EEG recording and analysis
One embodiment provides a method, including: obtaining EEG data from one or more single channel EEG sensor worn by a user; classifying, using a processor, the EEG data as one of nominal and abnormal; and providing an indication associated with a classification of the EEG data. Other embodiments are described and claimed.
1 . A method for detecting seizures using electroencephalogram (EEG) data, the method comprising:
obtaining EEG data collected by a plurality of discrete wireless EEG sensors positioned on a scalp of a patient being evaluated;
segmenting the EEG data into a plurality of EEG data segments;
identifying probabilities of occurrence of a seizure in the plurality of EEG data segments;
automatically identifying a discrete seizure event by evaluating the identified probabilities of a temporally sequential subset of the plurality of EEG data segments against one or more thresholds, wherein the subset includes multiple temporally sequential EEG data segments, and wherein a start and stop time of the discrete seizure event is determined;
automatically creating a plurality of EEG channels from the EEG data, the plurality of EEG channels including a channel from each discrete wireless EEG sensor of the plurality of discrete wireless EEG sensors and additional channels determined by, for each discrete wireless EEG sensor, subtracting its corresponding EEG data from a corresponding EEG data of the remaining discrete wireless EEG sensors; and
automatically annotating the EEG data with a graphical indicator that denotes the discrete seizure event,
wherein the method is performed under control of at least one processor.
2 . The method of claim 1 , wherein identifying probabilities of occurrence of the seizure is performed using a machine learning model trained with another plurality of EEG data segments obtained from EEG data collected by another plurality of discrete wireless EEG sensors positioned on a plurality of scalps of a plurality of patients different from the patient being evaluated.
3 . The method of claim 1 , further comprising generating a normalized EEG data to account for inter-patient and inter-sensor differences, wherein the segmenting comprises segmenting the normalized EEG data.
4 . The method of claim 1 , wherein the plurality of discrete wireless EEG sensors comprises four wireless EEG sensors positioned on the scalp of the patient at a left forehead, at a right forehead, behind a left ear, and behind a right ear.
5 . The method of claim 1 , wherein each discrete wireless EEG sensor of the plurality of discrete wireless EEG sensors comprises two electrodes forming a bipolar channel.
6 . The method of claim 1 , wherein the plurality of EEG channels comprises ten channels.
7 . The method of claim 1 , further comprising generating an alert indicating detection of the discrete seizure event.
8 . The method of claim 1 , wherein automatically annotating the EEG data with the graphical indicator comprises highlighting a region of the EEG data associated with the discrete seizure event.
9 . The method of claim 1 , wherein the graphical indicator denotes the start and stop time of the discrete seizure event.
10 . At least one non-transitory computer readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to:
obtain electroencephalogram (EEG) data collected by a plurality of discrete wireless EEG sensors positioned on a scalp of a patient being evaluated;
segment the EEG data into a plurality of EEG data segments;
identify probabilities of occurrence of a seizure in the plurality of EEG data segments;
automatically identify a discrete seizure event by evaluating the identified probabilities of a temporally sequential subset of the plurality of EEG data segments against one or more thresholds, wherein the subset includes multiple temporally sequential EEG data segments, and wherein a start and stop time of the discrete seizure event is determined;
automatically create a plurality of EEG channels from the EEG data, the plurality of EEG channels including a channel from each discrete wireless EEG sensor of the plurality of discrete wireless EEG sensors and additional channels determined by, for each discrete wireless EEG sensor, subtracting its corresponding EEG data from a corresponding EEG data of the remaining discrete wireless EEG sensors; and
automatically annotate the EEG data with a graphical indicator that denotes the discrete seizure event.
11 . The computer readable medium of claim 10 , wherein the instructions cause the at least one processor to:
identify probabilities of occurrence of the seizure using a machine learning model trained with another plurality of EEG data segments obtained from EEG data collected by another plurality of discrete wireless EEG sensors positioned on a plurality of scalps of a plurality of patients different from the patient being evaluated.
12 . The computer readable medium of claim 10 , wherein the instructions cause the at least one processor to:
generate a normalized EEG data to account for inter-patient and inter-sensor differences; and
segment the normalized EEG data into the plurality of EEG data segments.
13 . The at least one computer readable medium of claim 10 , wherein the plurality of EEG channels comprises ten channels.
14 . The computer readable medium of claim 10 , wherein the instructions cause the at least one processor to generate an alert indicating detection of the discrete seizure event.
15 . The computer readable medium of claim 10 , wherein the instructions cause the at least one processor to automatically annotate the EEG data with the graphical indicator by highlighting a region of the EEG data associated with the discrete seizure event.
16 . The non-transitory computer readable medium of claim 10 , wherein the graphical indicator denotes the start and stop time of the discrete seizure event.
17 . A system for detecting seizures using electroencephalogram (EEG) data, the system comprising:
a plurality of discrete wireless EEG sensors configured to be positioned on a scalp of a patient being evaluated; and
at least one non-transitory computer readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to:
obtain EEG data collected by the plurality of discrete wireless EEG sensors;
segment the EEG data into a plurality of EEG data segments;
identify probabilities of occurrence of a seizure in the plurality of EEG data segments;
automatically identify a discrete seizure event by evaluating the identified probabilities of a temporally sequential subset of the plurality of EEG data segments against one or more thresholds, wherein the subset includes multiple temporally sequential EEG data segments, and wherein a start and stop time of the discrete seizure event is determined;
automatically create a plurality of EEG channels from the EEG data, the plurality of EEG channels including a channel from each discrete wireless EEG sensor of the plurality of discrete wireless EEG sensors and additional channels determined by, for each discrete wireless EEG sensor, subtracting its corresponding EEG data from a corresponding EEG data of the remaining discrete wireless EEG sensors; and
automatically annotate the EEG data with a graphical indicator that denotes the discrete seizure event.
18 . The system of claim 17 , wherein the plurality of discrete wireless EEG sensors comprises four wireless EEG sensors configured to be positioned on the scalp of the patient at a left forehead, at a right forehead, behind a left ear, and behind a right ear.
19 . The system of claim 17 , wherein each discrete wireless EEG sensor of the plurality of discrete wireless EEG sensors comprises two electrodes forming a bipolar channel.
20 . The system of claim 17 , wherein the plurality of EEG channels comprises ten channels.
21 . The system of claim 17 , wherein the instructions cause the at least one processor to generate an alert indicating detection of the discrete seizure event.
22 . The system of claim 17 , wherein the instructions cause the at least one processor to automatically annotate the EEG data with the graphical indicator by highlighting a region of the EEG data associated with the discrete seizure event.
23 . The system of claim 17 , wherein the graphical indicator denotes the start and stop time of the discrete seizure event.