Methods for modeling neurological development and diagnosing a neurological impairment of a patient
Inventors
Le, Tan • Mackellar, Geoffrey Ross
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12036030-B2
Publication Date
2024-07-16
Expiration Date
Abstract
One variation of a method for modeling neurological development includes: aggregating electroencephalography (EEG) data that comprise multiple EEG signals of each user in a set of users, EEG signals of each user recorded on multiple distinct dates, the set of users comprising a plurality of users of various known neurological statuses; identifying a synchronization pattern trend within the EEG data of the set of users; and correlating the synchronization pattern trend with neurological development within the set of users.
Core Innovation
The invention relates to EEG signals recorded from users using a biosignal neuroheadset associated with each user and comprising a set of electrodes. For each EEG signal, the method determines a set of synchronization pattern features based on the EEG signal, and for the set of users it organizes the synchronization pattern feature sets into groups.
Each group is assigned a tag, where the tag is associated with a user state or a neurological status tag. The method determines a model for each group based on the tagged groups, and then uses the determined models to determine a predicted tag for a new user.
Using the predicted tag, the method recommends an activity for the new user to perform. The disclosed approach also includes determining synchronization-pattern features and trends from EEG to learn neurological development and impairment-related models, and using model comparison to diagnose and/or track neurological connectivity changes over time.
The disclosed system further associates user state with neurological status tags by determining the user state from user input and/or measurements received from physiological and/or environmental sensors, and attaching these tags to EEG for correlation and model building, optionally including mapping neural connectivity from EEG sub-signals and integrating additional neurological data.
Claims Coverage
The independent claim covers an end-to-end computer-implemented pipeline that records EEG with a biosignal neuroheadset, derives synchronization pattern features, groups feature sets and assigns user state tags, determines a model per group, predicts a tag for a new user using the determined models, and recommends an activity based on the predicted tag. Dependent claims further specify correlation-based prediction, grouping by clustering, neurological status tags, and user state determination using user input and/or sensor measurements.
Eeg synchronization pattern feature extraction using a biosignal neuroheadset
Recording a set of EEG signals from a user using a biosignal neuroheadset associated with the user, the biosignal neuroheadset comprising a set of electrodes, and determining a set of synchronization pattern features based on each EEG signal.
Grouping synchronization pattern features and assigning user state tags
Organizing the sets of synchronization pattern features into a set of groups and assigning a tag to each group of the set of groups, the tag associated with a user state.
Model determination per group based on tagged groups
Determining a model for each group of the set of groups based on the tagged groups.
Predicted tag determination for a new user using the determined models
Using the determined models to determine a predicted tag for a new user.
Activity recommendation based on the predicted tag
Recommending an activity for the new user to perform based on the predicted tag.
Correlation-based prediction for determining a predicted tag
Determining a predicted tag by comparing new user synchronization pattern features with determined models to compute correlations, and using the correlations to determine the predicted tag.
Group model training by aggregating feature sets
Determining the model for each group by aggregating synchronization pattern feature sets across the group.
Feature aggregation by clustering
Aggregating synchronization pattern features by clustering the synchronization pattern features.
Neurological status tag as the tag type
Using the tag as a neurological status tag.
User state determination from user input and/or sensor measurements
For each group, determining the user state associated with a tag based on at least one of a user input or a measurement received from a sensor.
Across the independent and dependent claims, the core inventive theme is tagging a user state from EEG synchronization pattern features by grouping feature sets, assigning tagged user states, determining models per group, predicting a tag for a new user using the determined models, and recommending an activity based on the predicted tag. Additional claim coverage constrains tags to neurological status and specifies that user state can be based on user input and/or sensor measurements.
Stated Advantages
Recommending an activity for a new user to perform based on the predicted tag.
Documented Applications
Modeling neurological development by aggregating multi-date EEG across users with known neurological statuses and learning synchronization-pattern trends for neurological impairment and connectivity-related modeling.
Diagnosing a neurological impairment of a patient by extracting synchronization-pattern trends from patient EEG and comparing them to a learned impairment model, including optional tracking of neural connectivity changes over time.
Automatically tagging EEG signals by determining user actions/status from physiological and/or environmental sensors and attaching these tags to EEG to improve correlation and model building.
Mapping neural connectivity from EEG sub-signals and optionally integrating additional neurological data in remote/feedback workflows.
Interested in licensing this patent?