Systems and methods for analyzing brain activity and applications thereof
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12402822-B2
Publication Date
2025-09-02
Expiration Date
Abstract
In some embodiments, the present invention provides an exemplary inventive system that includes: an apparatus to record: individual's brain electrical activity, a physiological parameter of the individual, and iii) an environmental parameter; a computer processor configured to perform: obtaining a recording of the electrical signal data; projecting the obtained recording of electrical signal data onto a pre-determined ordering of a denoised optimal set wavelet packet atoms to obtain a set of projections; normalizing the particular set of projections of the individual using a pre-determined set of normalization factors to form a set of normalized projections; determining a personalized mental state of the individual by assigning a brain state; determining a relationship between: the physiological parameter, the environmental parameter, and the personalized mental state; generating an output, including: a visual indication, representative of the personalized mental state, and) a feedback output configured to affect the personalized mental state of the individual.
Core Innovation
A system is configured to record brain electrical activity of an individual with an apparatus and to obtain electrical signal data representative of the brain electrical activity. The system also obtains at least one physiological parameter of the individual and/or at least one environmental parameter. A processor projects the obtained recording of electrical signal data onto a denoised optimal set of wavelet packet atoms to obtain a particular set of projections of the individual.
The denoised optimal set of wavelet packet atoms is determined based at least in part on applying at least one mother wavelet and at least one denoising algorithm to a plurality of individual-specific electrical signal data recordings representative of a plurality of individual-specific brain electrical activities of a plurality of individuals. The system determines at least one mental state of the individual by assigning at least one individual-specific brain state based on applying at least one machine learning algorithm to the particular set of projections of the individual.
The assigned individual-specific brain state is associated with the at least one mental state, at least one neurological condition, or both. The system determines a relationship between the at least one individual-specific brain state and at least one of the at least one physiological parameter or the at least one environmental parameter. Based on the relationship, the system generates an output comprising at least one of a visual indication representative of the at least one individual-specific brain state, or a feedback output expected to affect the at least one individual-specific brain state.
Claims Coverage
The independent claims cover four inventive features: recording brain electrical activity with physiological and/or environmental parameters, projecting the recording onto a denoised optimal set of wavelet packet atoms, assigning individual-specific brain states to determine mental states via machine learning, and generating a visual indication and/or feedback output expected to affect the brain state. Dependent claims further refine threshold-based output behavior, feedback modalities, iterative comparison and repetition, specific machine-learning algorithm types, and particular use cases.
Projecting brain electrical activity onto a denoised optimal set of wavelet packet atoms
Project the obtained recording of electrical signal data onto a denoised optimal set of wavelet packet atoms to obtain a particular set of projections of the individual; the denoised optimal set is determined by applying at least one mother wavelet and at least one denoising algorithm to individual-specific electrical signal data recordings representative of individual-specific brain electrical activities.
Assigning individual-specific brain states to determine mental states using machine learning
Determine at least one mental state of the individual by assigning at least one individual-specific brain state based on applying at least one machine learning algorithm to the particular set of projections; the assigned brain state is associated with the mental state, at least one neurological condition, or both.
Linking individual-specific brain states to physiological and/or environmental parameters
Determine a relationship between the at least one individual-specific brain state of the individual and at least one physiological parameter or at least one environmental parameter.
Generating a visual indication and/or feedback output expected to affect the brain state
Generate an output comprising at least one visual indication representative of the individual-specific brain state or a feedback output expected to affect, based on the relationship, the individual-specific brain state.
Threshold-based change detection for output generation
Generate an output by determining whether at least one physiological parameter or at least one environmental parameter changes beyond a pre-determined threshold.
Iterative comparison and repetition until a desired mental state or coherent response
Determine a desired mental state or coherent response by obtaining first and second mental states with corresponding first and second visual indications and first and second feedback outputs, comparing the visual indications, and repeating the process until the desired mental state or coherent response is obtained.
Selecting machine learning algorithm types for brain state assignment
Apply at least one machine learning algorithm chosen from logistic regression, support vector machine, or deep learning modeling algorithms.
Using feedback output categories including audible, visual, and physically-sensed signals
Generate the feedback output as one or more of an audible signal, a visual signal, a physically-sensed signal, or any combination thereof.
Pediatric musical stimulus-response feedback
Use an individual-specific brain state obtained from a child between years of 0 and 12 in response to a first musical stimulus to generate feedback in the form of at least one second musical stimulus.
Minimally conscious subject stimulus-response feedback
Apply the method to a minimally conscious subject by using an individual-specific brain state that reflects the subject’s response to a stimulus to generate feedback that outputs a second stimulus expected to affect the minimally conscious subject.
The independent claims center on projecting recorded brain electrical activity onto a denoised optimal set of wavelet packet atoms, assigning individual-specific brain states to determine mental states via machine learning, relating brain state to physiological and/or environmental parameters, and generating visual indication and/or feedback output expected to affect the brain state. Dependent claims refine this framework with iterative comparison and repetition, threshold-based change detection, machine-learning algorithm options, feedback modality options, and use cases including pediatric musical stimulus-response and minimally conscious subject stimulus-response.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Detecting or monitoring mental states and neurological conditions associated with individual-specific brain states.
Pediatric 0-12 months musical stimulus-response feedback using an individual-specific brain state.
Minimally conscious subject stimulus-to-stimulus feedback using an individual-specific brain state expected to affect the minimally conscious subject.
Alerting/caregiver monitoring when parameters change beyond a pre-determined threshold.
Generating visual indications and feedback outputs expected to affect an individual-specific brain state for neurofeedback and BCI applications.
Interested in licensing this patent?