# US 2016/0235324 A1 · Methods, Systems, and Apparatus For Self-Calibrating EEG Neurofeedback

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2016/0235324 A1
- Document ID: US20160235324A1
- Status: Published application
- Kind: A1
- Application number: 15/043,909
- Filed: 2016-02-15
- Published: 2016-08-18
- Earliest claimed: 2015-02-14
- Assignees: Massachusetts Institute of Technology
- Applicants: Massachusetts Institute of Technology
- Inventors: Andreas Mershin; Thrasyvoulos Karydis
- Web page: https://patentgrep.com/patent/US20160235324A1

## CPC

- A61B 5/375: using biofeedback
- A61B 5/4824: {Touch or pain perception evaluation (A61B5/4821 takes precedence)}
- A61B 5/6803: {Head-worn items, e.g. helmets, masks, headphones or goggles}
- A61B 5/6831: {Straps, bands or harnesses}
- A61B 5/7257: {using Fourier transforms}
- A61B 5/7267: {involving training the classification device}
- A61B 5/743: {Displaying an image simultaneously with additional graphical information, e.g. symbols, charts, function plots}
- G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Methods, systems, and apparatus implementing a generalizable self-calibrating protocol coupled with machine learning algorithms in an exemplary setting of classifying perceptual states as corresponding to the experience of perceptually opposite mental states (including pain or no pain) are disclosed. An embodiment presented represents inexpensive, commercially available, wearable EEG sensors providing sufficient data fidelity to robustly differentiate the two perceptually opposite states. Low-computational overhead machine learning algorithms that can be run on a mobile platform can be used to find the most efficient feature handles to classify perceptual states as self-calibrated by the user. The invention is generalizable to states beyond just pain and pave the way towards creating EEG NFB applications targeting arbitrary, self-calibrated perceptual states in at-home and wearable settings.

## Claims

### Claim 1

A method of classifying perceptual states as corresponding to the experience of “opposite” states, comprising:
a. teaching a classifier through self-calibration to arbitrary perceptual states, comprising the steps of:
i. extracting and then selecting features from an EEG signal;
ii. labeling the extracted and selected EEG features by indicating when the brain is in a desired (up) or undesired (down) perceptual state;
iii. calculating the classifier;
b. applying the classifier on newly extracted and selected EEG signals to predict which perceptually opposite state the brain is in for each signal.

### Claim 2 (depends on 1)

The method of claim 1, wherein the perceptually opposite perceptual states are specified as pain and no-pain, or zone and no-zone.

### Claim 3 (depends on 1)

The method of claim 1, further comprising the step of c. refining the self-calibration classifier through a feedback loop in which a user answers whether the classifier has correctly identified one or more perceptual states as being “up” or “down”.

### Claim 4 (depends on 1)

The method of claim 1, wherein the classifier is a multi-class classifier capable of predicting multiple non-binary perceptual states.

### Claim 5 (depends on 1)

The method of claim 1, wherein the method of indication in step (ii) is either active where the user performs an action, or passive where a proxy is recorded.

### Claim 6 (depends on 1)

The method of claim 1, wherein the extracted and selected features from an EEG signal are from time series traces of electroencephalographs (EEGs) obtained from a wearable headband.

### Claim 7 (depends on 1)

The method of claim 1, wherein the classifier is an off-the-shelf machine learning algorithm, the machine learning algorithm chosen from one of k-nearest neighbor, support vector machines, neural networks, or decision trees.

### Claim 8 (depends on 1)

The method of claim 1, wherein the states comprise a desired and an undesired state, and calculating the classifier consists of determining the posterior probability for both the desired and undesired states.

### Claim 9 (depends on 1)

The method of claim 1, wherein a computing device is capable of using fast Fourier transforms (FFTs) to act on the labeling by the individual in combination with the time series of brain waives to isolate brain wave patterns that may then be characterized as the first and the second perceived mental state.

### Claim 10 (depends on 1)

The method of claim 1, wherein the EEG signals are transferred to a neurofeedback and display platform via a wired or wireless link.

### Claim 11

A system for classifying perceptual states as corresponding to the experience of perceptually opposite perceptual states, comprising:
an EEG sensor and signal processing platform;
a neurofeedback platform comprising:
an input interface for providing timestamped markers on EEG data used as classifiers for machine learning of perceptually opposite perceptual brain state handles;
a processor to calculate a classifier for the prediction of a perceptually opposite perceptual state for new EEG signals.
a display to view EEG data and the result of classification;
and a link between the EEG sensor and neurofeedback platform.

### Claim 12 (depends on 11)

The system of claim 11, wherein the system further comprises cloud storage and processing.

### Claim 13 (depends on 11)

The system of claim 11, wherein the perceptually opposite perceptual states are specified as pain and no-pain, or zone and no-zone.

### Claim 14 (depends on 11)

The system of claim 11, wherein the classifier is refined through a feedback loop in which a user answers whether the classifier has correctly identified one or more perceptual states as being desired or undesired.

### Claim 15 (depends on 11)

The system of claim 11, wherein the classifier is a multi-class classifier capable of predicting multiple non-binary perceptual states.

### Claim 16 (depends on 11)

The system of claim 11, wherein the timestamped markers are provided actively where the user performs an action, or passively where a proxy is recorded.

### Claim 17 (depends on 11)

The system of claim 11, wherein the EEG sensor and signal processing platform is a wearable headband,

### Claim 18 (depends on 11)

The system of claim 11, wherein the link is wired or wireless.

### Claim 19 (depends on 11)

The system of claim 11, wherein the classifier is an off-the-shelf machine learning algorithm, the machine learning algorithm chosen from one of k-nearest neighbor, support vector machines, neural networks, or decision trees.

### Claim 20

An apparatus for classifying perceptual states as corresponding to the experience of perceptually opposite perceptual states, comprising:
an EEG sensor and signal processing platform;
a neurofeedback platform comprising:
an input interface for providing timestamped markers on EEG data used as classifiers for machine learning of perceptually opposite perceptual state handles;
a processor to calculate a classifier for the prediction of a perceptually opposite perceptual state for new EEG signals.
a display to view EEG data and the result of classification;
and a link between the EEG sensor and neurofeedback platform.

### Claim 21

A neurofeedback therapy system comprising:
a EEG brain wave monitor;
a computing device coupled to the brain wave monitor and configured to record a time series of brain waves of an individual being monitored by the EEG brain wave monitor;
an input interface coupled to the computing device, wherein the input interface permits the individual to label time points during the time series of brain waves;
wherein the computing device further comprises a calibration mode and a display mode, the calibration mode enabling the individual to label, using the input interface, a first perceived mental state and a second perceived mental state distinct from the first perceived mental state during the time series of brain waves; and
the display mode operating subsequent to the calibration mode, presenting to the individual a display of the computing device's determination of whether the individual is in the first or second perceived mental state.

### Claim 22 (depends on 21)

The system of claim 21 wherein the display mode additionally presents to the individual a display of the computing device's prediction of whether the individual is about to enter the first or second perceived mental state.

### Claim 23 (depends on 21)

The system of claim 21 wherein the first and second perceived mental states are perceived opposite mental states of the individual.

### Claim 24 (depends on 21)

The system of claim 21 wherein the computing device is capable of using one or more machine learning algorithms to act on the labeling by the individual in combination with the time series of brain waives to determine brain wave patterns that characterize the first and the second perceived mental state.

### Claim 25 (depends on 21)

The system of claim 21 wherein the machine learning algorithms are chosen from the list consisting of k-nearest neighbor, support vector machines, neural networks, and decision trees.

### Claim 26 (depends on 21)

The system of claim 21 wherein the computing device is capable of using fast Fourier transforms (FFTs) to act on the labeling by the individual in combination with the time series of brain waives to isolate brain wave patterns that may then be characterized as the first and the second perceived mental state.

### Claim 27

A memorization training system comprising:
an EEG brain wave monitor;
an electronic device coupled to the brain wave monitor and configured to record a time series of brain waves of an individual being monitored by the EEG brain wave monitor;
wherein the electronic device further comprises an input interface in a display mode that permits the individual to view the elements to be memorized in a serial way and indicate the status of memorization which can be (a) known element, or (b) unknown element; and
an algorithm trained within an initial calibration phase, capable of identifying whether the user of the system has truthfully memorized the given element.

### Claim 28

A dynamic tuning threshold system comprising:
an EEG brain wave monitor;
an electronic device coupled to the brain wave monitor and configured to record a time series of brain waves of an individual being monitored by the EEG brain wave monitor;
a computing device configured to train an algorithm capable of classifying perceptual states;
an input interface coupled to the computing device, wherein the input interface permits the individual to label time points during the time series of brain waves;
wherein the result of the training is a separating hyperplane, the hyperplane being used to achieve a directional arrow during a neurofeedback reward protocol that points towards the direction corresponding to being in one of two perceptual states by rewarding a highly mobile perceptual state.
