships mentioned in the sixth step are transmitted from the neurosurgical navigation system to the brain tissue deformation correction system. The handshake command is: CMDASKTRANS (request transmission transformation matrix, sent by the client); CMDREQACK (agreed to transfer, sent by the server); the end command is CMDSENDFIN (end of transmission, sent by the server);

9. After obtaining the coordinate relationship of the above four spaces, the brain tissue deformation correction system will register the deformed brain tissue surface and the initial brain tissue surface under the unified space coordinate system to complete the initial registration;

10. Obtain the mesh surface nodes of the craniotomy area (soft tissue surface before deformation) and the cortical point cloud of the deformed brain tissue (soft tissue surface after deformation) through manual delineation, where the mesh surface nodes contain 134 nodes, and the point cloud contains 18127 nodes ;

11. Divide the mesh surface nodes into 17 small blocks, the first 16 small blocks contain 8 nodes, and the last small block contains 6 nodes. Construct the energy minimum equation for each point set small block p _k to obtain the surface displacement of the point:

Wherein p _k ={p,o=1,2,...l} is a point set of small blocks, wherein the first 16 small blocks, l=8, and the last small block, l=6. Y={y,j=1,2,...N}, which is the point set of the point cloud, where N=18127. c _jk is a correlation matrix defining the corresponding probability between two point sets, and λ is a weighting factor. Use the Thin-plate Spline non-rigid registration algorithm to calculate the mapping function f between the deformed cortex (Y) and the undeformed cortex (p _k ), so as to obtain the surface displacement Disp _pk of the point set; collect each point Set the displacement Disp _pk of the small block to obtain the displacement Disp _x of all soft tissue surface grid nodes that need to be tracked;

12. Use the finite element method to convert the linear elastic model equation into a matrix form: Ka=P, where K is the rigid matrix, a is the node displacement, and P is the node load. K is determined by two parameters, Young's modulus of elasticity E and Poisson's ratio ν. Boundary conditions are introduced, the singularity of the matrix K is eliminated, the displacement vector a is obtained, and the displacement at any node inside the brain tissue is obtained. We use a linear solver based on PETSc (Portable, Extensible Toolkit for Scientific Computation) to realize multi-processor parallel computing to speed up the operation. Using a Dell workstation (Intel Core2 DuoE6850, 4 GB RAM), Windows XP operating system, the solution time is 50s;

13. From the deformed data field, calculate the position of all coordinate points (integer coordinate points) that contribute to the display before deformation according to the shape function, and use trilinear interpolation to calculate the gray value of the point;

14. Visualize the three-dimensional data field composed of all these integer coordinate points using the light transmission method for navigation surgery;

15. The predicted updated image is transmitted back to the neurosurgical navigation system through the communication module. The surgical navigation system acts as the server, and the brain tissue deformation correction system acts as the client. The client sends a request first, and the server responds according to the handshake command. First, the handshake is performed through the command, and then the data is transmitted. Placed in a one-dimensional array, the handshake commands of the application layer are as follows: CMDSENDREQ (request for image transmission, sent by the client); CMDREQACK (agreed to receive image update data, sent by the client); CMDSENDFIN transmission ends, sent by the client).

The test results show that the wireless transmission-based brain tissue deformation correction system of the present invention has reliable accuracy, and can be easily integrated into the existing neurosurgery navigation system to help realize intraoperative soft tissue deformation correction, thereby greatly improving the accuracy of the navigation system.

Claims (14)

1. a kind of based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that this system is by cerebral tissue deformation operation Platform and three-dimensional laser scanner workbench composition, described cerebral tissue deformation operation platform carries cerebral tissue straightening software system； The scanning resolution of three-dimensional laser scanner is more than 500 microns, more than precision 1mm.

2. as described in claim 1 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described brain group Knit straightening software system and include three-dimensional visualization module, demarcating module, brain tissue extraction module, gridding module, border Condition acquisition module, FEM calculation module, pre-operative image update module and communication module.

3. as described in claim 1 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described cerebral tissue Its workflow of straightening software system is：

(1) pre-operative image is transferred to from neurosurgery navigation system by cerebral tissue deformation operation platform by communication module；

(2) cerebral tissue is extracted from pre-operative image using automatically splitting the algorithm combining with manual segmentation, obtain target Tissue；

(3) with reference to class Octree and MT algorithm, three-dimensional grid is carried out to above-mentioned target；

(4) give corresponding biomechanicss attribute to each grid cell, set up the physical model of destination organization；

(5) obtain boundary condition by three-dimensional laser scanner and Subblock-wise surface tracking algorithm；

(6) carry out FEM calculation with reference to boundary condition and physical model, obtain the deformation of destination organization optional position；

(7) using turn back to algorithm acquisition deformation after grid update preoperative 3 d data field；

(8) transmitting image after forecast updating by communication module can neurosurgical Navigation System guides.

4. as described in claim 3 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (3) In three-dimensional grid be tetrahedral grid or hexahedral mesh.

5. as described in claim 3 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (5), in, described Subblock-wise surface tracking algorithm includes step：

(1) soft tissue surfaces point set after three-dimensional laser scanner obtains deformation, as the soft tissue surfaces after deformation；

(2) just registration will be carried out to soft tissue surfaces after deformation and initial soft tissue surfaces using Rigid Registration method；

(3) obtain the mapping between two point centrostigmas and point based on Subblock-wise energy function minimum non-rigid registration algorithm to close System.

6. as described in claim 3 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (6) in, using based on PETSc (Portable, Extensible Toolkit for Scientific Computation) structure The linear solution device built is realized parallel multiprocessor and is calculated, and accelerates arithmetic speed.

7. as described in claim 5 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (2) coordinate system based on different spaces for the Rigid Registration method in is changed.

8. as described in claim 3 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (5), in, its Subblock-wise surface tracking algorithm includes step：

(1) the soft tissue initial surface grid node X={ x, i=1,2 ... M } needing to follow the tracks of is divided into some fritters：P ={ p, k=1,2 ... ... L }, except fritter p_LIn addition, other each fritter p_kComprise l node, fritter p_LComprise M- (L-1) * L node；

(2) it is each point set fritter p_kBuild energy minimum equation and obtain its surface displacement of point：

Wherein p_k={ p, o=1,2 ... ... l } is point set fritter, and Y={ y, j=1,2 ... ... N } is the point set of a cloud, c_jkIt is fixed The correlation matrix of corresponding probability between adopted two point sets, λ is weight factor；With thin plate spline (Thin-plate Spline) Non-rigid registration algorithm calculates deformation cortex (Y) and not deformed cortex (p_k) between mapping function f, obtain point set fritter surface position Move Disp_pk；

(3) gather the displacement Disp of each point set fritter_pk, obtain the displacement of the soft tissue surfaces grid node needing to follow the tracks of Disp_x.

9. as described in claim 5 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that described step (2) coordinate system based on different spaces for the Rigid Registration method in is changed, and it is realized soft group after initial soft tissue surfaces and deformation The first registration knitting surface includes step；

(1) by three-dimensional laser scanner spatial alternation to trace tool space；

(2) by trace tool spatial alternation to tracking equipment space；

(3) by tracking equipment spatial alternation to frame of reference space；

(4) by frame of reference spatial alternation to image space, image space is soft tissue place space before deformation.

10. as described in claim 9 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that wherein, described Trace tool be fixed on three-dimensional laser scanner, the tracked equipment of trace tool follow the tracks of；Frame of reference is fixed on patient body On, the tracked equipment of frame of reference is followed the tracks of.

11. as described in claim 9 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that wherein, described (1) space coordinate transformation is by realizing in preoperative calibration process in cerebral tissue straightening software system；Described (2) (3) (4) space coordinate transformation is realized in operation guiding system, is transferred to cerebral tissue straightening software body by communication module System uses.

12. as described in claim 11 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that wherein, institute The preoperative calibration process stated includes step：

(1) start 3D scanner scanning alignment target, scanning institute invocation point cloud file is stored in cerebral tissue deformation operation platform；

(2) open a cloud file, choose 5-7 characteristic point using mouse from demarcation target spot cloud atlas picture；

(3) by tracking equipment and can be tracked probe demarcate target on choose and corresponding characteristic point in point cloud chart picture；

(4) using point registration Algorithm, the two registration obtains the seat between trace tool space for the three-dimensional laser scanner space Mark transformation relation.

13. as described in claim 1 based on the cerebral tissue straightening system being wirelessly transferred it is characterised in that cerebral tissue become Shape workbench carries out data exchange by wireless telecommunications and neurosurgery navigation system, and operation guiding system is as server End, cerebral tissue straightening system is as client.

Being navigated for neurosurgery based on the cerebral tissue straightening system being wirelessly transferred described in 14. claim 13 Purposes in the system integration；Including：Request is sent first by client, server end gives a response according to order of shaking hands； Transmission for space conversion matrices：First pass through order to be shaken hands, then transmission data, terminate to pass finally by order Defeated, all order string representations, started with CMD, order of wherein shaking hands is：CMDASKTRANS (request propagation and transformation square Battle array, client sends)；CMDREQACK (agrees to transmission, server end sends)；Terminate order for CMDSENDFIN (transmission knot Bundle, server end sends)；The transmission of pre-operative image is same with the mode of space conversion matrices transmission；For image after forecast updating Transmission, view data is placed in an one-dimension array, and the order of shaking hands of application layer is as follows：CMDSENDREQ (request transmission figure Picture, client sends)；CMDREQACK (agrees to receive data after image update, client sends)；CMDSENDFIN transmission knot Bundle, client sends).