Systems and methods for targeted neuromodulation

Abstract

To provide systems and methods for targeted neuromodulation.SOLUTION: Systems and methods for neuronavigation are illustrated according to embodiments of the present invention. The targeting systems and methods as described herein can generate personalized stimulation targets for the treatment of mental conditions. In many embodiments, direct stimulation of the personalized stimulation target indirectly affects brain structures that are more difficult to reach via the stimulation modality. In various embodiments, the psychiatric condition is major depressive disorder. In some embodiments, the psychiatric condition is suicidal ideation.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] (CROSS-REFERENCE TO RELATED APPLICATIONS) This application claims the benefit of and priority under 35 U.S.C. § 119(e) to U.S. Provisional Patent Application No. 63 / 090,680, filed October 12, 2020, entitled "Systems and Methods for Neuronavigation," the disclosure of which is incorporated herein by reference in its entirety for all purposes.

[0002] The present invention relates generally to neuromodulation therapy, and more particularly to generating personalized stimulation targets. [Background technology]

[0003] Brain stimulation therapy can be delivered in several ways, including (but not limited to) transcranial magnetic stimulation (TMS) and deep brain stimulation (DBS). Brain stimulation therapy is often delivered to or toward specific regions of a patient's brain to treat a patient's condition.

[0004] Radiation imaging allows for non-invasive scanning of internal organs. A common brain imaging technique involves the use of a magnetic resonance imaging (MRI) machine and a variant of MRI called functional MRI (fMRI), which is capable of measuring brain activity by measuring changes associated with blood flow. In contrast to fMRI, MRI is often referred to as "structural" because it examines only the anatomy of the brain and not brain activity. Summary of the Invention [Means for solving the problem]

[0005] Systems and methods for targeted neuromodulation according to embodiments of the present invention are illustrated. One embodiment includes a processor and a memory containing a targeting application, the targeting application acquiring patient brain data, the patient brain data comprising structural magnetic resonance imaging (sMRI) scans and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain, and mapping a reference region of interest (ROI) and at least one exploration ROI on the patient's brain using the sMRI scans and the at least one fMRI scan, the reference ROI describing a region to be indirectly affected via brain stimulation therapy, and the at least one exploration ROI describing a region to be directly targeted by the brain stimulation therapy. and a memory that describes at least one region and derives an individualized ROI parcelization map, the individualized ROI parcelization map describing a reference ROI as a plurality of reference parcels and describing at least one search ROI as a plurality of candidate parcels, extracting relationships between the plurality of candidate parcels and the plurality of reference parcels, calculating a target score for a candidate parcel in the plurality of candidate parcels based on the extracted relationships, selecting a target parcel from the plurality of candidate parcels based on the target score, and instructing a processor to provide the target parcel.

[0006] In another embodiment, the targeting application further instructs the processor to provide brain stimulation therapy to the target parcel to treat a psychiatric condition of the patient.

[0007] In a further embodiment, the psychiatric condition is major depressive disorder.

[0008] In still a further embodiment, the psychiatric condition is suicidal ideation.

[0009] In still further embodiments, the brain stimulation therapy is selected from the group consisting of transcranial magnetic stimulation, transcranial direct current stimulation, and electrical stimulation delivered via an implantable electrical stimulator.

[0010] In yet another embodiment, the targeting application further instructs the processor to discard fMRI scans that deviate from expected whole-brain network connectivity.

[0011] In still further embodiments, to derive the individualized map of ROI parcelization, the targeting application further instructs the processor to randomly subsample voxels within the reference and at least one search ROI, cluster the subsample of voxels, record the clustering assignments, and label the clusters in the clustering assignments as candidate parcels or reference parcels based on their location.

[0012] In another additional embodiment, to derive the individualized map of the ROI parcelization, the targeting application further instructs the processor to randomly subsample voxels within the reference and at least one search ROI as a first subsample of voxels, cluster the first subsample of voxels, record the first clustering assignment, randomly subsample voxels within the reference and at least one search ROI as a second subsample of voxels, cluster the second subsample of voxels, record the second clustering assignment, merge the first clustering assignment and the second clustering assignment using consensus clustering, and label the clusters in the merged clustering assignment as candidate parcels or reference parcels based on their location.

[0013] In yet an additional embodiment, to derive the individualized map of ROI parcelization, the targeting application further instructs the processor to divide the spatially disjoint clusters.

[0014] Again, in another embodiment, the target score is calculated based on at least one factor from the group consisting of parcel size, parcel depth, parcel shape, parcel homogeneity, functional connectivity strength relative to a reference ROI, and network connectivity score.

[0015] Again, in a further embodiment, the network connectivity score reflects the anti-correlation between the default mode network and the dorsal attention network of the patient's brain.

[0016] In yet another embodiment, a method of targeted neuromodulation includes acquiring patient brain data, the patient brain data including structural magnetic resonance imaging (sMRI) scans and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain, and mapping a reference region of interest (ROI) and at least one exploration ROI on the patient's brain using the sMRI scans and the at least one fMRI scan, the reference ROI describing a region to be indirectly affected via brain stimulation therapy and the at least one exploration ROI describing at least one region to be directly targeted by brain stimulation therapy. and deriving an ROI parcelization individualization map, wherein the ROI parcelization individualization map describes the reference ROI as a plurality of reference parcels and describes the at least one search ROI as a plurality of candidate parcels, extracting relationships between the plurality of candidate parcels and the plurality of reference parcels, calculating target scores for the candidate parcels in the plurality of candidate parcels based on the extracted relationships, selecting a target parcel from the plurality of candidate parcels based on the target scores, and providing the target parcel. In many embodiments, obtaining the patient brain data may be accomplished by accessing patient brain data previously uploaded or transmitted to the target identification system, requesting the patient brain data from a remote institution, computer system, or database, or by accessing hardware, such as MRI or other imaging hardware, to cause acquisition of the patient brain data.

[0017] In still yet a further embodiment, the method further comprises providing brain stimulation therapy to the target parcel to treat the patient's psychiatric condition.

[0018] In yet another additional embodiment, the psychiatric condition is major depressive disorder.

[0019] In still further additional embodiments, the psychiatric condition is suicidal ideation.

[0020] Again, in yet another embodiment, the brain stimulation therapy is selected from the group consisting of transcranial magnetic stimulation, transcranial direct current stimulation, and electrical stimulation delivered via an implantable electrical stimulator.

[0021] Again, in still a further embodiment, the method further comprises discarding fMRI scans that deviate from expected whole-brain network connectivity.

[0022] In yet another additional embodiment, deriving the individualized map of ROI parcelization includes randomly subsampling voxels within the reference and at least one search ROI, clustering the subsample of voxels, recording the clustering assignments, and labeling clusters in the clustering assignments as candidate parcels or reference parcels based on their location.

[0023] In yet a further additional embodiment, deriving the individualized map of ROI parcelization includes randomly subsampling voxels within the reference and at least one search ROI as a first subsample of voxels, clustering the first subsample of voxels, and recording the first clustering assignment, randomly subsampling voxels within the reference and at least one search ROI as a second subsample of voxels, clustering the second subsample of voxels, and recording the second clustering assignment, merging the first clustering assignment and the second clustering assignment using consensus clustering, and labeling clusters in the merged clustering assignment as candidate parcels or reference parcels based on their location.

[0024] Again, in yet another embodiment, deriving the individualization map of the ROI parcelization includes partitioning spatially disjoint clusters.

[0025] Again, in yet a further embodiment, the target score is calculated based on at least one factor from the group consisting of parcel size, parcel depth, parcel shape, parcel homogeneity, functional connectivity strength relative to a reference ROI, and network connectivity score.

[0026] Again, in another additional embodiment, the network connectivity score reflects the anti-correlation between the default mode network and the dorsal attention network of the patient's brain.

[0027] Again, in yet a further additional embodiment, a system for treating major depressive disorder includes a transcranial magnetic stimulation device, a neuronavigation device, a processor, and a memory containing a targeting application, the targeting application acquiring patient brain data, the patient brain data comprising structural magnetic resonance imaging (sMRI) scans and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain, and using the sMRI scans and the at least one fMRI scan, mapping a reference region of interest (ROI) and at least one exploration ROI on the patient's brain, the reference ROI describing a region to be indirectly affected via the transcranial magnetic stimulation device, and the at least one exploration ROI describing a region to be indirectly affected via brain stimulation therapy. and a memory that describes at least one region to be directly targeted; derives an ROI parcelization individualization map, the ROI parcelization individualization map describing a reference ROI as a plurality of reference parcels and describing at least one search ROI as a plurality of candidate parcels; extracts relationships between the plurality of candidate parcels and the plurality of reference parcels; calculates a target score for the candidate parcel in the plurality of candidate parcels based on the extracted relationships; selects a target parcel from the plurality of candidate parcels based on the target score; and instructs the processor to apply transcranial magnetic stimulation to the target parcel using a transcranial magnetic stimulation device and / or a neuronavigation device to treat major depressive disorder.

[0028] Again, in yet another additional embodiment, the target parcel is transmitted to the neuronavigation system from a cloud computing platform.

[0029] In yet another additional embodiment, a method of treating major depressive disorder includes acquiring patient brain data, the patient brain data comprising structural magnetic resonance imaging (sMRI) scans and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain; mapping a reference region of interest (ROI) and at least one exploration ROI on the patient's brain using the sMRI scans and the at least one fMRI scan, the reference ROI describing a region to be indirectly affected via brain stimulation therapy and the at least one exploration ROI describing at least one region to be directly targeted by brain stimulation therapy; and individualizing ROI parcelization. deriving a map, wherein the individualized map of ROI parcelization describes a reference ROI as a plurality of reference parcels and describes at least one search ROI as a plurality of candidate parcels; extracting relationships between the plurality of candidate parcels and the plurality of reference parcels; calculating a target score for a candidate parcel in the plurality of candidate parcels based on the extracted relationships; selecting a target parcel from the plurality of candidate parcels based on the target score; and treating major depressive disorder by applying transcranial magnetic stimulation to the target parcel using a transcranial magnetic stimulation device and / or a neuronavigation device.

[0030] In yet another additional embodiment, the transcranial magnetic stimulation is accelerated theta burst stimulation.

[0031] Additional embodiments and features are set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of this specification or may be learned by practice of the invention. A further understanding of the nature and advantages of the present invention may be realized by reference to the remaining portions of the specification and the drawings, which form a part of this disclosure. The present invention provides, for example, the following. (Item 1) 1. A targeted neuromodulation system comprising: a processor; A memory containing a targeted application, the targeted application comprising: acquiring patient brain data, the patient brain data comprising a structural magnetic resonance imaging (sMRI) scan and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain; mapping a reference region of interest (ROI) and at least one search ROI in the patient's brain using the sMRI scan and the at least one fMRI scan; the reference ROI describes a region to be indirectly affected via brain stimulation therapy; the at least one search ROI describes at least one region to be directly targeted by the brain stimulation therapy; and deriving an ROI parcelization individualization map, wherein the ROI parcelization individualization map describes the reference ROI as a plurality of reference parcels and describes the at least one search ROI as a plurality of candidate parcels; extracting relationships between the plurality of candidate parcels and the plurality of reference parcels; calculating a target score for a candidate parcel in the plurality of candidate parcels based on the extracted relationships; selecting a target parcel from the plurality of candidate parcels based on the target score; providing said target parcel; a memory for instructing the processor to A targeted neuromodulation system comprising: (Item 2) 2. The targeted neuromodulation system of item 1, wherein the targeting application further instructs the processor to provide the brain stimulation therapy to the target parcel to treat a mental condition of the patient. (Item 3) 3. The targeted neuromodulation system of item 2, wherein the psychiatric condition is major depressive disorder. (Item 4) 3. The targeted neuromodulation system of item 2, wherein the brain stimulation therapy is selected from the group consisting of transcranial magnetic stimulation, transcranial direct current stimulation, and electrical stimulation delivered via an implantable electrical stimulation device. (Item 5) 2. The targeted neuromodulation system of item 1, wherein the targeting application further instructs the processor to discard fMRI scans that deviate from expected whole-brain network connectivity. (Item 6) To derive the ROI parcelization individualization map, the targeting application further comprises: randomly subsampling voxels within the reference and at least one search ROI; clustering a sub-sample of said voxels; recording the clustering assignments; labeling clusters in said clustering assignment as candidate parcels or reference parcels based on their locations; 2. The targeted neuromodulation system of item 1, instructing the processor to: (Item 7) To derive the ROI parcelization individualization map, the targeting application further comprises: randomly subsampling voxels within the reference and at least one search ROI as a first subsample of voxels; clustering a first sub-sample of said voxels; recording the first clustering assignment; randomly subsampling voxels within the reference and at least one search ROI as a second subsample of voxels; clustering a second sub-sample of said voxels; recording the second clustering assignment; and merging the first clustering assignment and the second clustering assignment using consensus clustering; and labeling clusters in said merged clustering assignment as candidate parcels or reference parcels based on their locations; 2. The targeted neuromodulation system of item 1, instructing the processor to: (Item 8) 2. The targeted neuromodulation system of item 1, wherein the targeting application further instructs the processor to divide spatially disjoint clusters to derive the individualized map of ROI parcelization. (Item 9) Item 10. The targeted neuromodulation system of item 1, wherein the target score is calculated based on at least one factor from the group consisting of parcel size, parcel depth, parcel shape, parcel homogeneity, functional connectivity strength relative to the reference ROI, and network connectivity score. (Item 10) 10. The targeted neuromodulation system of item 9, wherein the network connectivity score reflects an anti-correlation between the default mode network and the dorsal attention network of the patient's brain. (Item 11) 2. The targeted neuromodulation system of item 1, wherein the target parcel is transmitted from a cloud computing platform to a neuronavigation system. (Item 12) 1. A method of targeted neuromodulation, comprising: acquiring patient brain data, the patient brain data comprising a structural magnetic resonance imaging (sMRI) scan and at least one functional magnetic resonance imaging (fMRI) scan of the patient's brain; mapping a reference region of interest (ROI) and at least one search ROI in the patient's brain using the sMRI scan and the at least one fMRI scan; the reference ROI describes a region to be indirectly affected via brain stimulation therapy; the at least one search ROI describes at least one region to be directly targeted by the brain stimulation therapy; and deriving an ROI parcelization individualization map, wherein the ROI parcelization individualization map describes the reference ROI as a plurality of reference parcels and describes the at least one search ROI as a plurality of candidate parcels; extracting relationships between the plurality of candidate parcels and the plurality of reference parcels; calculating a target score for a candidate parcel in the plurality of candidate parcels based on the extracted relationships; selecting a target parcel from the plurality of candidate parcels based on the target score; providing said target parcel; A method of targeted neuromodulation comprising: (Item 13) 13. The method of targeted neuromodulation described in item 12, further comprising providing the brain stimulation therapy to the target parcel to treat a psychiatric condition of the patient. (Item 14) Item 14. The targeted neuromodulation method of item 13, wherein the psychiatric condition is major depressive disorder. (Item 15) 14. The method of targeted neuromodulation according to item 13, wherein the brain stimulation therapy is selected from the group consisting of transcranial magnetic stimulation, transcranial direct current stimulation, and electrical stimulation delivered via an implantable electrical stimulation device. (Item 16) 13. The targeted neuromodulation method of item 12, further comprising discarding fMRI scans that deviate from expected whole-brain network connectivity. (Item 17) deriving the ROI parcelization individualization map, randomly subsampling voxels within the reference and at least one search ROI; clustering a sub-sample of said voxels; recording the clustering assignments; labeling clusters in said clustering assignment as candidate parcels or reference parcels based on their locations; Item 13. The targeted neuromodulation method of item 12, comprising: (Item 18) deriving the ROI parcelization individualization map, randomly subsampling voxels within the reference and at least one search ROI as a first subsample of voxels; clustering a first sub-sample of said voxels; recording the first clustering assignment; randomly subsampling voxels within the reference and at least one search ROI as a second subsample of voxels; clustering a second sub-sample of said voxels; recording the second clustering assignment; and merging the first clustering assignment and the second clustering assignment using consensus clustering; and labeling clusters in said merged clustering assignment as candidate parcels or reference parcels based on their locations; Item 13. The targeted neuromodulation method of item 12, comprising: (Item 19) Item 13. The targeted neuromodulation method of item 12, wherein deriving the individualized map of ROI parcelization comprises dividing spatially disjoint clusters. (Item 20) Item 13. The targeted neuromodulation method of item 12, wherein the target score is calculated based on at least one factor from the group consisting of parcel size, parcel depth, parcel shape, parcel homogeneity, functional connectivity strength relative to the reference ROI, and network connectivity score. (Item 21) Item 13. The targeted neuromodulation method of item 12, wherein the network connectivity score reflects the anti-correlation between the default mode network and the dorsal attention network of the patient's brain. (Item 22) Item 13. The targeted neuromodulation method of item 12, wherein the target parcel is transmitted from a cloud computing platform to a neuronavigation system. [Brief explanation of the drawings]

[0032] The description and claims are more fully understood with reference to the following figures and data graphs, which are presented as exemplary embodiments of the invention and should not be construed as a complete recitation of the scope of the invention.

[0033] [Figure 1] FIG. 1 illustrates a targeted neuromodulation system according to one embodiment of the present invention.

[0034] [Figure 2] FIG. 2 illustrates a target generator according to an embodiment of the present invention.

[0035] [Figure 3] FIG. 3 is a flow chart illustrating a targeting process for generating personalized targets according to an embodiment of the present invention.

[0036] [Figure 4] FIG. 4 is a flowchart illustrating a targeting process for assessing expected network connectivity according to an embodiment of the present invention.

[0037] [Figure 5] FIG. 5 is a flowchart illustrating a targeting process for deriving personalized ROI parcelization, according to an embodiment of the present invention.

[0038] [Figure 6] FIG. 6 is a flowchart illustrating a targeted process for dividing spatially disjoint clusters according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0039] Detailed Description Mental health conditions and other neurological problems are significant medical fields with profound implications for both patients and society at large. Depression and suicidal ideation, for example, represent chronic public health problems. However, treatment for these conditions has traditionally been addressed using pharmaceuticals and, in some treatment-resistant cases, surgery and / or electroconvulsive therapy (ECT). These methods can have significant side effects, both mental and physical. In contrast, a form of therapy called transcranial magnetic stimulation (TMS) has emerged as a viable, non-invasive treatment option with few reported side effects.

[0040] TMS involves applying a magnetic field to a specific region of the brain to depolarize or hyperpolarize neurons in the target region. Typically, a target region is selected by a medical professional based on its relationship to the patient's condition. For example, the dorsolateral prefrontal cortex (DLPFC) is known to be involved in major depressive disorder. However, the exact location of the DLPFC in an individual can be difficult to manually identify. Even when this can be identified, there may actually be a specific subregion of the DLPFC that would be the most effective target for an individual patient based on their unique brain. Furthermore, there may even be other regions in the brain that would provide a better stimulation target for the patient. Because every brain is at least slightly different, a personalized method of generating stimulation targets for each individual can provide better therapy outcomes.

[0041] An additional limitation of many TMS devices is the depth to which they can induce electrical currents within a patient's brain. In many cases, TMS devices cannot target deep brain structures. However, numerous large-scale networks throughout the brain have been identified. For example, the default mode network (DMN) is a network thought to be involved in numerous tasks, such as waking rest. As a further example, the dorsal attention network (DAN) is thought to be important in the voluntary orientation of visuospatial attention, and similarly, the ventral attention network (VAN) reorients attention toward salient stimuli. The connectivity between different regions of the brain offers opportunities for TMS and other brain stimulation therapies, whereby more superficial brain structures that are strongly connected to deeper brain structures can be stimulated to bring about changes in deeper brain regions. Furthermore, stimulation of connected networks can have significant effects on structures within or otherwise connected to the network. In particular, several networks, such as (but not limited to) the DMN, DAN, and VAN, have specific experimentally determined relationships to major depressive disorder and suicidal ideation. Networks with implications for the particular mental condition to be treated can be given additional priority.

[0042] Given the complex nature of the brain, when applying neuromodulation therapies (such as TMS), the location at which stimulation is delivered can have a significant impact on treatment outcomes. Targeting, as discussed herein, refers to the process of identifying target structures within a patient's brain for stimulation to treat mental health conditions. While current targeting methods can yield viable targets, many conventional methods have significant drawbacks. For example, targeting is often performed using a single scan from the patient and cannot incorporate multiple scans over time. Due to fMRI scanning noise and limited test-retest reliability, deriving targets based on a single scan is more likely to be affected by noise, leading to impaired levels of target reliability. Reliability limitations can be even more pronounced for methods employing voxel clustering for target detection, especially when the clustering procedure is highly sensitive to noise and signal loss. Furthermore, clustering procedures used for this purpose do not necessarily consider the spatial relationships between voxels, which can lead to impractical results. Turning now to the drawings, the systems and methods described herein seek to address these limitations and provide a more robust targeting framework that generates more effective, personalized stimulation targets for more effective treatments. In many embodiments, targets generated using the systems and methods described herein are subsequently used as targets in neuromodulation therapies such as (but not limited to) TMS, transcranial direct current stimulation (tDCS), as implantation sites for one or more stimulation electrodes, and / or as targets for any number of different neuromodulation modalities as appropriate to the requirements of a particular application of an embodiment of the invention. Targeting systems according to embodiments of the invention are discussed below.

[0043] (Targeted Neuromodulation System) The targeted neuromodulation system is capable of obtaining and / or accessing a scan of a patient's brain and identifying one or more personalized targets for brain stimulation therapy. In many embodiments, the targeting system may be integrated with other medical devices, such as (but not limited to) a TMS device or a neuronavigation device. In various embodiments, the targeting system not only generates personalized targets but also includes or is integrated with a neuronavigation device and can identify where a TMS coil should be placed to properly stimulate the targets. In many embodiments, the targeted neuromodulation system can further apply neuromodulation to the generated targets via a neuromodulation device, such as (but not limited to) a TMS device, a tDCS device, an implantable neurostimulator, and / or any other neurostimulation device as appropriate for the requirements of a particular application of embodiments of the present invention.

[0044] Turning now to Figure 1, a targeted neuromodulation system is illustrated, in accordance with an embodiment of the present invention. The targeted neuromodulation system 100 includes a target generator 110. The target generator can be implemented using any number of different computing platforms, such as (but not limited to) desktop computers, laptops, server computers and / or clusters, smartphones, tablet PCs, and / or any other computing platform capable of executing logical instructions as appropriate to the requirements of a particular application of an embodiment of the present invention. In many embodiments, the target generator determines personalized and / or partially personalized targets within an individual's brain.

[0045] Targeted neuromodulation system 100 further includes an fMRI machine 120 and a TMS device 130. In many embodiments, the fMRI machine is capable of acquiring both structural and functional MRI images of the patient. TMS device 130 is capable of delivering brain stimulation therapy to targets selected by target generator 110. However, as can be readily appreciated, alternative imaging modalities (e.g., computed tomography, positron emission tomography, electroencephalography, etc.) and alternative brain stimulation devices (e.g., implantable stimulators) can be used as appropriate to the requirements of a particular application of embodiments of the present invention; i.e., alternatively, targeting system 100 may not include its own imaging equipment and may receive imaging or other brain data from one or more imaging systems distinct from neuromodulation system 100.

[0046] In many embodiments, the targeted neuromodulation system 100 includes a neuronavigation device that guides the delivery of brain stimulation therapy by the TMS device 130 to targets selected by the target generator 110. This neuronavigation device may be integrated into the targeting generator 110 or may be separate from the targeting system 110 (not shown). In many embodiments, the neuronavigation device assists in delivering brain stimulation therapy to one or more targets generated by the targeting system, for example, by determining the rotational and translational positions of the stimulation coil and head and displaying images to guide the user to correctly position the stimulation coil, or additionally, by using a mechanical actuator, such as a robotic arm, to correctly position the stimulation coil. As can be readily appreciated, the specific functionality of the neuronavigation device can vary depending on the type of neuromodulation being applied.

[0047] In many embodiments, the fMRI, TMS device, targeting system, and / or neuronavigation device are connected via a network 140. The network may be a wired network, a wireless network, or any combination thereof. Indeed, any number of different networks may be combined to connect the components. However, it is not a requirement that all components of the system communicate via a network. The target generator may function without operational connections between other components. Indeed, as can be readily appreciated, although a specific targeted neuromodulation system is illustrated in FIG. 1, any number of different system architectures may be used without departing from the scope or spirit of the present invention. For example, in many embodiments, the targeted neuromodulation system may include different neuromodulation devices providing different stimulation modalities.

[0048] Once the targeting systems are provided with patient brain data, they are able to generate personalized targets. Turning now to FIG. 2, a target generator architecture is illustrated, according to an embodiment of the present invention. Target generator 200 includes a processor 210. However, in many embodiments, more than one processor can be used. In various embodiments, the processor can be made of any logic processing circuitry, such as (but not limited to) a central processing unit (CPU), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), and / or any other circuitry as appropriate to the requirements of a particular application of an embodiment of the present invention.

[0049] The target generator 200 further includes an input / output (I / O) interface 220. The I / O interface is capable of transferring data between connected components, such as (but not limited to) a display, a TMS device, an fMRI machine, other therapeutic and / or imaging devices, and / or any other computer component as appropriate to the requirements of a particular application of an embodiment of the present invention. The target generator further includes memory 230. The memory can be implemented using volatile memory, non-volatile memory, or any combination thereof. As can be readily appreciated, any machine-readable storage medium can be used as appropriate to the requirements of a particular application of an embodiment of the present invention.

[0050] The memory 230 contains a targeting application 232. The targeting application can instruct the processor to perform various target generation processes. The memory 230 can also store patient brain data 234. The patient brain data describes patient brain scans, such as, but not limited to, structural MRI and functional MRI scans. In numerous embodiments, the memory 230 can further contain normative connectivity data 236, which describes expected generalized connectivity networks for a standard brain model.

[0051] While specific target generator architectures and target generators are discussed in accordance with embodiments of the present invention above, any number of different architectures and hardware designs can be used without departing from the scope or spirit of the present invention. For example, in many embodiments, different stimulation modalities can be used. In various embodiments, transcranial direct current stimulation is used. In many embodiments, an implantable electrical neurostimulator is used to directly stimulate brain tissue. The target generation process for generating individualized stimulation targets is discussed in further detail below.

[0052] (Generation of individualized stimulus targets) Some brain stimulation methods will work with some degree of efficacy without personalized precision targeting.However, it is very beneficial to provide stimulation to specific brain regions and try to maximize the therapeutic impact of individuals.Various existing methodologies that try to generate personalized targets cannot fully consider the existing network connectivity in the brain and / or simply cluster regions in the brain.The target identification process described herein can provide a more accurate stimulation target for individuals based on their personal brain network connectivity.

[0053] Turning now to FIG. 3 , a flowchart of a target identification process for generating personalized stimulation targets for a patient is illustrated, according to an embodiment of the present invention. Process 300 includes acquiring patient brain data (310). As noted above, the patient brain data can include structural and / or functional brain scans. In many embodiments, the patient brain data includes both structural MRI and functional MRI scans. In various embodiments, multiple structural and / or functional MRI scans are included within the patient brain data, which may have been acquired at different times. The MRI scans can be checked for quality. In various embodiments, scan quality is examined using commonly used fMRI quality control (QC) tools and / or by matching the whole-brain connectivity structure to an expected normative connectivity structure. The target identification process for performing quality control using the expected normative connectivity structure is discussed in further detail in the following section with reference to FIG. 4 .

[0054] Process 300 further includes mapping (320) search and reference regions of interest (ROIs) onto the patient's brain. The ROIs can be any brain structure, substructure, or group of structures of interest within the brain as determined by the user. The reference ROIs are ROIs that describe regions where brain stimulation therapy should indirectly affect. In contrast, the search ROIs describe regions where individualized brain stimulation targets may reside. In this manner, applying stimulation to individualized brain stimulation targets within the search ROIs affects the reference ROI. ROIs can consist of one or more voxels, depending on the size of the particular ROI. In some embodiments, ROIs may overlap. In many embodiments, a brain atlas is used to map the ROIs onto a structural scan of the patient's brain. In various embodiments, the target ROIs are indicated by applying a mask to the brain structures, where the mask flags the desired target ROIs. In various embodiments, the mask can have different weighting metrics for different desired target ROIs. ROIs can also be mapped onto functional scans. In various embodiments, structural scans can be used as templates for aligning other functional scans. In various embodiments, multiple fMRI scans can be combined by integrating functional connectivity data to produce a "combined fMRI." In this way, multiple fMRI scans of a patient taken using similar or the same protocols can be merged to provide a more complete picture of an individual's network connectivity.

[0055] The fMRI signal (i.e., activity levels for a particular voxel or set of voxels over time) is extracted from the ROI (330). Voxels with poor signal quality can be excluded (335) and / or discarded. In numerous embodiments, poor-quality signals can be caused by various scanning device limitations, scanning parameters, and / or movement during the scanning process. In various embodiments, poor-quality signals are detected by calculating a voxel-level signal-to-noise ratio (SNR). By removing low-quality signals from consideration, targeting accuracy can be significantly increased. An ROI parcelization individualization map is derived from the extracted fMRI signal (340). The ROI parcelization individualization map describes multiple parcels (or groups of adjacent voxels). Candidate parcels are derived from the search ROI and constitute candidate targets for brain stimulation therapy. A reference parcel is derived from the reference ROI and constitutes the area of ​​the reference ROI that will be affected by stimulation. A method for deriving ROI parcelization according to an embodiment of the present invention is discussed in further detail below with respect to FIG.

[0056] Relationships between potential candidate parcels and the reference parcel are extracted (350), and target scores for the potential candidate parcels are generated (360). In many embodiments, functional connectivity between two parcels (candidate and reference) is measured, and the target score is based on the strength of the functional connection. Targets that have stronger functional connectivity to the reference ROI (e.g., any parcel within the reference ROI) and therefore more strongly influence the function of the reference may be given a higher target score. In many embodiments, other factors contribute to the score, including (but not limited to) parcel depth, other features of the parcel and / or surrounding brain structures, parcel size, shape, and homogeneity, fit to known / expected system / network-level connectivity profiles, and numerous other factors may be deemed appropriate to the requirements of specific applications of embodiments of the present invention. For example, a larger target may not have strong functional connectivity to the reference, but may be much larger and therefore easier to target with a specific brain stimulation device.

[0057] As an additional example, a network connectivity score can be included that incorporates network-level expectations regarding the brain region to be targeted. If the literature indicates that a particular brain structure or network (i.e., a set of structures) is believed to be involved in a particular condition, parcels that strongly interact with that brain structure / network may be weighted more heavily as potential targets. As noted above, the DLPFC is believed to be strongly associated with clinical depression and suicidal ideation, and therefore targets that strongly interact with that region may be more desirable based on current expectations.

[0058] As an example, in many embodiments, the difference between functional connectivity to the DAN and DMN can be calculated for each parcel. The anti-correlation between the DAN and DMN can be used as the network connectivity score, with a higher degree of anti-correlation indicating a stronger candidate parcel. In various embodiments, the difference between functional connectivity to the VAN and DMN is calculated and used as the network connectivity score. In some embodiments, a weighted average of the network connectivity scores for different networks can be used as the overall network connectivity score, with the weighting based on the relevance of the particular network to the condition in question. In various embodiments, functional connectivity is calculated for each voxel and averaged to obtain an overall parcel score.

[0059] An individualized target parcel is then selected from the group of candidate parcels based on the target score (370). In many embodiments, the candidate parcel with the highest score is selected. In many embodiments, the center for the target parcel is extracted (380) to more precisely determine TMS coil alignment. In many embodiments, the center is calculated by averaging the positions of each voxel that makes up the target candidate.

[0060] A particular method for generating personalized targets is illustrated in Figure 3, but as can be readily understood, any number of different modifications can be made without departing from the scope or spirit of the present invention. For example, not all quality control steps need to be performed or all parameters considered to generate a network score that is appropriate to the requirements of a particular application of an embodiment of the present invention. Furthermore, different weightings may be given to different parameters with respect to their relative importance in calculating the target score. Additional descriptions of the various steps of the above process are found below.

[0061] (Network Connectivity Quality Control) Patient brain data can include one or more fMRI scans; however, it is rarely immediately guaranteed that the data is of high quality (e.g., has a high SNR). Measurement noise and head movement are known causes of fMRI reliability limitations and are therefore estimated and partially addressed as a common practice during data preprocessing. However, in some cases, poor scan quality and / or preprocessing errors may be overlooked, which may lead to deriving targets based on incorrect brain functional connectivity structures. To prevent making clinical decisions based on erroneous data, additional measures are desirable.

[0062] Under the likely and acceptable assumption of global integrity in the systems-level organization of the human brain, matching measured whole-brain connectivity to expected normative connectivity can reduce errors from bad scans and, in some cases, alert medical professionals to the presence of atypical brains for further manual review. In many embodiments, identified bad scans are discarded. Turning now to Figure 4, a target identification process for measuring expected network connectivity is illustrated, according to one embodiment of the present invention.

[0063] Process 400 includes assigning each voxel to a predetermined network (410). Many large-scale brain networks, such as (but not limited to) the visual insulator (VIS), sensorimotor network (SMN), dorsal attention network (DAN), ventral attention network (VAN), limbic network, frontoparietal control network (FPCN), and default mode network (DMN), are known and have been mapped based on large population samples. These networks can be overlaid on a patient's MRI so that each voxel is assigned to at least one network. For each pair of voxels, a functional connectivity score (FC) can be calculated (420), where FC represents the strength of connectivity between voxels in the fMRI (including combined fMRI). All FC values ​​relating voxels assigned to the same network are averaged (430), resulting in an "intra-FC" value.

[0064] The "inter-FC" value is obtained by averaging all FC values ​​relating voxels from different networks (440). The inter-FC value is subtracted from the intra-FC value (450) to obtain the network fit for the voxel. Although individual voxels may vary in their network associations due to expected individual differences in brain function and structure, the average network fit across voxels (called the network quality control (QC) metric) is expected to remain positive (intra-FC > inter-FC). If the network QC metric is not significantly positive (meaning inter-FC >= intra-FC), this is an indicator that there may be a problem with either the scanning, preprocessing procedures, or significantly atypical structural issues occurring in the patient's brain. Statistical significance of the network QC metric can be obtained by randomly permuting the data while taking into account the spatial location of the voxels and repeating the network QC estimation process. In this way, the acquired fMRI can be screened for quality. If an fMRI scan is flagged as having a poor overall network fit, this can allow for detailed inspection of the data by medical experts and prevent deriving targets from erroneous information.

[0065] A specific method for QC control based on brain network connectivity is illustrated in accordance with one embodiment of the present invention in Figure 4, although network connectivity can be used as a control using any of several different algorithms as appropriate to the requirements of a particular application of an embodiment of the present invention. Ensuring high quality data can increase the accuracy of generated targets. A discussion of methods for parceling the brain into individualized ROIs is discussed further below.

[0066] (ROI parcelization) While the overall structure of the human brain is relatively conserved across individuals, it is well known that each person has unique brain functionality and circuitry based on any number of factors, both environmental and genetic. Therefore, simply segmenting the brain based on a standardized model may yield inaccurate or insufficient results. Previous attempts have been made to parcel the brain into ROIs, but the specific methodologies used often fail to robustly cluster voxels in an effective manner. Turning now to FIG. 5 , a target identification process for deriving an individualized map of ROI parcelization, according to one embodiment of the present invention, will be discussed.

[0067] Process 500 involves randomly subsampling (510) a percentage of all voxels. In many embodiments, the percentage is any number above 80%, however, depending on the amount of available data and calculations, the number may be less than 80%. The fMRI signals within the subsampled voxels are then clustered (520) based on signal similarity. Any number of different clustering processes can be used, including (but not limited to) agglomerative (hierarchical) clustering, cluster identification via connectivity kernel (CLICK) clustering, k-means clustering, and / or spectral clustering. In some embodiments, clustering methods that incorporate spatial information (e.g., spatially constrained spectral clustering) can be used.

[0068] The clustering assignments are recorded (530) and a new random subsampling (510) is obtained. The process can be repeated multiple times to increase accuracy. In many embodiments, the process is repeated 100 or more times to ensure sufficient data, although fewer may be sufficient. The subsample clustering solutions are then merged (540). In many embodiments, they are merged using a consensus clustering approach. Any resulting spatially disjoint clusters can then be divided into subclusters (550). Clusters (and any subclusters) are then labeled as either reference or search parcels (560) based on their location within the reference and search ROIs.

[0069] By repeatedly subsampling and clustering, noise in neural signals can be accounted for, and a more accurate picture of an individual's true brain connectivity can emerge. Furthermore, multiple fMRI scans can be performed through this process, and the resulting clusters can be integrated using consensus clustering. In this way, multiple fMRI scans, including those taken on different days, can contribute to the overall dataset used for targeting. In various embodiments, spatially disjoint clusters can be avoided by using a spatially constrained clustering process. However, depending on the requirements of a specific application of an embodiment of the present invention, it may be desirable to select a spatially unconstrained clustering process that can result in spatially disjoint clusters. A target identification process for dividing spatially disjoint clusters according to an embodiment of the present invention is illustrated in FIG. 6.

[0070] Process 600 involves recording the spatial location of each voxel in spatially disjoint clusters (610). A distance matrix indicating the physical distance between every two voxels is generated (620), which is then converted to a graph representation (630). Long edges in the graph (edges exceeding a predetermined threshold) are pruned (640), resulting in a partially connected graph, which is then split into connected subgraphs (components), if such occur. The set of voxels in each connected component can then be defined as a separate cluster (650).

[0071] In this way, disjoint clusters can be partitioned and used separately as potential candidate parcels for stimulation. In many embodiments, these disjoint clusters are problematic in that the "center" of a disjoint cluster may lie outside any part of the disjoint cluster and may not be anywhere near a viable target location.

[0072] Processes 300, 400, 500, and / or 600, and variations thereof, may be performed by a target identification system to provide a target parcel, which may then be archived, stored for later use, transmitted to a neuronavigation device, used in further analysis, or combined (e.g., by joining or intersecting) with one or more other target parcels to result in a composite target parcel. The target identification system may be distinct from, separate from, and / or integrated or partially integrated with the neuronavigation device. The target identification system may be implemented on a cloud computing platform, on a computing platform local to the site of care, on a computing platform incorporated into or part of the neuronavigation device, or on any combination of such platforms.

[0073] Although specific methods of ROI parcelization are discussed above, many different methods, including (but not limited to) those using different specific clustering processes and / or utilizing different thresholds and parameters, can be implemented in accordance with many different embodiments of the present invention. It should therefore be understood that the present invention may be practiced otherwise than as specifically described without departing from the scope and spirit of the present invention. Accordingly, the present embodiments are to be considered in all respects as illustrative and not restrictive. The scope of the present invention should, therefore, be determined not by the exemplified embodiments, but by the appended claims and their equivalents.

Claims

[Claim 1] The invention described in this specification.

Images