Abstract
Objective: To investigate the feasibility of improving the performance of an EEG-based multistate classifier (MSC) previously proposed by our group. Results: Using the random forest (RF) classifiers on the previously reported dataset of patients, but with three improvements to classification logic, the specificity of our alarm algorithm improves from 82.4% to 92.0%, and sensitivity from 87.9% to 95.2%. Discussion: The MSC could be a useful approach for seizure-monitoring both in the clinic and at home. Methods: Three improvements to the MSC are described. Firstly, an additional check using RF outputs is made prior to alarm to confirm increasing probability of a seizure onset state. Secondly, a post-alarm detection horizon that accounts for the seizure state duration is implemented. Thirdly, the alarm decision window is kept constant.
Highlights
NEED Features that could reliably identify a state prior to clinical seizure from scalp EEG are desirable for home-monitoring applications, as the engulfing fear of unpredictability of seizures has a major impact on quality of life for subjects living with epilepsy [1]
Table 2 summarizes the impact of each change on the multistage state classifier (MSC) performance metrics
The additional logical state checks and the detection delay lead to improved sensitivity and specificity of the MSC
Summary
Features that could reliably identify a state prior to clinical seizure from scalp EEG are desirable for home-monitoring applications, as the engulfing fear of unpredictability of seizures has a major impact on quality of life for subjects living with epilepsy [1]. A recent paper from our group [2] described a multistage state classifier (MSC) using cross frequency coupling (CFC) features in human scalp EEG. The MSC is based on three RF classifiers: IIS1 (outputs probability of S1 over II), IIS2, and S1S2. These classifiers detected CFC changes in advance of a clinical seizure onset, not necessarily at the EG onset. This suggested that state transitions similar to those occurring at EG could occur before EG onset. Using the same set of patient recordings as previously published, we show that these modifications significantly improve sensitivity and specificity of pre-clinical seizure state classification
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal of Translational Engineering in Health and Medicine
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
