Method for the quantification of human alertness
First Claim
Patent Images
1. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
- selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals at a sampling rate of up to 256 samples per second from a plurality of subjects during said at least one input class of states of alertness to provide EEG signal data;
selecting a baseline subset of EEG signal data from said EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
3 Assignments
0 Petitions
Accused Products
Abstract
The method for the quantification of EEG waveforms along the alertness continuum involves collecting and transforming EEG signal data, identifying and rejecting or decontaminating epochs containing various artifacts, and classifying individual EEG patterns along an alertness-drowsiness continuum. The results of the multi-level classification system are applied in real-time to provide feedback to the user via an audio or visual alarm, or are recorded for subsequent off-line analysis.
145 Citations
89 Claims
-
1. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals at a sampling rate of up to 256 samples per second from a plurality of subjects during said at least one input class of states of alertness to provide EEG signal data;
selecting a baseline subset of EEG signal data from said EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person based upon a classification model of alertness along an alertness-drowsiness continuum, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG; and
determining a classification of said EEG signals of the person along the alertness-drowsiness continuum based upon the person'"'"'s classification model. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 89)
increasing an EMG counter by one;
reading the EMG counter; and
providing an alarm to the subject when the EMG counter exceeds a predetermined threshold.
-
-
26. The method of claim 10, further comprising the step of identifying an artifact of gross head, eye or body movement in said EEG signals.
-
27. The method of claim 26, wherein if significant gross head, eye or body movement is identified in more than one overlap, further comprising the steps of:
-
increasing a movement counter by one;
reading the movement counter; and
providing an alarm to the subject when the movement counter exceeds a predetermined threshold.
-
-
28. The method of claim 10, further comprising the step of identifying epochs contaminated with high levels of environmental electrical interference.
-
29. The method of claim 28, wherein if a high level of environmental electrical interference is identified, further comprising the steps of:
-
increasing an environmental interference counter by one;
reading the environmental interference counter;
measuring the impedance of the electrode if the environmental electrical interference counter exceeds an environmental electrical interference artifact threshold for any electrode;
comparing the impedance of the electrode with an impedance threshold; and
providing an alarm if the impedance value exceeds the impedance threshold.
-
-
30. The method of claim 10, wherein said step of determining a classification of said EEG signals is based upon discriminant function analysis.
-
31. The method of claim 10, wherein said step of determining a classification of said EEG signals is based upon neural network analysis.
-
32. The method of claim 10, wherein said step of determining a classification of said EEG signals is based upon linear statistical analyses using categorical variables.
-
33. The method of claim 10, wherein said step of determining a classification of said EEG signals is based upon non-linear statistical analyses using categorical variables.
-
34. The method of claim 10, wherein said step of determining a classification of said EEG signals comprises determining a classification of said EEG signals according to a plurality of classes for each epoch.
-
35. The method of claim 34, wherein said step of determining a classification of said EEG signals further comprises determining a second level classification of the state of alertness for each epoch.
-
36. The method of claim 34, wherein said step of comparing said classification with a predetermined threshold classification of alertness comprises comparing said classification with said predetermined threshold classification of alertness for a plurality of epochs to determine a classification of the state of alertness for each of said plurality of epochs.
-
37. The method of claim 10, wherein said step of determining a classification of said EEG signals comprises classifying a first level classification of the state of alertness for each epoch as High Vigilance, Low Vigilance, Eyes Closed, or Sleepy.
-
38. The method of claim 10, wherein said step of determining a classification of said EEG signals comprises using the High Vigilance, Low Vigilance and Eyes Closed conditions to adjust the classification model to correctly classify Sleepy.
-
39. The method of claim 10, wherein said step of determining said classification of said EEG signals of the person along the alertness-drowsiness continuum comprises carrying out said step of comparing in real-time.
-
40. The method of claim 39, further comprising the step of providing an alarm to the person when said classification based upon said threshold classification of alertness.
-
41. The method of claim 40, wherein said alarm is a visual alarm.
-
42. The method of claim 40, wherein said alarm is an audio alarm.
-
43. The method of claim 42, wherein said audio alarm comprises spoken words with verbal instructions.
-
44. The method of claim 10, wherein at least one input class of baseline states of alertness from the subject and the database of EEG signal data from said EEG signals from said plurality of subjects are used to predict a state of alertness which corresponds with the subject'"'"'s onset of sleep.
-
45. The method of claim 10, wherein said step of determining a classification of said EEG signals of the person along the alertness-drowsiness continuum comprises determining at least one predictive variable to predict a state of alertness which corresponds with the person'"'"'s onset of sleep based upon at least one input class of states of alertness selected from states of alertness while sitting quietly with eyes open, while sitting quietly with eyes closed, and during a mental performance task.
-
46. The method of claim 10, wherein said database of EEG signals of a plurality of subjects is based upon at least one input class of states of alertness comprising a state of alertness derived during a mental performance task.
-
47. The method of claim 10, wherein said database of EEG signals of a plurality of subjects is based upon at least one input class of states of alertness comprising a state of alertness derived while the subjects are sitting quietly with eyes open.
-
48. The method of claim 10, wherein said database of EEG signals of a plurality of subjects is based upon at least one input class of states of alertness comprising a state of alertness derived while the subjects are sitting quietly with eyes closed.
-
49. The method of claim 10, wherein said database of EEG signals of a plurality of subjects is based upon at least one input class of states of alertness comprising a state of alertness derived at the onset of sleep.
-
50. The method of claim 10, wherein the step of adjusting the person'"'"'s classification model based on at least one database of EEG classifications of alertness along said alertness-drowsiness continuum comprises adjusting the person'"'"'s classification model based upon a plurality of databases of classifications of alertness along said alertness-drowsiness continuum.
-
51. The method of claim 10, further comprising the steps of identifying eye blink artifacts in said EEG signals, and decontaminating epochs with eye blink artifacts.
-
52. The method of claim 51, wherein said step of decontaminating epochs with eye blink artifacts comprises adjusting data in epochs with eye blink artifacts.
-
53. The method of claim 10, further comprising the steps of acquiring EEG signals from the person during a state of alertness derived during a mental performance task, and using the EEG signals acquired during the state of alertness during the mental performance task in combination with the EEG classification model of the person along said alertness-drowsiness continuum to assess the effects of alertness on performance of the person.
-
89. The method of claim 10, further comprising the step of storing said EEG signal data for subsequent analysis.
-
54. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness;
providing a database of EEG signal data from said EEG signals from said plurality of subjects;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining classifications based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum. - View Dependent Claims (55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69)
said step of selecting a baseline subset set of EEG signal data comprises selecting a plurality of baseline subsets of EEG signal data from said EEG signal data representative of each of said plurality of input classes of states of alertness of said plurality of subjects; and
said step of determining at least one predictive variable comprises determining a plurality of predictive variables for determining each of said at least one output class of states of alertness based upon said plurality of baseline subsets of EEG signal data.
-
-
61. The method of claim 60, wherein said plurality of input classes comprise a highly alert vigilant condition, a low vigilance condition, an eyes closed condition, and a sleep onset condition.
-
62. The method of claim 54, said step of selecting at least one input class of states of alertness comprises selecting a first input level of classes, determining an initial output level of classifications, selecting a second input level of classes based upon said first input level of classes and said initial output level of classifications for each epoch of EEG signal data.
-
63. The method of claim 54, wherein said step of determining a classification of said EEG signals comprises using the High Vigilance, Low Vigilance and Eyes Closed conditions to adjust the classifications to correctly classify Sleepy.
-
64. The method of claim 54, further comprising computing a ratio of two or more output classes to further classify the person along the alertness-drowsiness continuum.
-
65. The method of claim 64, wherein said step of computing a ratio of two or more output classes comprises computing a ratio between the number of epochs classified as High Vigilance to the number of epochs classified as Eyes Closed or Sleepy to further classify the person along the alertness-drowsiness continuum.
-
66. The method of claim 54, further comprising computing z-score variables for the classifications based upon the power spectral density of one-hz bins for a given epoch compared to the mean and standard deviation of the same one-hz bin for a baseline condition.
-
67. The method of claim 54, wherein at least one input class of baseline states of alertness from the subject and the database of EEG signal data from said EEG signals from said plurality of subjects are used to predict a state of alertness which corresponds with the subject'"'"'s onset of sleep.
-
68. The method of claim 54, wherein said step of determining at least one predictive variable comprises determining at least one predictive variable to predict a state of alertness which corresponds with the person'"'"'s onset of sleep based upon at least one input class of states of alertness selected from states of alertness while sitting quietly with eyes open, while sitting quietly with eyes closed, and during a mental performance task.
-
69. The method of claim 54, wherein said step of determining at least one predictive variable comprises determining at least one predictive variable to predict a state of alertness which corresponds with the person'"'"'s onset of sleep based upon the input classes of states of alertness while sitting quietly with eyes open, and during a mental performance task.
-
70. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data;
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum; and
determining when a sequence of output classes representing at least a portion of the alertness-drowsiness continuum reaches a level of drowsiness such that the person'"'"'s performance could be substantially improved by taking a short nap. - View Dependent Claims (71)
-
-
72. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum, wherein said at least one input class of states of alertness comprises a state of alertness derived from subjects diagnosed with attention deficit disorder;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
-
-
73. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum, wherein said at least one input class of states of alertness comprises a state of alertness derived from subjects diagnosed with attention deficit/hyperactivity disorder;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
-
-
74. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum, wherein said plurality of input classes of states of alertness comprises a highly alert vigilant condition, a low vigilance condition, an eyes closed condition, and a condition associated with attention deficit disorder;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
-
-
75. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum, wherein said plurality of input classes of states of alertness comprises a highly alert vigilant condition, a low vigilance condition, an eyes closed condition, and a condition associated with attention deficit/hyperactivity disorder;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
-
-
76. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum;
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum;
identifying epochs contaminated with high levels of electromyography;
increasing an EMG counter by one when significant EMG artifact is identified in more than one overlapping epoch;
reading the EMG counter; and
providing an alarm to the subject when the EMG counter exceeds a predetermined threshold.
-
-
77. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum;
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum; and
identifying epochs contaminated with high levels of environmental electrical interference.
-
-
78. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum;
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum;
identifying epochs contaminated with high levels of environmental electrical interference;
increasing an environmental interference counter by one if a high level of environmental electrical interference is identified;
reading the environmental interference counter;
measuring the impedance of the electrode if the environmental electrical interference counter exceeds an environmental electrical interference artifact threshold for any electrode;
comparing the impedance of the electrode with an impedance threshold; and
providing an alarm if the impedance value exceeds the impedance threshold.
-
-
79. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum;
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum;
identifying an artifact of gross head, eye or body movement in said EEG signals;
increasing a movement counter by one if significant gross head, eye or body movement is identified in more than one overlapping epoch;
reading the movement counter; and
providing an alarm to the subject when the movement counter exceeds a predetermined threshold.
-
-
80. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data;
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum; and
comparing said at least one output class to a database of normal subjects to determine a person'"'"'s state of alertness.
-
-
81. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
assigning each of a plurality of said baseline subsets of EEG signal data to cluster groups based on cluster analysis of said plurality of baseline subsets of EEG signal data and using the cluster group assignments for the plurality of baseline subsets to define a plurality of databases of EEG signal data;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data for each database; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum for each database.
-
-
82. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data;
assigning a plurality of said baseline subsets of EEG signal data to a database based on cluster analysis of said baseline subsets of EEG signal data; and
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum.
-
-
83. A method for determining a system for classifying electroencephalogram (EEG) data acquired from a person along at least a portion of an alertness-drowsiness continuum for monitoring levels of alertness of the person, comprising the steps of:
-
selecting at least one input class of states of alertness representative of at least a portion of an alertness-drowsiness continuum;
acquiring EEG signals from a plurality of subjects during said at least one input class of states of alertness to provide a database of EEG signal data;
selecting a baseline subset of EEG signal data from said database of EEG signal data representative of said at least one input class of states of alertness of said plurality of subjects;
determining at least one predictive variable for determining said at least one input class of states of alertness based upon said baseline subset of EEG signal data;
determining a discriminant function based upon said at least one predictive variable for assigning the EEG data acquired from a person to at least one output class representing at least a portion of the alertness-drowsiness continuum; and
wherein said plurality of input classes of states of alertness comprises a highly alert vigilant condition, a low vigilance condition, and an eyes closed condition.
-
-
84. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum;
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum; and
identifying an artifact in said EEG signal data, and generating an audio alarm of spoken words with verbal instructions when said artifact is identified.
-
-
85. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
identifying an artifact of amplifier saturation in said EEG signal data;
rejecting epochs with amplifier saturation;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum; and
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum. - View Dependent Claims (86, 87)
-
-
88. A method for monitoring levels of alertness of a person from electroencephalogram (EEG) data from the person, comprising the steps of:
-
providing a database of EEG signals of a plurality of subjects from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
acquiring EEG signals of a person from a plurality of EEG electrode sites for a plurality of sequential epochs of a predetermined duration;
identifying an artifact of a spike in said EEG signals;
decontaminating the epoch containing a spike by determining an adjusting amplitude for each data point of the spike, and adjusting the amplitude of each data point of the spike by subtracting the adjusting amplitude from each said data point of the spike;
adjusting the person'"'"'s classification model based on at least one said database of EEG classifications of alertness along the alertness-drowsiness continuum to overcome individual differences in the EEG;
determining a classification of said EEG signals of the person along an alertness-drowsiness continuum; and
comparing said classification with a predetermined threshold classification of alertness along said alertness-drowsiness continuum.
-
Specification