Seminar: Graduate Seminar
EMD-Enhanced EEG for Prehospital Large Vessel Occlusion Stroke Detection
Date:
May,06,2026
Start Time:
13:30 - 14:30
Location:
1061, Meyer Building
Add to:
Lecturer:
Alexander Yorov
Research Areas:
| This seminar presents an EEG-based framework for prehospital detection of large vessel occlusion (LVO) stroke, aimed at supporting early routing decisions to thrombectomy-capable centers. The method uses a compact bilateral EEG montage and combines fixed-band power spectral density (PSD) features with empirical mode decomposition (EMD), which provides adaptive intrinsic mode functions that can capture pathology-related spectral shifts. On the primary task of distinguishing LVO from non-LVO ischemic stroke, the Hybrid EMD-PSD representation achieved 88.1% recording-level accuracy and AUROC 0.893 under recording-aware cross-validation, outperforming PSD-only and EMD-only alternatives. Feature analysis showed that spectral ratios, alpha-related PSD activity, IMF2 and posterior/posterior-temporal regions were especially informative, supporting compact EEG with interpretable adaptive spectral features as a promising tool for prehospital LVO triage. |
| Alexander Yorov is an M.Sc. student in the Andrew and Erna Viterbi Faculty of Electrical and Computer Engineering at the Technion. His research focuses on EEG-based signal processing and machine learning for prehospital stroke triage, with an emphasis on large vessel occlusion detection using compact bilateral EEG, spectral analysis and empirical mode decomposition.
M.Sc. student under the supervision of Dr. Danny Lange and Prof. Kfir Yehuda Levy.
|