Brain state advisory system using calibrated metrics and optimal time-series decomposition

Patent metadata

Field	Value
Publication number	US-10531806-B2
Application number	US-201414574320-A
Country	US
Kind code	B2
Filing date	Dec 17, 2014
Priority date	Dec 17, 2013
Publication date	Jan 14, 2020
Grant date	Jan 14, 2020

Abstract

Various examples are provided for brain state advisory systems using calibrated metrics and optimal time-series decompositions. In one embodiment, a method includes monitoring brainwave activity of a subject and classifying a brain state of the subject based upon the brainwave activity and a model of the brainwave activity. In another embodiment, brain states of the subject are modeled based upon the brainwave activity of the subject.

First claim

Therefore, at least the following is claimed:

1. A method for determination of a clinical diagnostic condition of a subject's brain through a brain computer interface, comprising:
monitoring brainwave activity of a subject at multiple spatial locations via neural electric recordings, the brainwave activity being monitored at multiple temporal scales, where the brainwave activity comprises spike trains and local field potentials obtained via multichannel neural monitoring of the subject;
identifying, by at least one computing device, phasic events from neural waveforms of the brainwave activity, where the phasic events comprise temporally localized brain rhythms of the neural waveforms that are identified based upon a subject specific dictionary of filters associated with the subject;
generating, by the at least one computing device, a multidimensional metric by:
spatiotemporal integration of the phasic events, and
projection of results of the spatiotemporal integration to a lower dimensional space using metric learning to generate the multidimensional metric;

classifying, by the at least one computing device, a brain state of the subject based upon the multidimensional metric and a subject specific dynamical model of the brainwave activity of the subject, the brain state being a health state or a disease state of the subject; and
generating, by the at least one computing device, an indication of the brain state of the subject, the indication identifying the health state or disease state of the subject.

2. The method of claim 1 , wherein the brain state is classified based upon the subject specific dynamical model estimated through blind signal processing of the brainwave activity of that subject.

3. The method of claim 2 , wherein the blind signal processing is based upon blind waveform estimation of a single channel of the brainwave activity of the subject.

4. The method of claim 2 , wherein the blind signal processing is based upon separate blind waveform estimation of multiple individual channels of the brainwave activity of the subject.

5. The method of claim 1 , wherein the subject specific dictionary of filters comprises a set of filters determined through unsupervised blind waveform estimation of the brainwave activity of the subject by the at least one computing device.

6. The method of claim 5 , wherein the set of filters comprise temporal and spatial filters.

7. The method of claim 1 , wherein the brain state is classified with a temporal precision given by a sampling rate used for monitoring the brainwave activity.

8. The method of claim 1 , wherein the phasic events are identified based upon the subject specific dictionary of filters.

9. The method of claim 1 , wherein the spatiotemporal integration comprises pooling of energy between adjacent channels with similar neural waveforms.

10. The method of claim 1 , wherein the projection to the lower dimensional space is implemented using a supervised dimensionality reduction algorithm.

11. The method of claim 1 , wherein the brain state of the subject is classified based at least in part upon a projection of the subject specific dynamical model.

12. The method of claim 1 , wherein the brain state of the subject is classified based at least in part upon landmark points corresponding to classified brain states of the subject.

13. A method for determination of a clinical diagnostic condition of a subject's brain through a brain computer interface, comprising:
monitoring brainwave activity of a subject at multiple spatial locations, the brainwave activity being monitored by recording neural signals at multiple temporal scales, where the brainwave activity comprises spike trains and local field potentials obtained via multichannel neural monitoring of the subject;
decomposing, by at least one computing device, the neural signals of the brainwave activity to generate filtered components based upon a subject specific dictionary of filters associated with the subject, the subject specific dictionary of filters comprising filters determined through unsupervised blind waveform estimation of the brainwave activity of that subject by the at least one computing device;
modeling, by the at least one computing device, brain states of the subject based at least in part upon the filtered components from the brainwave activity and timing of the spike trains to provide a subject specific dynamical model of the brainwave activity of the subject;
classifying, by the at least one computing device, a brain state of the subject based upon a multidimensional metric determined from the brainwave activity and the subject specific dynamical model of the brainwave activity, the brain state being a health state or a disease state of the subject, where the multidimensional metric is determined from spatiotemporal integration of the filtered components and projection of results of the spatiotemporal integration to a lower dimensional space using metric learning; and
generating, by the at least one computing device, an indication of the brain state of the subject, the indication identifying the health state or disease state of the subject.

14. The method of claim 13 , wherein the filtered components comprise phasic events identified based upon the subject specific dictionary of filters.

15. The method of claim 13 , wherein the filters of the subject specific dictionary of filters comprise waveforms tuned to neural signals of the brainwave activity of the subject.

16. The method of claim 15 , wherein the neural signals comprise monitored EEG signals of the subject.

17. The method of claim 13 , wherein the brain states of the subject are modeled using the multidimensional metric.

18. The method of claim 8 , wherein the neural waveforms comprise EEG waveforms.

Assignees

• Univ Florida

Inventors

• Principe Jose
• Brockmeier Austin J

Classifications

• A61B5/16

  Devices for psychotechnics (using teaching or educational appliances G09B1/00 - G09B7/00); Testing reaction times {; Devices for evaluating the psychological state} · CPC title

• A61B5/742 Primary

  using visual displays (displays for heart-related electrical signals, e.g. ECG, A61B5/339) · CPC title

• A61B5/04012 Primary

  Human Necessities · mapped topic

• A61B5/0476

  Human Necessities · mapped topic

• A61B5/37

  Intracranial electroencephalography [IC-EEG], e.g. electrocorticography [ECoG] · CPC title