# US 12,178,588 B2 · Method for detecting and localizing brain silences using EEG > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,178,588 B2 - Document ID: US12178588B2 - Status: Granted - Kind: B2 - Application number: 18/410,446 - Filed: 2024-01-11 - Published: 2024-08-01 - Granted: 2024-12-31 - Earliest claimed: 2020-03-20 - Assignees: Carnegie Mellon University - Applicants: Carnegie Mellon University - Inventors: Pulkit Grover; Alireza Chamanzar; Marlene Behrmann Cohen - Web page: https://patentgrep.com/patent/US12178588B2 ## CPC - A61B 5/372: Analysis of electroencephalograms - A61B 5/0042: {for the brain} - A61B 5/055: involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging - A61B 5/374: Detecting the frequency distribution of signals, e.g. detecting delta, theta, alpha, beta or gamma waves - A61B 5/7246: {using correlation, e.g. template matching or determination of similarity} - A61B 5/7264: {Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems} - A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D) ## Abstract A novel method for using the widely-used electroencephalography (EEG) systems to detect and localize silences in the brain is disclosed. The method detects the absence of electrophysiological signals, or neural silences, using noninvasive scalp electroencephalography (EEG) signals. This method can also be used for reduced activity localization, activity level mapping throughout the brain, as well as mapping activity levels in different frequency bands. By accounting for the contributions of different sources to the power of the recorded signals and using a hemispheric baseline approach and a convex spectral clustering framework, the method permits rapid detection and localization of regions of silence in the brain using a relatively small amount of EEG data. ## Claims ### Claim 1 A method for localizing regions of silence in a brain comprising: localizing one or more contiguous regions of silence in a low-resolution grid of uncorrelated sources in the brain; using the one or more localized contiguous regions of silence in the low- resolution grid as an initial estimate of the location of the silence in a high-resolution grid of the sources in the brain; and iteratively localizing the one or more regions of silence in the high-resolution grid of correlated sources based on a covariance of the sources until a center-of-mass for each region of silence converges; wherein the contribution of each of the sources to the power component of the collected EEG signals is determined using a hemispheric baseline. ### Claim 2 (depends on 1) The method of claim 1 wherein the hemispheric baseline comprises: a ratio of a measured contribution in one hemisphere of the brain to a mirrored source from an opposite hemisphere of the brain. ### Claim 3 (depends on 2) The method of claim 2 wherein the mirrored sources are from a hemisphere of the brain having no regions of silence. ### Claim 4 (depends on 3) The method of claim 3 wherein the mirrored sources are measured with electrodes placed symmetrically with respect to a longitudinal fissure of the brain. ### Claim 5 (depends on 3) The method of claim 3 wherein readings from both hemispheres of the brain are recorded with respect to a reference electrode placed at a longitudinal fissure of the brain. ### Claim 6 (depends on 1) The method of claim 1 wherein the low-resolution grid is a coarse discretization of a cortex of the brain in which spatially-uncorrelated sources are localized. ### Claim 7 (depends on 1) The method of claim 1 wherein the high-resolution grid is a fine discretization of a cortex of the brain in which the sources are correlated. ### Claim 8 (depends on 1) The method of claim 1 wherein localizing the one or more contiguous regions of silence in the low-resolution grid further comprises: obtaining EEG readings of both hemispheres of the brain using a plurality of electrodes; determining, for each source, a cross-correlation coefficient indicating the contribution of the source to readings from each of the plurality of electrodes; calculating a normalized variance for each of the cross-correlation coefficients; deriving, based on the normalized variance, a source contribution measure for the low-resolution grid for each of the sources; and localizing silent sources in the low-resolution grid using convex spectral clustering based on the source contribution measures for the low-resolution grid. ### Claim 9 (depends on 8) The method of claim 8 wherein the cross-correlation coefficient is a ratio of the contribution of each uncorrelated source to a contribution of respective mirrored sources in the hemispheric baseline. ### Claim 10 (depends on 9) The method of claim 9 wherein a region of silence is indicated when the cross-correlation coefficient is close to zero. ### Claim 11 (depends on 1) The method of claim 1 wherein localizing the one or more regions of silence in the high-resolution grid further comprises: initializing a source variance variable, an exponential decay coefficient and a set containing a plurality of sources contributing to the one or more regions of silence; performing an iteration of the following steps: determining a power ratio for each of the plurality of electrodes indicating a reduction in the power of the electrode due to the one or more regions of silence; determining a set of electrodes having a minimum reduction in power; based on the signals from the set of electrodes having a minimum power reduction, and values of the source variance variable and the exponential decay coefficient from a previous iteration, recalculating the source variance variable and the exponential decay coefficient; calculating a source covariance matrix based on the source variance variable and the exponential decay coefficient; calculating a source contribution measure for the high-resolution grid for each of the sources based on the covariance matrix; localizing silent sources in the high-resolution grid using convex spectral clustering based on the source contribution measures for the high- resolution grid; and updating the set of silence sources. ### Claim 12 (depends on 1) The method of claim 1 further comprising: estimating a size of the one or more regions of silence. ### Claim 13 (depends on 1) The method of claim 1 wherein the iteration continues for a pre-determined number of iterations. ### Claim 14 (depends on 1) The method of claim 1 further comprising: calculating a source contribution measure for the low-resolution grid and the high-resolution grid by comparison of respective source contribution measures to the source contribution measure for the respective mirrored source in the hemispheric baseline.