Patient-specific cortical surface tessellation into dipole patches
The method improves dipole tessellation by varying patch numbers based on head size, enhancing the accuracy of electrical source imaging and patient-specific diagnoses through precise dipole patch representation.
Patent Information
- Application Number
- JP2021545695
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-02-07
- Filing Date
- 2020-02-07
- Publication Date
- 2024-11-14
- Estimated Expiration
- 2040-02-07
AI Technical Summary
Current dipole tessellation methods in electrical source imaging limit the number of dipole patches to a predetermined constant, resulting in inaccurate representation of cortical geometry due to simultaneous coverage of gyri and sulci, affecting the accuracy of electrical source imaging.
A method for cortical surface tessellation that varies the number of dipole patches based on head size, using a volumetric bitmask, triangular mesh representation, and topology-preserving mesh decimation to create accurate dipole patches with defined normals, improving the accuracy of electrical source imaging.
This method enhances the precision of electrical source imaging by accurately depicting dipole patches, leading to more accurate and patient-specific diagnoses and surgical planning.
Smart Images

Figure 0007583726000001 
Figure 0007583726000002 
Figure 0007583726000003
Abstract
Description
[Technical field]
[0001] The present invention relates generally to imaging of electrical sources, and more particularly to establishing forward models for use in imaging of electrical sources. [Background technology]
[0002] Epilepsy is a public health problem that affects nearly 3 million individuals in the United States alone. Surgery is the most effective treatment for medically refractory epilepsy and is curative if the seizure focus can be localized. Minimally invasive interventional procedures to remove the seizure focus have significant advantages over open surgery, such as providing better tissue preservation and better overall functional outcomes. Both minimally invasive ablation and open surgery require the use of high-resolution structural magnetic resonance imaging (MRI) to identify and visualize the seizure focus. Electrical source imaging (ESI) is a neuroimaging technique. ESI from electroencephalography (EEG) electrodes placed on the scalp can localize the seizure focus on the cortical surface. Using scalp potentials measured by electroencephalography (EEG) electrodes, ESI estimates the location of the sources in the brain that are responsible for these scalp potentials. The propagation of electrical activity is governed by partial differential equations. In ESI, these partial differential equations are used to establish a forward model and its inverse solution. The forward model is established to describe how electrical activity is generated by cortical sources.
[0003]
[0003] The accuracy of the forward model depends on the individual head geometry and the electrical conduction properties of the head tissue types, including gray matter, white matter, cerebrospinal fluid, scalp, skull, air spaces, and eyes. In establishing the forward model, the gray matter is tessellated into surface patches to further define the location and orientation of the dipoles. Tessellating this gray matter can be difficult. The human cerebral cortex is wrinkled to accommodate more brain tissue within the closed skull. The grooves that form these wrinkles are called sulci, and the ridges between them are known as gyri. Thus, the shape of the surface of the cortical gray matter is deeply convoluted. The goal of dipole tessellation is to divide the cortical surface into patches that may be the source of abnormal electrical discharges. A dipole patch is represented by the center of the patch, the normal to the patch, and the surface triangles that belong to the patch. These properties are used as boundary conditions in the forward model. An accurate representation of the geometry at hand is crucial for defining the propagation of electrical activity from the cortex to the scalp.
[0004]
[0004] Current dipole tessellation methods limit the number of dipoles to a predefined constant, resulting in dipole patches that simultaneously cover gyri and sulci. As a result, the normals of the dipole patches are calculated as averages that do not represent the cortical geometry. Summary of the Invention [Problem to be solved by the invention]
[0005] One object of the present invention is to improve current methods of cortical surface tessellation. [Means for solving the problem]
[0006] In a first aspect of the present invention, to better address the above concerns, a method for tessellating the cortical surface with a number of dipole patches, the number of which is variable and head dimension dependent, can be used to improve the accuracy of the forward model, resulting in improved imaging of electrical sources and, more generally, more accurate and patient-specific diagnoses and outcomes.
[0007]
[0006] In one embodiment, the cortical surface consists of the gray matter surface of the cortex. EEG source localization is generally considered an underdetermined problem due to the large number (e.g., infinite) of possible inverse configurations that can result in the same scalp potential. One mechanism for imposing constraints on the solution is to constrain the cortical surface to gray matter.
[0008]
[0007] In one embodiment, the method further comprises the steps of: receiving a volumetric bit mask of the cortex, where each voxel of a plurality of voxels is labeled according to tissue type; providing a triangular mesh representation of the segmented cortical surface based on the volumetric bit mask; providing a high-resolution triangular mesh including triangles with uniform edge lengths by remeshing the triangular mesh representation; and providing a simplified triangular mesh representation by applying a topology-preserving mesh decimation algorithm to the high-resolution triangular mesh representation. The dipole patches are obtained by using the simplified mesh as a template. The mesh decimation algorithm preserves the cortical folds and geometry, providing a more realistic model and therefore improved accuracy for achieving electrical source localization.
[0009]
[0008] In one embodiment, the method further comprises obtaining the plurality of dipole patches by relating each triangle from the high resolution triangular mesh representation to a nearest triangle from the simplified triangular mesh representation, where the final amount of the plurality of dipole patches is equal to the amount of triangles of the simplified triangular mesh representation. In this way, the number of dipole patches is not limited to a predetermined constant, thus eliminating or reducing the risk of merging sulci and gyri.
[0010] In one embodiment, each of the dipole patches has properties that are used as boundary conditions in a forward model used for electrical source imaging, the accuracy of which depends on the individual head geometry and electrical conduction properties of head tissue types including grey matter, white matter, etc., and improved determination of these properties helps to achieve more accurate electrical source imaging results.
[0011]
[0010] In one embodiment, the center of each dipole patch is determined by determining the center of mass of all triangles of the high resolution triangular mesh representation associated with a particular one of the closest triangles, where the center of the high resolution mesh triangle closest to the center of mass is selected as the center of the dipole patch, and the normal of the dipole patch corresponds to the average of all normals of the triangles of the high resolution triangular mesh representation. In this way, a more accurate depiction of the dipole patch may be obtained.
[0012]
[0011] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. [Brief description of the drawings]
[0013] [Figure 1] FIG. 1 is a schematic diagram illustrating an exemplary environment in which a cortical surface tessellation method is implemented, according to one embodiment of the present invention. [Diagram 2]
[0014] FIG. 2 is a flow chart illustrating an exemplary cortical surface tessellation method, according to one embodiment of the present invention. [Diagram 3]
[0015] FIG. 3 is a flow diagram conceptually illustrating an exemplary workflow for selected steps of the cortical surface tessellation method of FIG. 2, in accordance with one embodiment of the present invention. [Figure 4]
[0016] FIG. 4 is an image of an exemplary cortical surface including dipole patches and patch normals resulting from implementation of a cortical surface tessellation method according to one embodiment of the present invention. [Diagram 5]
[0017] FIG. 5 is a block diagram illustrating an exemplary apparatus for implementing all or part of the cortical surface tessellation method according to one embodiment of the present invention. [Figure 6]
[0018] FIG. 6 is a flow chart illustrating an exemplary cortical surface tessellation method, according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0014]
[0012] Many aspects of the present invention can be better understood by reference to the drawings, which are schematic. The components in the figures are not necessarily to scale, emphasis instead being placed on clearly illustrating the principles of the invention. Moreover, in the drawings, like reference characters indicate corresponding parts throughout the several views.
[0015]
[0019] Disclosed herein are specific embodiments of a cortical surface tessellation method (and associated devices and systems) that provides an (automatic) tessellation of the cortical surface into dipole patches necessary for accurate electrical source imaging (ESI) localization. Particular embodiments of the cortical surface tessellation method provide an accurate representation of the propagation of electrical activity from the cortex to the scalp by having dipole patches with normals perpendicular to all triangles within a patch and by having the number of patches dependent on the head dimensions.
[0016]
[0020] As a brief digression, current dipole tessellation methods limit the number of dipole patches to a predefined constant, resulting in dipole patches that simultaneously cover gyri and sulci. In other words, one or more dipole patches may each span several gyri and sulci, resulting in inaccurate dipole normals. In contrast, certain embodiments of the cortical surface tessellation method address this and other issues, improving electrical source imaging by tessellating cortical gray matter within dipole patches and calculating more accurate normals associated with these dipole patches.
[0017]
[0021] Having summarized certain features of the cortical surface tessellation method of the present disclosure, the cortical surface tessellation method will now be described in detail as illustrated in the drawings. While the cortical surface tessellation method will be described in conjunction with these drawings, there is no intent to limit the method to the embodiment or embodiments disclosed herein. Moreover, while the description may identify or describe details of one or more embodiments, such details may not necessarily be part of every embodiment, and not all of the various noted advantages may necessarily pertain to a single embodiment. On the contrary, the intent is to cover all alternatives, modifications, and equivalents included within the principles and scope of the disclosure as defined by the appended claims. For example, two or more embodiments may be interchangeably combined or in any combination. Moreover, in the context of the present disclosure, the claims should not necessarily be limited to the specific embodiments set forth in the description.
[0018]
[0022] FIG. 1 is a schematic diagram illustrating an exemplary environment 10 in which the cortical surface tessellation method is implemented. The environment 10 may include a clinical setting in a hospital or doctor's office or an educational or research facility. The environment 10 includes a subject 12 fitted with an electroencephalogram (EEG) net 14 including a number of electrodes / sensors (hereinafter referred to as electrodes, but used interchangeably herein with sensors). The EEG net 14 may include 32, 64, 128 or 256 EEG electrodes, and thus EEG electrode channels. The EEG net 14 is coupled to a processing unit (PRC DEV) 16. The processing unit 16 has an amplifier configured to filter, measure and sample the EEG signals acquired by the EEG net 14 and then forward the digitized samples to a computation unit (CMPUT) 18 for use in head modeling (e.g., solving forward and inverse problems) and, in general, in estimating cortical activity. Electrical source imaging (ESI) from electrodes of the EEG net 14 placed on the scalp of the subject 12 can be used to localize the seizure focus on the cortical surface. In other words, signals to / from the EEG electrodes of the EEG net 14 are used to record brain activity of the subject 12 from cortical regions of the brain. The environment 10 further includes a magnetic resonance imaging (MRI) system 20. The MRI system 20 acquires brain images and provides (to the computing device 18) high resolution structural images that enable identification and / or visualization of the seizure focus.
[0019]
[0023] The MRI scan of the brain is provided to a computing device 18 where the MRI brain image is segmented to identify or delineate head tissue types including grey matter, white matter, cerebrospinal fluid, scalp, skull, air spaces and eyes. The computing device 18 can render the image on a user interface (UI) 22 (e.g., a display device or display screen) and / or store the image data in a data repository (STORE) 24. For example, the user interface 22 allows for a visual representation of a tessellated cortical surface including an image of the head (or a portion thereof) with one or any combination of dipole patches, centers of dipole patches, normals of dipole patches or cortical sources. The user interface 22 can be integrated into the computing device 18 or embodied as a separate device. Similarly, the processing device 16 can be integrated into the computing device 18 or embodied as a separate device. The data repository 24 may include one or more of a picture archiving and communication system (PACS), a radiology information system (RIS), a hospital information system (HIS), an electronic medical record (EMR) database, a surgical navigation system, a server, a computer, and / or other data repository. The computing device 18 may export data corresponding to brain activity from the cortical surface to the MRI space for import into a surgical navigation system, which may enable planning of optimal ablation trajectories based on retrospective data (e.g., ablation area, activity in the tracts, laser trajectory, patient outcomes, etc.) from the same procedure in different subjects. The user interface 22 and / or the data repository 24 may be located locally to the MRI system 20 or remotely (e.g., in another room, building, or area) and accessed via a network. In some embodiments, the functions of the computing device 18 may be performed at multiple locations (remote and local) according to a distributed processing system, including locally (local to the clinical facility), remotely, or via a cloud platform.
[0020]
[0024] In some embodiments, a cortical surface tessellation system implementing the corresponding method may include an MRI system 20, an EEG net 14, a processing device 16, a computing device 18, a user interface 22, and a data repository 24. In some embodiments, the cortical surface tessellation system may include fewer components. For example, in one embodiment, the cortical surface tessellation system may consist of a non-transitory computer readable storage medium from the computing device 18, or in some embodiments, of or associated with the computing device 18, the medium having executable code / instructions that provide the functionality of a cortical surface to volume mapping system. In some embodiments, the environment 10 may include additional systems and / or devices, including, for example, a computed tomography (CT) system that may be used to characterize skull density and provide images that are registered to the MR images. In some embodiments, an electrode registration system or mobile device may be used that may be used to register electrode positions on the head. For example, a Philips geodesic photogrammetry system (GPS) may be used for electrode registration. In some embodiments, other types of devices may be used for registration.
[0021]
[0025] In general, a cortical surface tessellation system acquires MRI data, generates a head model based on the MRI data (e.g., to solve a forward problem), records EEG data, and then uses the head model to solve an inverse problem to localize a targeted source of brain activity (e.g., a seizure source) based on the EEG data. Based on the localized source, brain surgery (which is usually preceded by a pre-planning step) can be used to remove the source of activity (e.g., a seizure). With respect to generating the head model, a cortical surface tessellation system segments head tissues (e.g., gray matter, skull, etc.), associates different electrical conductivities with each tissue class or type (e.g., using known electrical conductivities from studies and / or literature), identifies or extracts the cortical surface from the MRI data by topology correction and tessellation into dipole patches, aligns electrode locations to the subject's head, and calculates lead field matrices to predict how current will flow from the cortex to the skull (e.g., using one of the finite difference method (FDM), finite element method (FEM), boundary element method (BEM), etc.). For example, the lead field matrices are used to relate dipole activity to changes in measured scalp EEG potentials and are defined at least in part by an accurate model of head conductivity based on the MRI data (or in some embodiments, by an atlas model). Particular embodiments of the cortical surface tessellation method are further described below with a focus on the head modeling stage, and in particular the dipole tessellation process (recognizing that the tessellation process advances the objectives of the overall electrical source imaging process).
[0022]
[0026] Referring now to Figure 2, what is shown is a flow chart illustrating one embodiment of an exemplary cortical surface tessellation method 26. Reference is also made in conjunction with the description of Figure 2 to Figure 3, where Figure 3 conceptually illustrates selected features of method 26. As briefly described above, the head modeling process generally begins with receiving a magnetic resonance imaging scan of a subject's head (e.g., from an MRI system 20 (Figure 1)) and segmenting the magnetic resonance imaging data, which in one embodiment includes tissue segmentation of the cortical surface of the cortex. For example, the image segmentation software of computing device 18 (Figure 1) can access an anatomical model of the brain from an anatomical model bank, which includes a surface representation of a shape-constrained deformable brain model. Examples of suitable brain models are described in L. Zagorchev, A. Goshtasby, K. Paulsen, T. McAllister, S. Young, and J. Weese, “Manual annotation, 3-D shape reconstruction, and traumatic brain injury analysis,” International Workshop on Multimodal Brain Image Analysis (MBIA), Toronto, California, September 2011, and L. Zagorchev, C. Meyer, T. Stehle, R. Kneser, S. Young, and J. Weese, “Evaluation of Traumatic Brain Injury patients using a shape-constrained deformable model,” International Workshop on Multimodal Brain Image Analysis (MBIA), Toronto, California, September 2011, all of which are incorporated by reference in their entireties.Additional information regarding shape-constrained deformable models can be found in Jurgen Weese, Irina Wachter-Stehle, Lyubomir Zagorchev, and Jochen Peters, "Shape-constrained deformable models and applications in medical imaging," SCDM Book, pp. 151-184, Chapter 1400, and Fabian Wenzel, Carsten Meyer, Thomas Stehle, Jochen Peters, Susanne Siemonsen, Christian Thaler, and Lyubomir Zagorchev, "Rapid fully automatic segmentation of subcortical brain structures by shape-constrained surface adaptation," Alzheimer's Disease Neuroimaging Initiative, Medical Image Analysis 46 (2018), pp. 141-161, each of which is also incorporated by reference. Other models are contemplated herein. The image segmentation software is configured to segment the brain based on the anatomy represented in the model by performing an initial registration between the model and the MRI brain image, converting the model to the brain anatomy based on a transformation (such as the Hough transform), performing a parametric adaptation of the model, and performing a deformable adaptation of the model. Other known techniques can also be used.
[0023]
[0027] Referring to the cortical surface tessellation method 26 of FIG. 2, the method 26 includes a step 28 of receiving (e.g., obtained from a memory or network storage) a volumetric bitmask of the cortex in which each voxel is associated with a tissue type.
[0024]
[0028] In step 30, method 26 generates (or obtains) a triangular mesh representation of the segmented gray matter of the cortex based on the bitmask (e.g., via a marching cubes algorithm). For example, referring to FIG. 3, a dipole tessellation workflow 32 corresponding to selected steps of method 26 is shown, in particular a triangular mesh representation based on the bitmask, shown as triangular mesh representation 34. The cortical bitmask is obtained from a segmentation of an MRI scan, resulting in dipole patches with normals perpendicular to the dipole surface. The number of dipoles generated is not fixed and depends on the dimensions (size) of the head (cortex). The dipole tessellation can be performed using a volumetric bitmask of the cortex, where each voxel is labeled according to tissue type.
[0025]
[0029] The marching cubes algorithm can be applied to the bit mask to extract a triangular mesh representation of the segmented gray matter representing the cortex. The marching cubes algorithm generates a polygonal representation of a surface of constant density from a three-dimensional array of data, and can be generally described as having the steps of reading four slices into memory, scanning two slices, creating a cube from the four adjacent nodes on the next slice, calculating the index of the cube by comparing the eight density values of the cube's vertices with a surface constant, using the index to look up a list of edges from a pre-computed table, finding the surface / edge intersections via linear interpolation using the density at each edge's vertex, calculating the unit normal at each cube vertex using the center difference (and interpolating the normal of each triangle vertex), and outputting the triangle vertices and vertex normals. In effect, the marching cubes algorithm determines how the surface intersects with each cube and moves or advances to the next cube. Further information on the marching cubes algorithm can be found in "Marching Cubes: A High Resolution 3D Surface Construction Algorithm" by William E. Lorensen et al., Computer Graphics, Vol. 21, No. 4, July 1997 (see also SIGGRAPH 1987, pp. 163-169), and "The Asymptotic Decider: Resolving the Ambiguity in Marching Cubes" by Gregory M. Nielson et al., IEEE Visualization 1991: pp. 83-93, all of which are incorporated herein by reference in their entireties. In some embodiments, other surface generation algorithms can be used.
[0026]
[0030] The triangular mesh representation of the cortex obtained by the marching cubes can then be remeshed into a high-resolution triangular mesh with triangles having uniform edge lengths, and another simplified triangular mesh obtained with a topology-preserving mesh decimation algorithm. In other words, the method 26 further comprises steps of generating a higher-resolution triangular mesh (e.g., see higher-resolution triangular mesh 36 in FIG. 3 ) based on the triangular mesh representation, where the triangles of the higher-resolution triangular mesh have uniform edge lengths (e.g., 1 mm edges) 38, and generating a simplified triangular mesh (e.g., see simplified triangular mesh 40 in FIG. 3 ) using a mesh decimation algorithm, where the cortical folding and geometry are maintained (e.g., topology-preserving) 42. That is, the simplified mesh is obtained using a topology-preserving mesh decimation that preserves the cortical folding and geometry. For example, the mesh decimation algorithm may be accomplished using open source software performing 3D computer graphics, image processing and / or visualization, including the visualization toolkit in "Mesh Decimation Using VTK" by Knapp, M., 2002, Institute of Computer Graphics and Algorithms, Vienna University of Technology, and further edited by Shao Yiyang, "Allen-Shao / Mesh Simplification Wiki - Github," all of which are incorporated herein by reference in their entireties.
[0027]
[0031] At step 44, the method 26 further includes generating a dipole patch using the simplified triangular mesh, as illustrated by cortical dipole patch 46 in Figure 3. That is, the dipole patch is obtained by using the simplified mesh as a template.
[0028]
[0032] In step 48, the method 26 associates triangles in the higher resolution triangular mesh with each nearest triangle in the simplified triangular mesh. Every triangle from the high resolution uniform edge length mesh can be assigned to the nearest triangle from the simplified mesh. In this way, the final number of dipole patches is equal to the number of triangles in the simplified mesh.
[0029]
[0033] In some embodiments, each dipole patch must have a patch center and a patch normal to define its orientation. In this regard, the method 26 further comprises determining 50 a patch center for each dipole patch. In one embodiment, the method 26 determines the patch center by finding the centroids or centers of mass of all high resolution mesh triangles associated with (e.g., assigned to) each dipole patch, and defining that the center of the high resolution mesh triangle closest to the centroid or center of mass of a particular dipole patch may be selected as the patch center of the corresponding dipole patch.
[0030]
[0034] At step 52, the method 26 further includes determining a patch normal for each dipole patch by calculating the average of the normals of all high resolution triangles associated with (eg, assigned to) each corresponding dipole patch.
[0031]
[0035] The patch normals, patch centers and / or tessellation are then used in a model to determine the location of cortical sources based on electrical potentials measured on the subject's scalp. The resulting dipole patches and patch normals are shown in image 54 of FIG.
[0032]
[0036] It should be noted that in some embodiments, variations to the number and / or order of steps in the above method 26, or to the substance of the steps, are contemplated within the scope of this disclosure.
[0033]
[0037] Having described an embodiment of the cortical surface tessellation method 26 (FIG. 2), attention is directed to FIG. 5, which illustrates an embodiment of an exemplary device 56 for performing the cortical surface tessellation method 26. The device 56 may be the same as or similar to the computing device 18 (FIG. 1). In the illustrated embodiment, the functionality of the cortical surface tessellation system performing the method 26 is embodied as a device having co-located software and hardware elements collectively embodied as a computing device (which may include a medical device, a personal computer, a tablet, etc.). It should be understood that in some embodiments, the functionality of the cortical surface tessellation system may be performed in one or more devices that are local to a clinical, imaging machine or system, or that are remote from the imaging machine / system (in a cloud-based platform, server farm, web server, application server, etc.). In some embodiments, multiple devices that are remote from each other (e.g., in a client-server relationship) may collectively perform the functionality of the cortical surface tessellation system in a distributed manner. Those skilled in the art should understand that, in the context of this disclosure, the exemplary device referred to hereinafter as computing device 56 is merely illustrative of one embodiment, and that some embodiments of a computing device may include fewer or additional components, and in some embodiments, some of the functionality associated with the various components illustrated in Figure 5 may be combined or further distributed among additional modules or computing devices. It should be understood that certain well-known elements of computer systems have been omitted herein to avoid obscuring the relevant functionality of computing device 56.
[0034]
[0038] In one embodiment, the computing device 56 has one or more processors (P) 58 (e.g., 58A, ..., 58N), an input / output (I / O) interface 60, one or more user interfaces (UI) 62 (which may include one or more of a keyboard, mouse, microphone, speaker, display, etc.), and memory (MEM) 64, all coupled to one or more data buses, such as a data bus (DBUS) 66. In some embodiments, the user interface 62 may be directly coupled to the data bus 66. The user interface 62 may include the user interface 22 (FIG. 1). The memory 64 may include any one or combination of volatile memory elements (e.g., random access memory RAM, such as DRAM and SRAM) and non-volatile memory elements (e.g., ROM, flash, solid state, EPROM, EEPROM, hard drive, tape, CD-ROM, etc.). The memory 64 may store a native operating system, one or more native applications, an emulation system, or emulated applications for any of the various operating systems and / or emulated hardware platforms, emulated operating systems, etc. In some embodiments, a separate storage device (STOR DEV) may be coupled to the data bus 66 or as a networked device (or devices) via the I / O interface 60 and one or more networks. In some embodiments, the storage device may function as the data repository 24 (FIG. 1). In the illustrated embodiment, the computing device 56 may be coupled to the imaging system (e.g., the MRI system 20) and the processing device 16 (FIG. 1) via the I / O interface 60, although it should be understood that in some embodiments, the coupling may be accomplished via one or more networks or according to other known connections or interconnections. The storage device may be embodied as a persistent memory (e.g., optical, magnetic, and / or semiconductor memory and associated drives).In some embodiments, the storage or memory 64 may store a model bank, one or more bitmasks, scans, meshes, among other primary information. The memory 64 and / or the storage may each be considered a non-transitory computer-readable storage medium.
[0035]
[0039] In the embodiment shown in FIG. 5, memory 64 includes an operating system (OS) 68 (e.g., LINUX, macOS, Windows, etc.) and electrical source imaging (ESI) software 70 including functionality for performing electrical source imaging and support functions. In one embodiment, ESI software 70 includes cortical surface tessellation (CST) software 72. Cortical surface tessellation software 72 includes one or more modules (executable code, or generally instructions) corresponding to the functionality of cortical surface tessellation method 26 (FIG. 2) shown in and described in connection with the method. That is, the functionality of the cortical surface tessellation software 72 has been described in connection with the cortical surface tessellation method 26 shown in FIG. 2, and thus a description of the software is omitted here for the sake of brevity. Note that references to software may include software, firmware, middleware, and / or microcode or microinstructions. In some embodiments, the functionality of the software may be implemented through hardware (e.g., circuitry including an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a vector processor, a tensor processing unit, etc.). Memory 64 further includes a communications module (COMM) 74 that includes software configured to enable communication of information (via I / O interface 60) between other systems and / or devices (e.g., of environment 10 of FIG. 1).
[0036]
[0040] Execution of the cortical surface tessellation software 72 (and ESI software 70) may be performed by one or more processors 58 under the direction and / or control of the operating system 68. The processor 58 may be embodied as a custom-made or commercially available processor, including a single-core or multi-core central processing unit (CPU), a tensor processing unit (TPU), a graphics processing unit (GPU), a vector processing unit (VPU), or co-processors among several processors, a semiconductor-based microprocessor (in the form of a microchip), a microprocessor, one or more application-specific integrated circuits (ASICs), a field programmable gate array (FPGU), a plurality of appropriately configured digital logic gates, and / or any other known electrical configuration including elements both individually and in various combinations for coordinating the overall operation of the computing device 56.
[0037]
[0041] I / O interface 60 comprises hardware and / or software for providing one or more interfaces to other systems or devices, including data repository 24 (FIG. 1). I / O interface 60 may include cable, wireless and / or cellular modems and / or may establish communications with other devices or systems using one or more of a variety of communication protocols (e.g., TCP / IP, UDP, etc.) over an Ethernet connection, hybrid / fiber coaxial (HFC), copper cable (e.g., digital subscriber line (DSL), asymmetric DSL, etc.). Typically, the I / O interface 60 comprises appropriate hardware to cooperate with the communications module 74 to enable communication of information using TCP / IP, UDP, HTTP, DSL, via PSTN (Public Switched Telephone Network), POTS, Integrated Services Digital Network (ISDN), Ethernet, Fiber, DSL / ADSL, Wi-Fi, Cellular (e.g., 3G, 4G, 5G, Global System for Mobile Communications (GSM), General Packet Radio Service (GPRS), etc.), Bluetooth, Near Field Communication (NFC), ZigBee, among others.
[0038]
[0042] The user interface 62 may include a keyboard, mouse, microphone, display, immersive headset, etc., that allow for input and / or output by a clinician, technician, or other user. For example, the user interface 62 may visually represent results obtained by the method 26 (e.g., locations of patch centers, patch normals, etc., as shown in FIG. 4 ) and / or may visually represent one or more cortical sources. In some embodiments, the user interface 62 may be integrated into the computing device 56, and in some embodiments, may be coupled to the computing device 56 via one or more wired and / or wireless networks. In some embodiments, the user interface 62 may cooperate with associated software to enable augmented or virtual reality, or visualization may be achieved in conjunction with other devices via the I / O interface 60.
[0039]
[0043] It should be noted that, if a particular embodiment of the computing device 56 is embodied at least in part in software (including firmware, middleware, microcode, etc.), the cortical surface tessellation software 72 (and the ESI software 70) can be stored on any of a variety of non-transitory computer-readable (storage) media for use by or in connection with a variety of computer-related systems or methods. In the context of this specification, a computer-readable storage medium can have an electronic, magnetic, optical, or other physical device or apparatus that can store or store computer programs (e.g., executable code or instructions) for use by or in connection with a computer-related system or method. The software can be embedded in a variety of computer-readable storage media for use by or in connection with an instruction execution system, device, or apparatus, such as a computer-based system, a system that includes a processor, or other system that can retrieve instructions from and execute those instructions.
[0040]
[0044] When a particular embodiment of computing device 56 is implemented at least in part in hardware, such functionality may be implemented using any or a combination of the following technologies, all of which are well known in the art: discrete logic circuits having logic gates for performing logical functions on data signals, application specific integrated circuits (ASICs) with appropriate combinatorial logic gates, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), TPUs, GPUs and / or other accelerators / co-processors, etc.
[0041]
[0045] In view of the above description, it should be understood that one embodiment of an exemplary cortical surface tessellation method, depicted in FIG. 6 and denoted as method 76 delimited by a start and an end, comprises receiving magnetic resonance imaging data 78; segmenting the magnetic resonance imaging data 80, the segmentation including a tissue segmentation of a cortical surface of the cortex; tessellating the segmented cortical surface with a plurality of dipole patches, the amount of which is variable and dependent on head dimensions 82; and providing an image of the head or a portion thereof according to the tessellation for display 84.
[0042]
[0046] It should be noted that methods 26 (FIG. 2) and 76 may be performed by device 56, or in some embodiments, by multiple devices.
[0043]
[0047] Any process description or block in a flowchart should be understood to represent a module, segment, or portion of code that includes one or more executable instructions for implementing a particular logical function or step in the process, and the scope of the embodiments also includes alternative configurations in which functions are performed out of the order shown or described, including substantially simultaneously or in reverse order depending on the functionality involved, as will be understood by those skilled in the art of this disclosure. In some embodiments, one or more steps can be omitted or additional steps can be added.
[0044]
[0048] Embodiments of the cortical surface tessellation methods, devices, and systems described herein may improve the accuracy of electrical source imaging in a clinically feasible manner, which may ultimately lead to more accurate, patient-specific diagnoses and better outcomes.
[0045]
[0049] In one embodiment, a method is disclosed comprising the steps of receiving magnetic resonance imaging data; segmenting the magnetic resonance imaging data, the segmentation comprising a tissue segmentation of a cortical surface of the cortex; tessellating the segmented cortical surface with a plurality of dipole patches, the quantity of the plurality of dipole patches being variable and dependent on head dimensions; and providing an image of the head or a portion thereof according to the tessellation for display.
[0046]
[0050] In one embodiment, the method comprises the cortical surface comprising a grey matter surface of the cortex.
[0047]
[0051] In one embodiment, any one of the methods further comprises the steps of receiving a volumetric bit mask of the cortex, where each voxel of a plurality of voxels is labeled according to tissue type; providing a triangular mesh representation of the segmented cortical surface based on the volumetric bit mask; re-meshing the triangular mesh representation to provide a high resolution triangular mesh comprising triangles of uniform edge length; and applying a topology-preserving mesh decimation algorithm to the high resolution triangular mesh representation to provide a simplified triangular mesh representation.
[0048]
[0052] In one embodiment, any one of the methods is wherein the triangular mesh representation of the segmented cortical surface is based on applying a marching cubes algorithm to the bitmask.
[0049]
[0053] In one embodiment, any one of the methods is wherein the topology-preserving mesh decimation algorithm preserves cortical folding and geometry of the segmented cortex.
[0050]
[0054] In one embodiment, any one of the methods further comprises the step of obtaining the plurality of dipole patches by associating each triangle from the high resolution triangular mesh representation with its nearest triangle from the simplified triangular mesh representation, such that a final quantity of the plurality of dipole patches is equal to the number of triangles in the simplified triangular mesh representation.
[0051]
[0055] In an embodiment, any one of the methods has the property that each of the plurality of dipole patches is used as a boundary condition in a forward model used for electrical source imaging.
[0052]
[0056] In one embodiment, any one of the methods, wherein the properties include a patch center, a patch normal, and a surface triangle associated with each of the plurality of dipole patches.
[0053]
[0057] In one embodiment, any one of the methods further comprises determining each dipole patch center by determining the center of mass of all triangles of the high resolution triangular mesh representation associated with a particular one of the closest triangles, and the center of the high resolution mesh triangle closest to the center of mass is selected as the center of the dipole patch, the normal of which corresponds to the average of all normals of the triangles of the high resolution triangular mesh representation.
[0054]
[0058] In one embodiment, any one of the methods further comprises the steps of: aligning electrode positions on the head; predicting electrical activity of the head based on the tessellated images; recording electroencephalographic signals from the head; and predicting a source of target brain activity based on the recordings.
[0055]
[0059] In one embodiment, any one of the methods further comprises displaying an image of the head or part thereof according to the tessellation, the displayed image further comprising visual representations of one or any combination of patch centers, patch normals and cortical sources.
[0056]
[0060] In one embodiment, an apparatus is disclosed having a memory having instructions; and one or more processors configured to perform functions corresponding to any one of the methods according to the instructions.
[0057]
[0061] In one embodiment, a system is disclosed having the device, the system further comprising: a magnetic resonance imaging system configured to scan a subject's head and provide magnetic resonance imaging data to one or more processors of the device; and a plurality of EEG electrodes configured to sense EEG signals, the one or more processors of the device being configured to receive the EEG signals measured by the EEG electrodes positioned on the subject's scalp.
[0058]
[0062] In one embodiment, the system further comprises a display device coupled to the device, the display device configured to display an image of the head or part thereof according to the tessellation.
[0059]
[0063] In one embodiment, any one of the systems is such that the displayed image further comprises visual representations of one or any combination of patch centers, patch normals, and cortical sources.
[0060]
[0064] While the present invention has been illustrated and described in detail in the drawings and the foregoing description, such illustrations and descriptions are to be considered as illustrative or exemplary and not restrictive. That is, the present invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art from a study of the drawings, the disclosure and the appended claims in carrying out the claimed invention. It should be noted that various combinations of the disclosed embodiments can be used, and thus a reference to an embodiment or an embodiment does not mean to exclude the features of that embodiment from being used with the features of other embodiments. In the claims, the word "comprises" does not exclude other elements or steps, and the singular does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain means are recited in mutually different dependent claims does not indicate that a combination of these means cannot be used to advantage. A computer program can be stored / distributed on an appropriate medium, such as an optical medium or a solid medium, provided together with or as part of other hardware, but can also be distributed in other forms. Any reference signs in the claims should not be construed as limiting the scope.
Claims
1. receiving magnetic resonance imaging data; segmenting the magnetic resonance imaging data, the segmentation comprising a tissue segmentation of a cortical surface of the cortex; tessellating the segmented cortical surface with a plurality of dipole patches, the quantity of which is variable and dependent on head dimensions; providing an image of the head or a portion of the head according to said tessellation for display; receiving a volumetric bit mask of the cortex, each voxel of a plurality of voxels being labeled according to tissue type; providing a triangular mesh representation of the segmented cortical surface based on the volumetric bitmask; remeshing the triangular mesh representation to provide a high resolution triangular mesh representation comprising triangles of uniform edge length; applying a topology-preserving mesh decimation algorithm to the high resolution triangular mesh representation to provide a simplified triangular mesh representation; obtaining the plurality of dipole patches by associating each triangle of the high resolution triangular mesh representation with a nearest triangle of the simplified triangular mesh representation, such that a final amount of the plurality of dipole patches is equal to an amount of triangles of the simplified triangular mesh representation; The method comprising:
2. The method of claim 1 , wherein the cortical surface comprises a gray matter surface of the cortex.
3. 3. The method of claim 1 or claim 2, wherein the triangular mesh representation of the segmented cortical surface is based on application of a marching cubes algorithm to a bitmask.
4. The method of claim 1 , wherein the topology-preserving mesh decimation algorithm preserves cortical folding and geometry of the segmented cortex.
5. The method of claim 1 , wherein each of the plurality of dipole patches has properties that are used as boundary conditions in a forward model used in electrical source imaging.
6. The method of claim 5 , wherein the properties include a patch center, a patch normal, and a surface triangle associated with each of the plurality of dipole patches.
7. 2. The method of claim 1, wherein each dipole patch center is determined by determining the center of mass of all triangles of the high resolution triangular mesh representation associated with a particular one of the closest triangles, the center of the high resolution mesh triangle closest to the center of mass being selected as the dipole patch center, and a normal of the dipole patch corresponds to an average of all normals of the triangles of the high resolution triangular mesh representation.
8. aligning electrode positions on the head; predicting electrical activity of the head based on the tessellated images; recording electroencephalographic signals from the head; predicting a source of the target's brain activity based on said recordings; The method of claim 1 , further comprising:
9. 9. The method of claim 1 , further comprising the step of displaying an image of the head or a portion of the head according to the tessellation, the displayed image further comprising visual representations of one or any combination of patch centers, patch normals and cortical sources.
10. a memory having instructions; One or more processors adapted to carry out the method according to any one of claims 1 to 9. An apparatus having the above configuration.
11. A system comprising the device of claim 10, a magnetic resonance imaging system configured to scan a subject's head and provide the magnetic resonance imaging data to the one or more processors of the apparatus; A plurality of EEG electrodes for detecting EEG signals; Further comprising: the one or more processors of the device receive the EEG signals measured by the EEG electrodes placed on the subject's scalp. system.
12. The system of claim 11 , further comprising a display device coupled to the device, the display device displaying an image of the head or a portion of the head according to the tessellation.
13. The system of claim 12 , wherein the displayed image further comprises a visual representation of one or any combination of patch centers, patch normals, and cortical sources.
Citation Information
Patent Citations
Brain's insular cortex activity extraction method
JP2018023586A
Computerised Cortex Boundary Extraction From Mr Images
US20080170791A1
Deep brain source imaging with m / EEG and anatomical MRI
WO2016073985A1
