Processing

Please wait...

PATENTSCOPE will be unavailable a few hours for maintenance reason on Tuesday 26.10.2021 at 12:00 PM CEST
Settings

Settings

Goto Application

1. CA2991350 - EEG BRAIN-COMPUTER INTERFACE PLATFORM AND PROCESS FOR DETECTION OF CHANGES TO MENTAL STATE

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]
WHAT IS CLAIMED IS:
1. A system for a brain-computer interface (BCI) comprising: an output unit configured to trigger a series of mental tasks for the patient; a device having a plurality of electrodes to continuously capture real-time raw electroencephalography (EEG) data from a patient; a server having: an acquisition unit configured to receive the electrode data; a processor configured to detect, using the electrode data, real-time changes in brain-state of the patient in response to the series of mental tasks for the patient, the processor configured to: generate a set of features based upon a frequency domain analysis of the EEG data; reduce the dimensionality of the set of features by implementing a feature clustering process to account for redundancy in EEG signal features of the EEG data; and classify the features into a mental state; a presentation unit configured to generate visual elements for an interface in real-time; and a display controller configured to issue control commands to update the interface using the generated visual elements; and a display device to display and update the interface with the visual elements based on the issued control commands from the server.
2. The system of claim 1, wherein the display device is part of a virtual reality headset.
3. The system of claim 1, wherein the visual elements are part of a train animation that moves in response to the real-time changes in brain-state of the patient.
4. The system of claim 1, wherein the visual elements are part of game that
moves one or more features of a virtual character in response to the real-time changes in brain-state of the patient.
5. The system of claim 1, wherein the interface comprises a topographic map
representing a plurality of portions of a brain of the patient, wherein the visual elements comprise an overlay of activated portions based on the real-time changes in brain-state of the patient.
6. The system of claim 1, wherein the device having the plurality of electrodes or electrodes is an in-ear electroencephalography device having an over-ear support arm and an earpiece.
7. The system of claim 6, wherein the earpiece has two electrodes and the overear support arm has a reference electrode.
8. The system of claim 1, further comprising a collector device coupled to the plurality of electrodes for pre-processing the real-time raw EEG data and correlating to the series of mental tasks on a common timeline.
9. The system of claim 1, wherein the output unit is configured to attempt to elicit a change in a mental state by a sequential trigger of the series of mental tasks based on a dynamic selection specific to the patient, wherein: each mental task is selected from the set of an arithmetic task, an anagram task, and a grid-recall task; each mental task is to be performed within a time period for that mental task;
and each mental task type comprises at least two difficulty levels ranging from easy to difficult.
10. The system of claim 9, wherein: the time period for each arithmetic task is 30 seconds; the time period for each anagram task is 20 seconds; and the time period for each grid-recall task is 15 seconds.
11. The system of claim 10, wherein: the number of difficulty levels for each task type is five; the difficulty levels for each task type comprise very easy, easy, average, difficult and very difficult; and
the mental tasks are triggered in pseudo-random order with some clustering of very easy tasks and very difficult tasks.
12. The system of claim 1, wherein the visual elements represent the mental state of the patient displayed on the display device.
13. The system of claim 1, wherein the mental state of the patient is monitored via passive BCI monitoring in parallel with active BCI monitoring.
14. The system of claim 1, where the output unit and the display device are the same device.
15. The system of claim 1, wherein the processor is further configured to oversample data collected at a more contemporaneous time, such that the data collected at the more contemporaneous time is weighted more heavily than historical data.
16. The system of claim 1, wherein to generate the set of features, the processor is further configured to: compute a fast Fourier Transform (FFT) for each signal received from each electrode, resulting in a frequency spectrum; and compute a total spectral power within non-overlapping frequency ranges in the frequency spectra.
17. The system of claim 15, wherein the non-overlapping frequency ranges comprise a one Hz frequency range from zero-one Hz to 29-30 Hz.
18. The system of claim 1, wherein each spectral power measurement comprises a feature for classification.
19. The system of claim 1, wherein to reduce the dimensionality of the set of features, the processor is further configured to: apply a clustering process to group the features from each electrode into data-
sensitive frequency bands; and apply a fast correlation-based filter to select between two and 20 features for classification.
20. The system of claim 1, wherein to classify the features into mental states, the processor is further configured to: apply a shrinkage linear discrimination analysis to the frequency spectra data of selected features for classification; and determine the mental state based on the frequency ranges having higher spectral power.
21. The system of claim 1, wherein the mental state is one of: fatigue when features originated from frontal and central electrodes;
frustration when features originated from alpha band activity from posterior electrodes and other electrodes in the central and frontal regions; and attention when features originated from alpha band activity from posterior electrodes and not from other electrodes in the central and frontal regions.
22. A method of detecting a mental state from multichannel EEG data continuously received from electrodes located relative to a patient, the method comprising: at a processor, generating a set of features based upon a frequency domain analysis of the EEG data; reducing the dimensionality of the set of features using a feature clustering process to account for redundancy in EEG signal features of the EEG data; classifying the features into a mental state; generating visual elements for an interface in real-time, the visual elements representing the mental state of the patient; and triggering the display of the visual elements for the interface on a display device.
23. The method as claimed in claim 22, further comprising, at the processor, oversampling data collected at a more contemporaneous time, such that the data
collected at the more contemporaneous time is weighted more heavily than historical data.
24. The method as claimed in claim 22, wherein the step of generating a set of features comprises: computing a fast Fourier Transform (FFT) for each signal received from each electrode, resulting in a frequency spectrum; and computing a total spectral power within non-overlapping frequency ranges in the frequency spectra.
25. The method as claimed in claim 24, wherein the non-overlapping frequency ranges comprise a one Hz frequency range from zero-one Hz to 29-30 Hz.
26. The method as claimed in claim 22, wherein each spectral power measurement comprises a feature for classification.
27. The method as claimed in claim 22, wherein the step of reducing the dimensionality comprises: applying a clustering process to group the features from each electrode into data-sensitive frequency bands; and applying a fast correlation-based filter to select between two and 20 features for classification.
28. The method as claimed in claim 22, wherein the step of classifying the features into the mental state comprises:
applying a shrinkage linear discrimination analysis to the frequency spectra data of selected features for classification; and determining the mental state based on the frequency ranges having higher spectral power.
29. The method as claimed in claim 28, wherein the mental state is one of: fatigue when features originated from frontal and central electrodes; frustration when features originated from alpha band activity from posterior electrodes and other electrodes in the central and frontal regions; and attention when features originated from alpha band activity from posterior electrodes and not from other electrodes in the central and frontal regions.
30. The method as claimed in claim 22, further comprising, at a collector device, pre-processing the real-time electrode data and correlating to the series of mental tasks on a common timeline.
31. A non-transitory computer-readable storage medium comprising computerexecutable instructions for causing a processor to: generate a set of features based upon a frequency domain analysis of the EEG data; reduce the dimensionality of the set of features using a feature clustering process to account for redundancy in EEG signal features of the EEG data; classify the features into a mental state;
generate visual elements for an interface in real-time, the visual elements representing the mental state of the patient; and trigger the display of the visual elements for the interface on a display device.
32. A processing device for real-time brain monitoring comprising: a network interface for acquisition of real-time raw EEG data for a patient's brain; a server for processing the real-time raw EEG data to compute real-time changes in brain state of the patient using feature clustering to account for redundancy in EEG signal features of the EEG data; a storage device for storing the real-time changes in brain state of the patient; and a display device having the interface to generate and update a visual representation of the real-time changes in brain state of the patient based on
the issued control commands from the server.
33. A system for a brain-computer interface (BCI) comprising: a device having a plurality of electrodes to continuously capture real-time raw electroencephalography (EEG) data from a patient; a server having: an acquisition unit configured to receive the electrode data; and a processor configured to detect, using the electrode data, real-time changes in brain-state of the patient in response to the series of mental tasks for the patient, the processor configured to:
generate a set of features base upon a frequency domain analysis of the EEG data; reduce the dimensionality of the set of features by implementing a feature clustering process to account for redundancy in EEG signal features of the EEG data; and classify the features into a mental state.
34. The system of claim 33, further comprising a display device that is part of a virtual reality headset.
35. The system of claim 33, wherein the server further comprises: a presentation unit to generate visual elements for an interface in real-time, the visual elements representing real-time changes in brain-state of the patient, wherein the visual elements are part of a train animation that moves in response to the real-time changes in brain-state of the patient; and a display controller to issue control commands to update the interface using the generated visual elements.
36. The system of claim 33, wherein the visual elements are part of game that moves one or more features of a virtual character in response to the real-time changes in brain-state of the patient.
37. The system of claim 33, wherein the interface comprises a topographic map representing a plurality of portions of a brain of the patient, wherein the visual elements
comprise an overlay of activated portions based on the real-time changes in brain-state of the patient.
38. The system of claim 33, wherein the device having the plurality of electrodes or electrodes is an in-ear electroencephalography device having an over-ear support arm and an earpiece.
39. The system of claim 38, wherein the earpiece has two electrodes and the overear support arm has a reference electrode.
40. The system of claim 33, further comprising a collector device coupled to the plurality of electrodes for pre-processing the real-time raw EGG data and correlating to the series of mental tasks on a common timeline.
41. The system of claim 33, wherein the output unit is configured to attempt to elicit a change in a mental state by a sequential trigger of the series of mental tasks based on a dynamic selection specific to the patient, wherein: each mental task is selected from the set of an arithmetic task, an anagram task, and a grid-recall task; each mental task is to be performed within a time period for that mental task;
and each mental task type comprises at least two difficulty levels ranging from easy to difficult.
42. The system of claim 41, wherein: the time period for each arithmetic task is 30 seconds; the time period for each anagram task is 20 seconds; and the time period for each grid-recall task is 15 seconds.
43. The system of claim 42, wherein: the number of difficulty levels for each task type is five; the difficulty levels for each task type comprise very easy, easy, average, difficult and very difficult; and the mental tasks are triggered in pseudo-random order with some clustering of very easy tasks and very difficult tasks.
44. The system of claim 35, wherein the visual elements represent the mental state of the patient displayed on the display device.
45. The system of claim 33, wherein the mental state of the patient is monitored via passive BCI monitoring in parallel with active BCI monitoring.
46. The system of claim 33, wherein the processor is further configured to oversample data collected at a more contemporaneous time, such that the data collected at the more contemporaneous time is weighted more heavily than historical data.
47. The system of claim 33, wherein to generate the set of features, the processor is further configured to: compute a fast Fourier Transform (FFT) for each signal received from each electrode, resulting in a frequency spectrum; and compute a total spectral power within non-overlapping frequency ranges in the frequency spectra.
48. The system of claim 47, wherein the non-overlapping frequency ranges comprise a one Hz frequency range from zero-one Hz to 29-30 Hz. 18. The system of claim 33, wherein each spectral power measurement comprises a feature for classification.
49. The system of claim 33, wherein to reduce the dimensionality of the set of features, the processor is further configured to: apply a clustering process to group the features from each electrode into data-
sensitive frequency bands; and apply a fast correlation-based filter to select between two and 20 features for classification.
50. The system of claim 33, wherein to classify the features into mental states, the processor is further configured to:
apply a shrinkage linear discrimination analysis to the frequency spectra data of selected features for classification; and determine the mental state based on the frequency ranges having higher spectral power.
51. The system of claim 50, wherein the mental state is one of: fatigue when features originated from frontal and central electrodes; frustration when features originated from alpha band activity from posterior electrodes and other electrodes in the central and frontal regions; and attention when features originated from alpha band activity from posterior electrodes and not from other electrodes in the central and frontal regions.
52. A method of detecting a mental state from multichannel EEG data continuously received from electrodes located relative to a patient, the method comprising: at a processor, generating a set of features based upon a frequency domain analysis of the EEG data; reducing the dimensionality of the set of features using a feature clustering process to account for redundancy in EEG signal features of the EEG data; and classifying the features into a mental state.
53. The method as claimed in claim 52, further comprising:
generating visual elements for an interface in real-time, the visual elements representing the mental state of the patient; and triggering the display of the visual elements for the interface on a display device.
54. The method as claimed in claim 52, further comprising, at the processor, oversampling data collected at a more contemporaneous time, such that the data
collected at the more contemporaneous time is weighted more heavily than historical data.
55. The method as claimed in claim 52, wherein the step of generating a set of features comprises: computing a fast Fourier Transform (FFT) for each signal received from each electrode, resulting in a frequency spectrum; and computing a total spectral power within non-overlapping frequency ranges in the frequency spectra.
56. The method as claimed in claim 55, wherein the non-overlapping frequency ranges comprise a one Hz frequency range from zero-one Hz to 29-30 Hz.
57. The method as claimed in claim 55, wherein each spectral power measurement comprises a feature for classification.
58. The method as claimed in claim 52, wherein the step of reducing the dimensionality comprises: applying a clustering process to group the features from each electrode into data-sensitive frequency bands; and applying a fast correlation-based filter to select between two and 20 features for classification.
59. The method as claimed in claim 52, wherein the step of classifying the features into the mental state comprises: applying a shrinkage linear discrimination analysis to the frequency spectra data of selected features for classification; and determining the mental state based on the frequency ranges having higher spectral power.
60. The method as claimed in claim 59, wherein the mental state is one of: fatigue when features originated from frontal and central electrodes; frustration when features originated from alpha band activity from posterior electrodes and other electrodes in the central and frontal regions; and attention when features originated from alpha band activity from posterior electrodes and not from other electrodes in the central and frontal regions.
61. The method as claimed in claim 52, further comprising, at a collector device, pre-processing the real-time electrode data and correlating to the series of mental tasks on a common timeline.
62. A non-transitory computer-readable storage medium comprising computerexecutable instructions for causing a processor to: generate a set of features based upon a frequency domain analysis of the EEG data; reduce the dimensionality of the set of features using a feature clustering process to account for redundancy in EEG signal features of the EEG data; and classify the features into a mental state.