Automated network detection and confidence mapping using functional neuroimaging data

Abstract

A functional network map of the somatomotor cognitive action network (SCAN) is generated from magnetic resonance data based on a template matching of time-course signals with one or more functional network templates. A first network map is generated using correlations of the time-course signals with the functional network template(s) at a first threshold value. The first network map is then updated by reassigning grayordinates in the first network map based on correlations generated using a second threshold value that is higher than the first threshold value. A network confidence map can be generated to indicate the confidence in network assignments. The network confidence map is generated using a permutation-based analysis in which dense time series data are partitioned, the partitioned shuffled based on a number of permutations. Network assignments are generated for each permutation and compared across shuffles to assess network confidence.

Description

UMN 2024-350Aty. Docket: 920171.00660AUTOMATED NETWORK DETECTION AND CONFIDENCE MAPPING USING FUNCTIONAE NEUROIMAGING DATACROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application Serial No. 63 / 700,419, filed on September 27, 2024, and entitled “Automated Detection of the Somatomotor Cognitive Action Network and Network Confidence Mapping Using Functional Neuroimaging Data,” which is herein incorporated by reference in its entirety'.STATEMENT OF FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with government support under MH096773 awarded by the National Institutes of Health. The government has certain rights in the invention.BACKGROUND

[0003] Functional magnetic resonance imaging (fMRI) is used to investigate brain activity and connectivity. The somatomotor cognitive action network (SCAN) integrates regions involved in motor control and cognitive processes related to action planning and execution. The SCAN encompasses areas traditionally associated with the somatosensory and motor systems, such as the primary motor cortex, supplementary motor area, and parietal lobes, along with regions implicated in higher-order cognitive functions, such as the prefrontal cortex.

[0004] Traditional methods of analyzing fMRI data often focus on localized brain activity, potentially overlooking the intricate interconnections that characterize the functionality of the SCAN. Accurately identifying the SCAN through functional network mapping is challenging because it is susceptible to subjective biases and methodological limitations. Existing methods fall short in either automation or reliability.SUMMARY OF THE DISCLOSURE

[0005] It is an aspect of the present disclosure to provide a method for generating a functional network map of a somatomotor cognitive action network (SCAN) from magnetic resonance image data acquired from a subject using a magnetic resonance imaging (MRI) system. The method includes accessing magnetic resonance image data with a computer system, where the magnetic resonance image data comprise a time-series of images whose voxels depict blood-oxygen-level-dependent (BOLD) signals. Time course signal data1QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 are formed for each of a plurality of gray ordinates associated with the subject using the computer system. The time course signal data are formed for each of the plurality of gray ordinates as BOLD signals at each of the plurality of gray ordinates measured over the time series of images. A first correlation matrix is computed from the time course signal data for each of the plurality of grayordinates using the computer system and a first threshold value. Functional network template data are accessed with the computer system, where the functional network template data include at least a SCAN template indicative of grayordinates associated with a SCAN and a motor network template indicative of gray ordinates associated with a motor network. Similarity7values are computed between the first correlation matrix and each functional network template in the functional network template data. Functional network maps are then generated with the computer system by assigning gray ordinates to the functional networks in the functional network data based on the similarity values. Similarity values are recomputed from the functional network maps for each of the plurality7of gray ordinates using the computer system at a second threshold value that is higher than the first threshold value. Updated similarity values are computed for each functional network template in the functional network template data. A SCAN map is then generated with the computer system by reassigning grayordinates to the SCAN based on the updated similarity7values.

[0006] It is another aspect of the present disclosure to provide a method for generating a network confidence map that indicates confidence in functional network assignments based on magnetic resonance image data. The method includes accessing magnetic resonance image data with a computer system, where the magnetic resonance image data comprise a time-series of images whose voxels depict BOLD signals. Dense time series data are formed from all voxels across the time-series images in the magnetic resonance image data acquired from the subject. The dense time series data are formed by recording temporal fluctuations in the BOLD signals over time. The dense time series data are partitioned into a plurality of partitions, and permutated dense time series data are generated by shuffling the partitioned dense time series data according to a plurality of permutations. The permutated dense time series data are split into first split-half dense time series data and second split-half dense time series data. First network assignment data are generated from the first split-half dense time series data and second network assignment data are generated from the second split-half dense time series data. A network confidence map that indicates confidence in functional netw ork assignments is then generated based on a comparison of the first network assignment data and the second2QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 network assignment data. The network confidence map may be output with the computer system.BRIEF DESCRIPTION OF THE DRAWINGS

[0007] FIG. l is a flowchart of an example method for generating an individual-specific functional network map of a somatomotor cognitive action network (SCAN).

[0008] FIGS. 2A-2E illustrate an example workflow for automated SCAN identification. The seed-based spatial pattern of connectivity for a given subject is compared to a series of network templates, which have been generated in an independent dataset (FIG. 2A). The number of networks can vary (t pi cal ly 8-20 different networks), and only three are shown here for visualization purposes. The template is thresholded to Z-score > 1 (or about the top 15% of connections). After a measure of similarity is calculated (e g. using correlation or eta-squared), that region gets the assignment of the most similar network (FIG. 2B). To identify the scan network more accurately, only the motor systems are thresholded (FIG. 2C). Here, the somatomotor dorsal (SM Dorsal) netw ork, the somatomotor lateral networks (SM lateral), and the SCAN are shown. A comparison against the network templates at a higher threshold (Z- score > 3) is then recalculated (FIG. 2D). The motor regions get reassigned based on the assignment of the most similar network (FIG. 2E).

[0009] FIG. 3 is a flow chart of an example method for generating netw ork confidence maps using a permutation-based analysis.

[0010] FIGS. 4A-4G illustrate an example workflow' for generating permutation-based network confidence maps. Time courses of functional activity of even’ vertex are extracted and combed into a dense time series (FIG. 4A). Dense time series (dtseries) are concatenated across all runs and sessions into one dtseries that contains the time courses of all brain vertices across all the scan times (FIG. 4B). Rows represent the areas of the brain, and columns represent time. The total time of the dtseries is calculated and split into X partitions (FIG. 4C). Partitions of data are shuffled for N permutations (FIG. 4D). Data from shuffled dtseries are run through a community detection algorithm to identify which vertex belongs to which functional network (FIG. 4E). For each vertex, netw ork assignment is checked and confidence is assessed to identify which vertices have consistent network assignment across shuffles and which assignments change (FIG. 4F). Dense scalars from each permutation are combined to make the network confidence map, highlighting the probabilistic variability across shuffled runs (FIG. 4F). Lighter colors indicate lower confidence, and darker higher confidence.3QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660

[0011] FIGS. 5A-5E illustrate an example workflow for generating permutation-based network confidence maps using a bootstrap resampling approach.

[0012] FIG. 6 is a block diagram of an example system for generating SCAN maps and associated network confidence maps.

[0013] FIG. 7 is a block diagram of example components that can implement the system of FIG. 6.

[0014] FIG. 8 is a block diagram of an example magnetic resonance imaging (MRI) system.DETAILED DESCRIPTION

[0015] Described here are systems and methods for the automated identification of the Somatomotor Cognitive Action Network (SCAN), which is advantageous for assessing brain function and guiding interventions (e.g., neuromodulation). Alternatively, the systems and methods can be adapted to identify other functional networks of interest, such as the default mode network, visual network, frontal parietal network, or other brain networks. The SCAN is a distributed brain system for integrating abstract behavioral plans with movements and autonomic functions of the body. The SCAN can include nodes such as the supplementary motor area (SMA), the middle insula, inter-effector regions in the primary motor cortex, the dorsal putamen, the ventralis intermedius (VIM) nucleus of the thalamus, the centromedian (CM) nucleus of the thalamus, the red nucleus, the subthalamic nucleus (STN), the substantia nigra, and the dorsal motor nucleus of the vagus nerve, among others. In some instances, the SCAN may alternatively be referred to as the mind-body interface (MBI).

[0016] In general, the disclosed systems and methods combine advanced network mapping algorithms (e.g., template matching, precision functional mapping (PFM)) with a permutation-based analysis to ensure the reliability and objectivity of SCAN detection. As one non-limiting example, the permutation-based analysis can include permutation-based data shuffling.

[0017] The methodology developed offers broad commercial potential, particularly for creating precise targeting maps for functional neurosurgical implants and pain management strategies. Its reliability and automation capabilities make it a valuable tool for advancing neurosurgical planning and therapeutic interventions.

[0018] Referring now to FIG. 1, a flowchart is illustrated as setting forth the steps of an example method for generating an individual-specific network map that identifies the SCAN4QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 region, or other functional network region, of the individual. In general, the method implements a comprehensive approach to analyze the functional connectivity across cortical and subcortical areas of the brain, leveraging high-resolution neuroimaging data (e.g., functional magnetic resonance imaging (fMRI) data).

[0019] The method includes accessing magnetic resonance image data with a computer system, as indicated at step 102. Accessing the magnetic resonance image data can include retrieving previously acquired data from a memory or other data storage device or medium. Additionally or alternatively, accessing the magnetic resonance image data can include acquiring the data with a magnetic resonance imaging (MRI) system and transferring or otherwise communicating the data to the computer system, which in some embodiments may be a part of the MRI system.

[0020] In general, the magnetic resonance image data include images acquired with an MRI system. The images can include a time-series of functional images acquired while a subject is performing a task (e.g., a functional task), while a subject is in a resting-state, or both. In these instances, the magnetic resonance image data includes task-based functional MRI data (e.g., data acquired while a subject is performing a functional task), resting-state functional MRI data, or both. As an example, a functional task may include a motor task (e.g., finger tapping), a stop signal task, an emotional n-back task, and the like. In some instances the magnetic resonance data may additionally include structural image data (e g., anatomical images such as MP RAGE images. T1 -weighted images, T2-weighted images, etc.).

[0021] A functional image depicts a region or volume-of-interest (e.g., a slice, slab, or volume) imaged within a subject’s brain, and the time-series of functional images represents the time course of magnetic resonance signals (e.g., blood-oxygen-level dependent (BOLD) signals) in that region or volume over the duration of time during which the time-series of functional images was acquired. The time-varying magnetic resonance signals measured at a pixel or voxel location can be referred to as a time course, or time course signal data.

[0022] In some embodiments, time course signal data can be constructed by tracking the time-varying magnetic resonance signals measured at a grayordinate or other brainordinate over the time-series of functional images. A brainordinate is a coordinate (e.g., a particular location) within a subject’s brain and a gray ordinate is a brainordinate within the gray matter of a subject’s brain. As one example, a brainordinate can be specified by a surface vertex, or node. As another example, a brainordinate can be specified by a volume voxel. In still other examples, a brainordmate can include other suitable spatial units associated with magnetic5QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 resonance imaging. Thus, a gray ordinate corresponds to a particular location in the gray matter that can be specified as gray-matter surface vertices (e.g., cortical gray matter), gray-matter volume voxels (e.g., subcortical gray matter), or both.

[0023] In some embodiments, the magnetic resonance image data accessed with the computer system have been preprocessed (e.g., to denoise the images, the perform bias field correction, to perform brain extraction, to perform motion correction). In other embodiments, the magnetic resonance image data can be preprocessed using the computer system after accessing the data.

[0024] In some implementations, volumetric time course signal data (e.g., BOLD functional MRI volumetric data in the magnetic resonance image data) can be constrained to the cortical surface. In these instances, the volumetric time course signal data are mapped to the cortical surface, after which they are deformed and resampled to the original surface. The left and right surfaces can, in some instances, be combined with volumetric midbrain and hindbrain time course signal data into a CIFTI (“Connectivity Informatics Technology Initiative") format.

[0025] Functional network template data for a plurality of different functional networks are accessed with the computer system, as indicated at step 104. Accessing the functional network template data can include retrieving previously generated data from a memory7or other data storage device or medium. Additionally or alternatively, accessing the functional network template data can include generating the functional network template data with the computer system. Functional network templates can be generated from magnetic resonance image data (i.e., functional images and / or time course signal data) obtained from a group of subjects, or participants, as group-average network assignments in the brain (e.g.. in the cortex).

[0026] As one example, the functional network template data can include templates for the SCAN in addition to other functional networks, such as the sensorimotor dorsal network (“SMd”) and the sensorimotor lateral network (“SMI”) (as shown in FIG. 2A). Additionally or alternatively, the functional netw ork template data can include templates for other functional networks such as the default mode network (“DMN”), the visual network (“VIS”), the frontal parietal network (“FPN”), the dorsal attention network (“DAN”), the ventral attention network (“VAN”), the salience network (“Sal”), the cingulo-opercular network (“CO”), the auditory network (“AUD”), the temporal pole network (“Tpole”), the medial temporal network (“MTL”), the parietal occipital network (“PON”), and / or the parietal medial network (“PMN"). Sensory and motor systems can be combined due to the coupled nature of activation. In other6QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 implementations, the templates can include fewer of these functional networks and / or can include additional functional networks.

[0027] As a non-limiting example, functional network template data can be generated using an Infomap community detection algorithm. Brain network organization can be described using a two-level system of networks and nodes, respectively. Infomap is a network-describing algorithm based on the duality of finding community structure in networks and minimizing the description length of a random walk on a network. For example, the Infomap algorithm can minimize the number of bits (e.g., using Huffman coding) necessary to describe the whole network and using a random walk algorithm that uses connection weights to determine the minimum descriptor code length.

[0028] To generate functional network template data using an Infomap community detection algorithm, a correlation matrix is first generated using motion-censored dense time series data. For example, a voxelwise correlation matrix can be computed by correlating the time course signal data (i.e., BOLD time series) at each grayordinate with the time course signal data of each other gray ordinate. The correlation matrices for each participant in a group can be transformed and averaged across participants. For example, the correlation matrices can be transformed using a Fisher transform and the inverse Fisher transform can be applied to the group-average matrix.

[0029] A group-average matrix is then applied to the Infomap algorithm in order to identify functional networks across a range of edge density thresholds. For example, each upper triangle of the correlation matrix can be thresholded to various top percentages (e.g., 0.3, 0.4, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0) of the connections. Those thresholded connections can then be used as the input for the Infomap algorithm. Infomap uses a random walk to minimize bit-wise code length necessary to describe the whole system structure. The final network labels can be determined by generating a consensus across thresholds. In some embodiments, the Jaccard index of the spatial arrangement of gray ordinates from the detected network can then be compared with those found in the group.

[0030] Additionally averaging can be applied. For example, the average time course signal of all gray ordinates labeled for a particular network in the group-average consensus map can be extracted. The average time course signal can then be correlated with all other grayordinate time course signals in order to generate a network seedmap, which can be averaged across group participants. This can be repeated for each network, thereby generating a group-average functional network template map for each brain network.7QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660

[0031] For example, in some embodiments, an independent template is generated using a seed-based correlation (e.g., using an average time series correlated to all the grayordinates) for all networks. Seed-based correlations can be generated using a dense time series from each template participant that were smoothed with a within-frame spatial Gaussian smoothing kernel (e.g., with 2.55 mm smoothing) using each participant’s own midthickness surfaces. The resulting networks can be converted to a CIFTI file format and applied to the smooth dense time series to generate an average time series for each network. The time series of the seed can then be correlated with the times series of all other grayordinates.

[0032] Seed-based correlation values can be averaged across all the participants in the template group, resulting in a vector of average correlation values for each network correlated with each gray ordinate. Each network vector can be averaged independently across subjects in the template group to generate seed-based templates for each network. Each network template can then be thresholded (e.g., at Z > 1).

[0033] An individual-specific functional network map identifying the SCAN is then generated, as generally indicated at process block 106. The individual-specific functional network map may be referred to as a SCAN map, since the resulting individual-specific functional network map depicts or otherwise identifies the SCAN in the individual subject whose magnetic resonance image data are processed. In alternative embodiments, the methods described herein can be adapted to identify other functional networks beyond the SCAN, such as the default mode network, visual network, frontal parietal network, or other networks of interest.

[0034] While the following description focuses on template matching approaches, it should be understood that other methods for generating individual-specific network maps can be used, including other template-based, graph-based, and / or generative approaches. Examples of such alternative approaches include Infomap / Leiden community detection algorithms, Order Statistics Local Optimization Method (OSLOM), Non-Negative Matrix Factorization (NMF), Independent Component Analysis (ICA), spectral clustering, Multi-Session Hierarchical Bayesian Models (MS-HBM), stochastic block models, and the like.

[0035] By way of example, alternative template-based approaches may include probabilistic template mapping methods, which incorporate probabilistic network assignments rather than deterministic assignments, providing uncertainty estimates for network boundaries.

[0036] As another example, flow-based and / or modularity-based graph community detection approaches could be used. For instance, Infomap algorithms use flow-based8QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 approaches and are often robust on correlation graphs, utilizing information-theoretic principles to identify communities. Louvain and Leiden algorithms maximize modularity, with Leiden being the safer and faster successor to Louvain. Multilayer modularity approaches, such as GenLouvain, are designed for time-resolved and dynamic functional connectivity7analysis. Markov Stability methods identify communities using random-walk multiscale approaches. Asymptotical Surprise provides a significance-based alternative to traditional modularity optimization.

[0037] Overlapping-community methods, such as OSLOM, could also be used. OSLOM finds statistically significant and possibly overlapping communities. NMF-based approaches also yield soft and overlapping network structures. In other cases, matrix factorization and / or decomposition approaches could be used. As a non-limiting example, NMF provides parts-based components and can be used for resting-state network discovery and group / subject mapping. ICA represents another approach for resting-state network identification.

[0038] Bayesian and / or generative Models, such as MS-HBM provide individualspecific cortical parcellations across multiple sessions. Weighted Stochastic Block Models (WSBM) and Stochastic Block Models (SBM) offer explicit generative models of mesoscale brain network structure.

[0039] Using the functional network template data, individual-specific network assignments are then determined. A voxelwise or brainordinate-wise correlation matnx is generated from the time course signal data, as indicated at step 108. For example, the correlation matrix can be generated by correlating the BOLD signals for each gray ordinate with the BOLD signals every other grayordinate represented in the time course signal data. In some embodiments, the correlation matrix can be thresholded to a percentage of top connectivity values (e g., the top 5% connectivity7values) across gray ordinates.

[0040] The similarity7between the correlation matrix and one or more of the templates in the functional network template data is then computed, as indicated at step 110. As one example, an eta-squared (rf2) value, which is a measure of association or similarity at each gray ordinate, can be calculated between the remaining gray ordinates and each of the network templates. Alternatively, other similarity metrics can be computed, such as correlation, Dice coefficients, or the like.

[0041] Based on the similarity7values, each grayordinate is assigned to one or more functional networks, generating output as one or more individual-specific functional network9QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 maps, as indicated at step 112. For instance, each gray ordinate is given a network assignment based on the maximum eta-squared value (e.g., as shown in FIG. 2B).

[0042] To identify the SCAN more accurately, the motor networks are then thresholded. For instance, after the motor networks have been identified (e.g., as shown in FIG. 2C), the motor networks are then compared again against the motor templates at a higher threshold, as indicated at step 114 and illustrated in FIG. 2D. For example, each template can be thresholded using a higher threshold than the one used in step 106. As a non-limiting example, the higher threshold can be Z > 3 or Z = 3. This higher thresholding approach can similarly be applied to identify' other functional networks more accurately by focusing on the strongest connectivity patterns and reducing noise in network assignments. Updated similarity values are then calculated using the second correlation matrix generated using the higher thresholding, as indicated at step 116. The grayordinates are then reassigned to the functional networks to which they have the maximum eta-squared value (or other similarity measure), as indicated at step 118 and illustrated in FIG. 2E.

[0043] As indicated at step 120, after the individual-specific functional network maps are generated, they can be stored for later use, displayed to a user, or both. For example, the individual-specific functional network maps can be stored in a memory or other data storage device or medium using the computer system, where the individual-specific functional network maps can be later accessed for further processing or displayed to a user. In some embodiments, the individual-specific functional network maps can be displayed to a user using the computer system.

[0044] As another example, the individual-specific functional network maps may be analyzed to monitor and / or measure the efficacy of targeted brain stimulation or other neuromodulation therapies that have been delivered or otherwise administered to the subject. For instance, the individual-specific functional network maps may be compared to reference or baseline maps to monitor and / or measure the efficacy of the targeted brain stimulation or other neuromodulation therapies. The comparison may be performed on a gray ordinate basis, on a brainordinate basis, a network basis, or so on. For example, the individual-specific functional network map(s) generated for the subject can be compared with the reference or baseline to assess whether the topography (e.g., size, extent, brainordinate locations) or other characteristics or features of the subject's functional networks have changed in response to the targeted brain stimulation or other neuromodulation therapies.10QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660

[0045] Referring now to FIG. 3. a flowchart is illustrated as setting forth the steps of an example method for measuring the reliability’ of the individual-specific functional network maps using a permutation-based data analysis. For example, to ensure the reliability- of the functional network mapping described in the present disclosure, a constructed dense time series (as shown in FIGS. 4A and 4B) is assessed in different configurations by splitting the data into partitions (as illustrated in FIG. 4C) and shuffling these for a number (N) of permutations (as illustrated in FIG. 4D). This introduces sampling variability while preserving the brain’s connectivity structure. Each shuffled dense time series is then bisected for comparative analysis through the template matching mentioned above and in FIG. 1, providing a measure of reliability’ of functional connectivity patterns of SCAN.

[0046] The method includes constructing dense time series data from the magnetic resonance image data, as indicated at step 302, as illustrated in FIG. 4A. By recording the temporal fluctuations in the high resolution BOLD signal over time in the magnetic resonance data a dense time series (dtseries) can be constructed from all voxels across all functional MRI runs of a given subject. These volumetric data can be projected onto a 3D model of the cortical surface following the CIFTI structure.

[0047] The dense time series data can then be concatenated across all runs and sessions to form unified dense time series data that contains the time courses of all brain vertices across all the scan times, as indicated at step 304. For instance, the unified dense time series data can include rows indicating the vertices (gray ordinates) of the left cortex followed by the right cortex and the subcortical regions, and can include columns indicating the time points of each collected frame of data, which may correspond to each repetition time (TR). This step constructs a comprehensive representation of brain activity across the entire imaging sessions.

[0048] The total time of the unified dense times series is calculated and split into X partitions based on a percentage of the total length of the scans, as indicated at step 306. An example of this is illustrated in FIG. 4C. The partitioned dense time series data are then shuffled for a total of N permutations, as indicated at step 308 and illustrated in FIG. 4D. The shuffled dense time series data are then split, as indicated at step 310 and illustrated in FIG. 4D. In the illustrated example, the shuffled dense time series data are split into two halves. Alternatively, instead of splitting the shuffled partitions into halves, surrogate dense time series data can be generated by sampling the partitions with or without replacement and concatenating the sampled partitions according to a plurality of resampling draws.11QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660

[0049] The stability of template matching network assignments can then be evaluated at the individual voxel or gray ordinate level across the two halves of each shuffled dense time series permutation. For each grayordinate, the network assignment is identified in both split halves. For example, the method illustrated in FIG. 1 can be used to generate network assignments for each split half of the dense time series data, as indicated at step 312. Whether the assignment remains consistent or changes is logged to quantify the consistency of network delineation, as indicated at step 314 and illustrated in FIG. 4F. This provides a robust measure of the reliability of the network mapping techniques described in the present disclosure.

[0050] In an alternative embodiment, as illustrated in FIGS. 5A-5E, rather than performing a split-half analysis, a bootstrap resampling approach may be employed to assess the reliability of network assignments. In these instances, steps 310-314 can be replaced with use cases where all of the data are used for a bootstrap with replacement, and then reliability and confidence is assessed across permutations, rather than between split halves. In this approach, the total time of the unified dense time series is calculated and split into X partitions based on a percentage of the total length of the scans. As described above, the dense time series data are partitioned into a plurality of partitions, and surrogate dense time series data are generated by sampling the partitions with or without replacement and concatenating the sampled partitions according to a plurality of resampling draws. This bootstrap approach may utilize all of the available data or a user-chosen percentage of the data for each resampling iteration.

[0051] The stability of template matching network assignments can then be evaluated using the surrogate dense time series data generated through bootstrap resampling. For each surrogate dense time-series, network assignment data are generated using the template matching methods described herein. This approach may provide a more comprehensive assessment of network reliability by utilizing the full dataset in each permutation rather than splitting the data into separate halves. The bootstrap resampling may preserve the temporal structure of the data while introducing variability through the random sampling of partitions.

[0052] One or more network confidence maps can then be generated for the functional networks identified based on the network assignments from the split-half or otherwise shuffled dense time series data, as indicated at step 316 and illustrated in FIG. 4G. As a non-limiting example, dense scalar maps from each permutation can be aggregated to construct a network confidence map. This map highlights the probabilistic variability of network assignments across shuffled runs, offering insights into the stability and reliability of functional connectivity12QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 patterns. Areas of high and low confidence are delineated, enriching the identification and reliability’ of the SCAN or other functional networks.

[0053] In the case where reliability is assessed across permutations rather than split halves, a network confidence map indicative of confidence in functional network assignments may be generated by aggregating, for each spatial location, the plurality of network assignment data across the surrogates to compute a consistency measure of assignment to one or more networks. This aggregation process may involve calculating the frequency with which each grayordinate is assigned to a particular network across all bootstrap iterations. The resulting confidence map may highlight areas where network assignments are highly consistent across resampling iterations, as well as regions where assignments may be more variable. Areas of high confidence may indicate robust network boundaries, while areas of lower confidence may suggest regions of network overlap or transition zones.

[0054] Additionally or alternatively, mode maps and mode proportion maps can be generated from the network assignment data across permutations. A mode map indicates the most frequently assigned network (i.e., the mode) for each gray ordinate across all permutations, providing a consensus view of network assignments that represents the most stable network identity for each spatial location. A mode proportion map indicates the frequency or proportion with which each network’s mode assignment occurs across permutations, quantifying how often the most common network assignment is observed relative to the total number of permutations. These maps provide complementary information to the network confidence maps by highlighting not only the reliability of assignments but also the dominant network identify and the strength of that dominance at each spatial location.

[0055] As indicated at step 318, after the network confidence maps are generated, they can be stored for later use. displayed to a user, or both. For example, the network confidence maps can be stored in a memory or other data storage device or medium using the computer system, where the network confidence maps can be later accessed for further processing or displayed to a user. In some embodiments, the network confidence maps can be displayed to a user using the computer system

[0056] FIG. 5 shows an example of a system 500 for automated detection of the somatomotor cognitive action network (SCAN), or other functional network, and network confidence mapping in accordance with some embodiments described in the present disclosure. As shown in FIG. 5, a computing device 550 can receive one or more types of data (e.g., magnetic resonance image data, functional image data, structural image data, functional13QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 network template data) from data source 502. In some embodiments, computing device 550 can execute at least a portion of a network detection and confidence mapping system 504 to generate a functional network map (e.g., a SCAN map) and / or network confidence map from data received from the data source 502.

[0057] Additionally or alternatively, in some embodiments, the computing device 550 can communicate information about data received from the data source 502 to a server 552 over a communication network 554, which can execute at least a portion of the network detection and confidence mapping system 504. In such embodiments, the server 552 can return information to the computing device 550 (and / or any other suitable computing device) indicative of an output of the network detection and confidence mapping system 504.

[0058] In some embodiments, computing device 550 and / or server 552 can be any suitable computing device or combination of devices, such as a desktop computer, a laptop computer, a smartphone, a tablet computer, a wearable computer, a server computer, a virtual machine being executed by a physical computing device, and so on. The computing device 550 and / or server 552 can also reconstruct images from the data.

[0059] In some embodiments, data source 502 can be any suitable source of data (e.g., measurement data, images reconstructed from measurement data, processed image data), such as an MRI system, another computing device (e.g., a server storing measurement data, images reconstructed from measurement data, processed image data), and so on. In some embodiments, data source 502 can be local to computing device 550. For example, data source 502 can be incorporated with computing device 550 (e.g., computing device 550 can be configured as part of a device for measuring, recording, estimating, acquiring, or otherwise collecting or storing data). As another example, data source 502 can be connected to computing device 550 by a cable, a direct wireless link, and so on. Additionally or alternatively, in some embodiments, data source 502 can be located locally and / or remotely from computing device 550, and can communicate data to computing device 550 (and / or server 552) via a communication network (e.g., communication network 554).

[0060] In some embodiments, communication network 554 can be any suitable communication network or combination of communication networks. For example, communication network 554 can include a Wi-Fi network (which can include one or more wireless routers, one or more switches, etc.), a peer-to-peer network (e.g., a Bluetooth network), a cellular network (e g., a 3G network, a 4G network, etc., complying with any suitable standard, such as CDMA. GSM, LTE, LTE Advanced. WiMAX, etc.), other types of14QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 wireless network, a wired network, and so on. In some embodiments, communication network 554 can be a local area network, a wide area network, a public network (e.g., the Internet), a private or semi-private network (e.g., a corporate or university intranet), any other suitable type of netw ork, or any suitable combination of networks. Communications links show n in FIG. 5 can each be any suitable communications link or combination of communications links, such as wired links, fiber optic links, Wi-Fi links. Bluetooth links, cellular links, and so on.

[0061] The network detection and confidence mapping system 504 provides significant technical improvements in the field of functional magnetic resonance imaging by implementing automated, objective methods for identifying functional brain networks. Traditional fMRI analysis methods often rely on subjective interpretation and manual processing, which can introduce variability and bias in network identification. The network detection and confidence mapping system 504 addresses these limitations by implementing the automated template matching algorithms described with respect to FIG. 1 , which systematically compare individual subject data against standardized functional network templates using quantitative similarity measures such as eta-squared values. This automated approach eliminates subjective biases and provides consistent, reproducible network assignments across different subjects and analysis sessions.

[0062] Furthermore, the network detection and confidence mapping system 504 implements the dual-threshold approach described in the method of FIG. 1, wherein initial network assignments are refined using a higher threshold value to improve the accuracy of network identification. This technical innovation allows the system to first capture broad network patterns at a low er threshold and then focus on the strongest connectivity patterns at a higher threshold, thereby reducing noise and improving the precision of network boundaries. The computing device 550 and / or server 552 are specifically configured to execute these dualthreshold computations efficiently, processing large correlation matrices derived from high- resolution fMRI data while maintaining computational performance suitable for clinical and research applications.

[0063] The network detection and confidence mapping system 504 also implements the permutation-based reliability’ assessment methods described with respect to FIG. 3, providing a technical advancement in quantifying the confidence of functional netw ork assignments. Traditional fMRI analysis methods typically do not provide measures of reliability’ for network assignments, leaving clinicians and researchers uncertain about the stability' of their results. The network detection and confidence mapping system 504 addresses this limitation by15QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 implementing automated data partitioning, shuffling, and reliability analysis procedures that generate quantitative confidence maps through either split-half analysis or bootstrap resampling approaches. These confidence maps provide spatial information about which brain regions have stable network assignments and which regions may have variable or uncertain assignments, thereby improving the interpretability and clinical utility of functional network maps.

[0064] The technical implementation of the permutation-based analysis within the network detection and confidence mapping system 504 represents a significant computational advancement, as the system is configured to efficiently manage and process multiple permutations of dense time series data while preserving the temporal structure of brain connectivity patterns. The computing device 550 and / or server 552 are configured with sufficient memon and processing capabilities to handle the computationally intensive operations required for generating multiple shuffled datasets, computing correlation matrices for each permutation, and aggregating results across permutations to generate confidence maps. This automated processing capability enables the generation of reliability measures that would be impractical to compute manually, thereby advancing the field of precision functional mapping.

[0065] The network detection and confidence mapping system 504 provides additional technical improvements through its integration of both functional network mapping and confidence assessment in a unified computational framework. This integration allows for the simultaneous generation of functional network maps and their associated reliability measures, providing a comprehensive analysis pipeline that enhances the clinical applicability' of functional neuroimaging. The system’s ability to process high-resolution grayordinate data across both cortical and subcortical regions represents an advancement over traditional voxelbased approaches, as it provides improved spatial precision and anatomical accuracy for network mapping. These technical improvements make the network detection and confidence mapping system 504 particularly valuable for applications requiring precise targeting, such as neurosurgical planning and neuromodulation therapy guidance.

[0066] Referring now to FIG. 6, an example of hardware 600 that can be used to implement data source 502, computing device 550, and server 552 in accordance with some embodiments of the systems and methods described in the present disclosure is shown.

[0067] As shown in FIG. 6, in some embodiments, computing device 550 can include a processor 602, a display 604, one or more inputs 606, one or more communication systems16QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660608, and / or memory 610. In some embodiments, processor 602 can be any suitable hardware processor or combination of processors, such as a central processing unit (CPU), a graphics processing unit (GPU), and so on. In some embodiments, display 604 can include any suitable display devices, such as a liquid crystal display (UCD) screen, a light-emitting diode (LED) display, an organic LED (OLED) display, an electrophoretic display (e.g., an “e-ink’‘ display), a computer monitor, a touchscreen, a television, and so on. In some embodiments, inputs 606 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0068] In some embodiments, communications systems 608 can include any suitable hardware, firmware, and / or software for communicating information over communication network 554 and / or any other suitable communication networks. For example, communications systems 608 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 608 can include hardware, firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0069] In some embodiments, memory 610 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 602 to present content using display 604, to communicate with server 552 via communications system(s) 608, and so on. Memory 610 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, memory 610 can include random-access memory (RAM), read-only memory (ROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM), other forms of volatile memory, other forms of non-volatile memory', one or more forms of semivolatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, and so on. In some embodiments, memory 610 can have encoded thereon, or otherwise stored therein, a computer program for controlling operation of computing device 550. In such embodiments, processor 602 can execute at least a portion of the computer program to present content (e.g., images, user interfaces, graphics, tables), receive content from server 552, transmit information to server 552, and so on. For example, the processor 602 and the memory 610 can be configured to perform the methods described herein (e.g., the method of FIG. 1, the method of FIG. 3).

[0070] The processor 602 is specifically configured to execute the computationally intensive operations required for implementing the functional network mapping methods17QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 described with respect to FIG. 1. These operations include computing correlation matrices from time course signal data across thousands of gray ordinates, calculating similarity values between individual subject data and multiple functional network templates, and performing the dual-threshold analysis that refines network assignments. The processor 602 may be optimized for parallel processing operations, such as a multi-core CPU or GPU. to efficiently handle the matrix computations and template matching algorithms that form the core of the functional network identification process. The technical implementation enables real-time or near-realtime processing of high-resolution fMRI data, which represents a significant improvement over traditional manual analysis methods that may require hours or days of processing time.

[0071] The memory 610 is configured to store not only the computer program instructions for executing the disclosed methods, but also the large datasets required for functional network analysis. This includes storage of dense time series data, functional network template data, correlation matrices, and intermediate processing results generated during the permutation-based confidence analysis described with respect to FIG. 3. The memory 610 may include high-speed RAM for active processing operations and larger capacity storage devices for archiving processed results and template data. The memory architecture is designed to support the data-intensive operations of the permutation analysis, which may generate hundreds or thousands of shuffled datasets and their corresponding network assignments for confidence assessment.

[0072] In some embodiments, server 552 can include a processor 612. a display 614. one or more inputs 616, one or more communications systems 618, and / or memory 620. In some embodiments, processor 612 can be any suitable hardware processor or combination of processors, such as a CPU, a GPU, and so on. In some embodiments, display 614 can include any suitable display devices, such as an LCD screen, LED display, OLED display, electrophoretic display, a computer monitor, a touchscreen, a television, and so on. In some embodiments, inputs 616 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0073] In some embodiments, communications systems 618 can include any suitable hardware, firmware, and / or software for communicating information over communication network 554 and / or any other suitable communication networks. For example, communications systems 618 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 618 can include18QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 hardware, firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0074] In some embodiments, memory 620 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 612 to present content using display 614, to communicate with one or more computing devices 550, and so on. Memory 620 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, memory 620 can include RAM, ROM, EPROM, EEPROM, other types of volatile memory, other ty pes of non-volatile memory7, one or more types of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, and so on. In some embodiments, memory 620 can have encoded thereon a server program for controlling operation of server 552. In such embodiments, processor 612 can execute at least a portion of the server program to transmit information and / or content (e.g., data, images, a user interface) to one or more computing devices 550, receive information and / or content from one or more computing devices 550, receive instructions from one or more devices (e.g., a personal computer, a laptop computer, a tablet computer, a smartphone), and so on.

[0075] In some embodiments, the server 552 is configured to perform the methods described in the present disclosure. For example, the processor 612 and memory7620 can be configured to perform the methods described herein (e.g., the method of FIG. 1, the method of FIG. 3).

[0076] The server 552 provides enhanced computational capabilities for implementing the disclosed methods across multiple subjects or in high-throughput environments. The processor 612 may be configured as a high-performance computing system with multiple processors or processor cores specifically optimized for the parallel processing requirements of the permutation-based confidence analysis described in FIG. 3. This server-based implementation enables the processing of large cohorts of subjects simultaneously, generating population-level functional network templates, and performing batch processing of functional network mapping and confidence assessment across multiple datasets. The server architecture provides technical advantages in terms of processing speed, data throughput, and scalability compared to single-device implementations.

[0077] The memory 620 of server 552 is configured to maintain comprehensive databases of functional network templates, processed results, and intermediate analysis products that can be shared across multiple computing devices 550. This centralized data19QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 management approach provides technical improvements in data consistency, version control, and collaborative research capabilities. The server 552 can maintain updated functional network templates derived from large populations, ensuring that individual analyses benefit from the most current and comprehensive template data available. This server-based template management represents a significant advancement over traditional approaches where individual researchers maintain separate, potentially outdated template datasets.

[0078] In some embodiments, data source 502 can include a processor 622, one or more data acquisition systems 624, one or more communications systems 626, and / or memory 628. In some embodiments, processor 622 can be any suitable hardware processor or combination of processors, such as a CPU, a GPU, and so on. In some embodiments, the one or more data acquisition systems 624 are generally configured to acquire data, images, or both, and can include an MRI system. Additionally or alternatively, in some embodiments, the one or more data acquisition systems 624 can include any suitable hardware, firmware, and / or software for coupling to and / or controlling operations of an MRI system. In some embodiments, one or more portions of the data acquisition system(s) 624 can be removable and / or replaceable.

[0079] Note that, although not shown, data source 502 can include any suitable inputs and / or outputs. For example, data source 502 can include input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, a trackpad, a trackball, and so on. As another example, data source 502 can include any suitable display devices, such as an LCD screen, an LED display, an OLED display, an electrophoretic display, a computer monitor, a touchscreen, a television, etc., one or more speakers, and so on.

[0080] In some embodiments, communications systems 626 can include any suitable hardware, firmware, and / or software for communicating information to computing device 550 (and. in some embodiments, over communication network 554 and / or any other suitable communication networks). For example, communications systems 626 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 626 can include hardware, firmware, and / or software that can be used to establish a wired connection using any suitable port and / or communication standard (e.g., VGA, DVI video, USB, RS-232, etc.), Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0081] In some embodiments, memory' 628 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 622 to control the one or more data acquisition systems 624. and / or20QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 receive data from the one or more data acquisition systems 624; to generate images from data; present content (e.g.. data, images, a user interface) using a display; communicate with one or more computing devices 550; and so on. Memory 628 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, memory 628 can include RAM, ROM, EPROM, EEPROM, other ty pes of volatile memory', other types of non-volatile memory, one or more types of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical dnves, and so on. In some embodiments, memory 628 can have encoded thereon, or otherwise stored therein, a program for controlling operation of data source 502. In such embodiments, processor 622 can execute at least a portion of the program to generate images, transmit information and / or content (e.g., data, images, a user interface) to one or more computing devices 550, receive information and / or content from one or more computing devices 550, receive instructions from one or more devices (e.g., a personal computer, a laptop computer, a tablet computer, a smartphone, etc.), and so on.

[0082] In some embodiments, any suitable computer-readable media can be used for storing instructions for performing the functions and / or processes described herein. For example, in some embodiments, computer-readable media can be transitory or non-transitory. For example, non-transitory computer-readable media can include media such as magnetic media (e.g., hard disks, floppy disks), optical media (e g., compact discs, digital video discs, Blu-ray discs), semiconductor media (e.g.. RAM, flash memory, EPROM. EEPROM), any suitable media that is not fleeting or devoid of any semblance of permanence during transmission, and / or any suitable tangible media. As another example, transitory' computer- readable media can include signals on networks, in wires, conductors, optical fibers, circuits, or any suitable media that is fleeting and devoid of any semblance of permanence during transmission, and / or any suitable intangible media.

[0083] As used herein in the context of computer implementation, unless otherwise specified or limited, the terms “component,” “system,” “module.” “framework,” and the like are intended to encompass part or all of computer-related systems that include hardware, software, a combination of hardware and software, or software in execution. For example, a component may be, but is not limited to being, a processor device, a process being executed (or executable) by a processor device, an object, an executable, a thread of execution, a computer program, or a computer. By way of illustration, both an application running on a computer and the computer can be a component. One or more components (or system, module.21QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 and so on) may reside within a process or thread of execution, may be localized on one computer, may be distributed between two or more computers or other processor devices, or may be included within another component (or system, module, and so on).

[0084] In some implementations, devices or systems disclosed herein can be utilized or installed using methods embodying aspects of the disclosure. Correspondingly, description herein of particular features, capabilities, or intended purposes of a device or system is generally intended to inherently include disclosure of a method of using such features for the intended purposes, a method of implementing such capabilities, and a method of installing disclosed (or otherwise known) components to support these purposes or capabilities. Similarly, unless otherwise indicated or limited, discussion herein of any method of manufacturing or using a particular device or system, including installing the device or system, is intended to inherently include disclosure, as embodiments of the disclosure, of the utilized features and implemented capabilities of such device or system.

[0085] Referring particularly now to FIG. 7, an example of an MRI system 700 that can implement the methods described here is illustrated. The MRI system 700 includes an operator workstation 702 that may include a display 704, one or more input devices 706 (e.g., a keyboard, a mouse), and a processor 708. The processor 708 may include a commercially available programmable machine running a commercially available operating system. The operator workstation 702 provides an operator interface that facilitates entering scan parameters into the MRI system 700. The operator w orkstation 702 may be coupled to different servers, including, for example, a pulse sequence server 710, a data acquisition server 712, a data processing server 714, and a data store server 716. The operator workstation 702 and the serv ers 710, 712, 714, and 716 may be connected via a communicarion system 740, which may include wired or wireless network connections.

[0086] The pulse sequence server 710 functions in response to instructions provided by the operator workstation 702 to operate a gradient system 718 and a radiofrequency (“RF”) system 720. Gradient waveforms for performing a prescribed scan are produced and applied to the gradient system 718. which then excites gradient coils in an assembly 722 to produce the magnetic field gradients Gx, G , and G7that are used for spatially encoding magnetic resonance signals. The gradient coil assembly 722 forms part of a magnet assembly 724 that includes a polarizing magnet 726 and a whole-body RF coil 728.

[0087] RF waveforms are applied by the RF system 720 to the RF coil 728, or a separate local coil to perform the prescribed magnetic resonance pulse sequence. Responsive magnetic22QB\920171.00660\98569921.4UMN 2024-350Aty. Docket: 920171.00660 resonance signals detected by the RF coil 728, or a separate local coil, are received by the RF system 720. The responsive magnetic resonance signals may be amplified, demodulated, filtered, and digitized under direction of commands produced by the pulse sequence server 710. The RF system 720 includes an RF transmitter for producing a wide variety of RF pulses used in MRI pulse sequences. The RF transmitter is responsive to the prescribed scan and direction from the pulse sequence server 710 to produce RF pulses of the desired frequency, phase, and pulse amplitude waveform. The generated RF pulses may be applied to the whole-body RF coil 728 or to one or more local coils or coil arrays.

[0088] The RF system 720 also includes one or more RF receiver channels. An RF receiver channel includes an RF preamplifier that amplifies the magnetic resonance signal received by the coil 728 to which it is connected, and a detector that detects and digitizes the I and Q quadrature components of the received magnetic resonance signal. The magnitude of the received magnetic resonance signal may, therefore, be determined at a sampled point by the square root of the sum of the squares of the I and Q components:

[0089] and the phase of the received magnetic resonance signal may also be determined according to the following relationship:

[0090] The pulse sequence server 710 may receive patient data from a physiological acquisition controller 730. By way of example, the physiological acquisition controller 730 may receive signals from a number of different sensors connected to the patient, including electrocardiograph (“ECG”) signals from electrodes, or respiratory signals from a respiratory bellows or other respiratory’ monitoring devices. These signals may be used by the pulse sequence server 710 to synchronize, or “gate.” the performance of the scan with the subject's heart beat or respiration.

[0091] The pulse sequence server 710 may also connect to a scan room interface circuit 732 that receives signals from various sensors associated with the condition of the patient and the magnet system. Through the scan room interface circuit 732. a patient positioning system 734 can receive commands to move the patient to desired positions during the scan.

[0092] The digitized magnetic resonance signal samples produced by the RF system 720 are received by the data acquisition server 712. The data acquisition server 712 operates23QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 in response to instructions downloaded from the operator workstation 702 to receive the realtime magnetic resonance data and provide buffer storage, so that data is not lost by data overrun. In some scans, the data acquisition server 712 passes the acquired magnetic resonance data to the data processor server 714. In scans that require information derived from acquired magnetic resonance data to control the further performance of the scan, the data acquisition server 712 may be programmed to produce such information and convey it to the pulse sequence server 710. For example, during pre-scans, magnetic resonance data may be acquired and used to calibrate the pulse sequence performed by the pulse sequence server 710. As another example, navigator signals may be acquired and used to adjust the operating parameters of the RF system 720 or the gradient system 718, or to control the view order in which k-space is sampled. In still another example, the data acquisition server 712 may also process magnetic resonance signals used to detect the arrival of a contrast agent in a magnetic resonance angiography (“MRA”) scan. For example, the data acquisition server 712 may acquire magnetic resonance data and processes it in real-time to produce information that is used to control the scan.

[0093] The data processing server 714 receives magnetic resonance data from the data acquisition server 712 and processes the magnetic resonance data in accordance with instructions provided by the operator workstation 702. Such processing may include, for example, reconstructing two-dimensional or three-dimensional images by performing a Fourier transformation of raw k-space data, performing other image reconstruction algorithms (e.g.. iterative or backproj ection reconstruction algorithms), applying fdters to raw k-space data or to reconstructed images, generating functional magnetic resonance images, or calculating motion or flow images.

[0094] Images reconstructed by the data processing server 714 are conveyed back to the operator workstation 702 for storage. Real-time images may be stored in a data base memory cache, from which they may be output to operator display 702 or a display 736. Batch mode images or selected real time images may be stored in a host database on disc storage 738. When such images have been reconstructed and transferred to storage, the data processing server 714 may notify the data store server 716 on the operator workstation 702. The operator workstation 702 may be used by an operator to archive the images, produce fdms, or send the images via a network to other facilities.

[0095] The MRI system 700 may also include one or more networked workstations 742. For example, a networked workstation 742 may include a display 744, one or more input24QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 devices 746 (e.g., a keyboard, a mouse), and a processor 748. The networked workstation 742 may be located within the same facility as the operator workstation 702, or in a different facility, such as a different healthcare institution or clinic.

[0096] The networked workstation 742 may gain remote access to the data processing serv er 714 or data store server 716 via the communication system 740. Accordingly, multiple networked workstations 742 may have access to the data processing server 714 and the data store server 716. In this manner, magnetic resonance data, reconstructed images, or other data may be exchanged between the data processing server 714 or the data store server 716 and the networked workstations 742, such that the data or images may be remotely processed by a networked workstation 742.

[0097] The present disclosure has described one or more preferred embodiments, and it should be appreciated that many equivalents, alternatives, variations, and modifications, aside from those expressly stated, are possible and within the scope of the invention.25QB\920171.00660\98569921.4

Claims

UMN 2024-350Aty. Docket: 920171.00660CLAIMS1. A method for generating a functional network map of a somatomotor cognitive action network (SCAN) from magnetic resonance image data acquired from a subject using a magnetic resonance imaging (MRI) system, the method comprising: accessing magnetic resonance image data with a computer system, wherein the magnetic resonance image data comprise a time-series of images whose voxels depict blood-oxygen-level-dependent (BOLD) signals; forming time course signal data for each of a plurality of grayordinates associated with the subject using the computer system, wherein the time course signal data are formed for each of the plurality of gray ordinates as BOLD signals at each of the plurality’ of gray ordinates measured over the time series of images; computing a first correlation matrix from the time course signal data for each of the plurality of gray ordinates using the computer system and a first threshold value; accessing functional network template data with the computer system, wherein the functional network template data comprise at least a SCAN template indicative of grayordinates associated with a SCAN and a motor network template indicative of grayordinates associated with a motor network; computing similarity7values between the first correlation matrix and each functional network template in the functional network template data; generating functional network maps with the computer system by assigning gray ordinates to the functional networks in the functional network data based on the similarity7values; recomputing the similarity values from the functional network maps for each of the plurality of grayordinates using the computer system and second threshold value that is higher than the first threshold value; computing updated similarity’ values for each functional network template in the functional network template data; generating a SCAN map with the computer system by reassigning grayordinates to the SCAN based on the updated similarity values.

2. The method of claim 1 , wherein the similarity’ values are eta-squared (q2) values.26QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.006603. The method of claim 1. wherein the first threshold value comprises z-score values greater than 1.

4. The method of claim 1 or 3, wherein the second threshold value comprises z- scores greater than 3.

5. The method of claim 1, wherein the motor network template comprises a plurality of motor network templates indicative of grayordinates associated with each of a plurality of motor networks.

6. The method of claim 5, wherein the plurality7of motor networks comprises a somatomotor dorsal network and a somatomotor lateral network.

7. The method of claim 1. further comprising generating an updated motor network map by reassigning grayordinates to the motor network based on the updated similarity values.

8. A method for generating a network confidence map that indicates confidence in functional network assignments based on magnetic resonance image data, the method comprising: accessing magnetic resonance image data with a computer system, wherein the magnetic resonance image data comprise a time-series of images whose voxels depict blood-oxygen-level-dependent (BOLD) signals; forming dense time series data from all voxels across the time-series images in the magnetic resonance image data acquired from the subject, wherein the dense time series data are formed by recording temporal fluctuations in the BOLD signals over time; partitioning the dense time series data into a plurality of partitions; generating permutated dense time series data by' shuffling the partitioned dense time series data according to a plurality of permutations; splitting the permutated dense time series data into first shuffled dense time series data and second shuffled dense time series data;27QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.00660 generating first network assignment data from the first shuffled dense time series data and second network assignment data from the second shuffled dense time series data; generating a network confidence map that indicates confidence in functional network assignments based on a comparison of the first network assignment data and the second network assignment data; and outputting the network confidence map with the computer system.

9. The method of claim 8, wherein generating the network confidence map comprises quantifying a stability of template matching network assignments in the first network assignment data and the second network assignment data.

10. The method of claim 9, wherein the stability of the template matching network assignments is quantified based on whether a network assignment for each grayordinate represented in the first and second network assignment data is consistent between the first network assignment data and the second network assignment data.

11. The method of claim 8, wherein generating the network confidence map comprises aggregating dense scalar maps from each of the plurality of permutations.

12. The method of claim 8, wherein partitioning the dense time series data comprises concatenating the dense time series data across a plurality of scan times for the subject to form a unified dense time series data that contains time courses of all brain vertices across all of the plurality of scan times, and wherein the partitioned dense time series data are generated from the unified dense time series data.

13. The method of claim 8, wherein the first network assignment data are generated by template matching the first shuffled dense time series data with at least one functional network template, and where the second network assignment data are generated by template matching the second shuffled dense time series data with the least one functional network template.28QB\920171.00660X98569921.4UMN 2024-350Aty. Docket: 920171.0066014. The method of claim 13, wherein the at least one functional network template comprises a somatomotor cognitive action network (SCAN) template.

15. The method of claim 14, wherein the first network assignment data and the second network assignment data are each updated by reassigning network assignments for each grayordinate in the first network assignment data and the second network assignment data using an updated template matching using a higher threshold value than used in the template matching.

16. The method of claim 8. wherein the first shuffled dense time series data comprise a first split-half of the dense time series data and the second shuffled dense time series data comprise a second split-half of the dense time series data.

17. The method of claim 8. wherein generating the permutated dense time series data comprises sampling the partitioned dense time series data and concatenating the sampled partitions according to a plurality of resampling draws.

18. The method of claim 17, wherein the partitioned dense time series data are resampled with replacement of a selected percentage of the partitioned dense time series data.29QB\920171.00660X98569921.4

Images