Log In

Forgot Password?

OR

Not a member? Continue as a nonmember.

Become a Member

By becoming a member of the AAN, you can receive exclusive information to help you at every stage of your career. Benefits include:

Join Now See All Benefits

Loading... please wait

Abstract Details

Predicting Seizure Freedom through Computational Processing of Routine EEG
Epilepsy/Clinical Neurophysiology (EEG)
S30 - Epilepsy/Clinical Neurophysiology (EEG): Technology (3:42 PM-3:54 PM)
002

The presence of interictal epileptiform discharges (IEDs) on routine EEG predicts seizure recurrence. However, 29–55% of routine EEGs do not show IEDs in patients with epilepsy (PWE), limiting its prognostic yield.

We aim to evaluate the performance of computational markers to predict seizure freedom after routine EEG.

We selected all patients undergoing 30-minute EEG at the CHUM in 2018. Exclusion criteria were absence of follow-up after EEG, unclear epilepsy diagnosis at last follow-up, and electrical seizure(s) on EEG. Medical charts were reviewed for seizure freedom after EEG. EEGs were segmented into non-overlapping 10s windows. Several features were extracted for each electrode and each time window, each feature capturing a distinct linear or non-linear property of the EEG signal. The extracted features were: band power, peak alpha frequency, Hurst exponent, entropy (ApEn, SampEn, MsEn, SpecEn, FuzzEn, and PermEn), line length, and correlation dimension. We trained a regularized boosted trees machine learning model on multichannel EEG segments to predict seizure freedom during follow-up. We averaged predictions over windows to obtain one prediction per EEG. We tuned hyperparameters with cross-validation on a validation ensemble (80% of data) and evaluated the model on a testing set (20% of data). 

656 EEGs from 534 subjects were included; 345 (65%) were from PWE. 527 EEGs (80%) did not capture IEDs. Median follow-up was 92 weeks (IQR 40–126). The classifier, using all features as input, could predict seizure freedom with above chance performance (area under the receiver-operating characteristic curve [AUC ROC] 0.68, 95%CI 0.65–0.72), even in the subset of patients with no IEDs on EEG (0.66, 0.61–0.70). Among the features, band power showed the best performance, followed by FuzzEn and line length.

A computational biomarker could potentially increase the prognostic yield of routine EEG in the clinical setting.

Authors/Disclosures
Emile Lemoine, MD
PRESENTER
Dr. Lemoine has received research support from Canadian Institute of Health Research.
Denahin Hinnoutondji Toffa, MD, PhD, CSCN (CHUM Research Center) Denahin Hinnoutondji Toffa has nothing to disclose.
Frédéric Lesage, PhD (École Polytechnique Montreal) Prof. Lesage has stock in Labeo Technologies . Prof. Lesage has received intellectual property interests from a discovery or technology relating to health care.
Dang Khoa Nguyen, MD,PhD (Centre Hospitalier de l'Université de Montréal) The institution of Dr. Nguyen has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Paladin Labs. The institution of Dr. Nguyen has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for UCB. The institution of Dr. Nguyen has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for Eisai.
Elie Bou Assi, PhD (University of Montreal) Prof. Bou Assi has received personal compensation in the range of $0-$499 for serving as a Consultant for Akcea Therapeutics, Canada Inc.