Neural by Nature

Research

My work focuses on EEG-based machine learning for clinical applications.

Publications

  • Predicting ICU seizures using 1 hour of cEEG

    Phadikar S, Philibert-Rosas S, Haworth SE, Fatima S, Grobbelaar M, et al. Journal of Clinical Neurophysiology, 2026.

    Identifies and evaluates a comprehensive set of engineering and clinical biomarkers for machine learning based seizure prediction in hospitalized patients. Examines the interplay between EEG-derived signal features and clinical covariates to improve prediction performance in acute care settings, with emphasis on features that generalize across patient populations and monitoring configurations.

  • Duration of EEG monitoring needed to ensure a low risk of seizure recurrence in hospitalized patients

    Krishnamurthy PV, Philibert-Rosas S, Rivkin M, Brace CJ, Haworth SE, et al. Journal of Neurology, 2026;273(1):70.

    Presents a survival analysis of 117 hospitalized patients with electrographic or electroclinical seizures undergoing cEEG at the University of Wisconsin Hospital between 2018 and 2022, establishing data-driven thresholds for monitoring duration after seizure detection. Demonstrates that status epilepticus is the only clinical feature significantly associated with increased recurrence risk, and quantifies the EEG duration required to reduce that risk below 5%: 36.8 hours in patients with status epilepticus and 21.2 hours in those without.

  • Frequent cannabis use and the association with seizure burden in drug-resistant epilepsy

    Philibert-Rosas S, Guston KA, Kalkach-Aparicio M, Keller A, Semia S, Imhoff-Smith T, Adluru N, Brace CJ, Haworth SE, et al. Epilepsia, 2026.

Preprints

  • Neural signatures of conscious experience during sleep: a serial awakening study using high-density EEG

    Selte A, Haworth SE, Vanasse T, Alauddin T, Gjini K, Philibert-Rosas S, et al. bioRxiv, 2026. Co-first author, under review.

    Investigates the electrophysiologic and network-level features distinguishing conscious experience (CE) from no conscious experience (NCE) during NREM stage 2 and 3 sleep using HD-EEG combined with machine learning classification. In a serial-awakening paradigm across 140 participants and 699 awakenings, spectral power, functional connectivity via lagged coherence, and graph-theoretic metrics including modularity, global efficiency, and rich-club coefficient were extracted and screened using false discovery rate correction. A probabilistic ensemble classifier achieves ROC-AUC 0.80 and average precision 0.80 with stable 5-fold participant-level cross-validation, with the largest discriminating effects observed in posterior alpha, delta, and gamma-to-alpha power ratios.

  • Thermodynamic rigidity of harmonic brain states relates to general mental ability in juvenile myoclonic epilepsy

    De Paiva FB, Zhao M, Zhao M, Philibert-Rosas S, Brace CJ, Moe E, Haworth SE, et al. bioRxiv, 2026.

In progress

  • Argus: A CNN-LSTM Architecture for Real-Time Seizure Detection in Long-Term EEG Monitoring

    Haworth SE, Philibert-Rosas S, Phadikar S, Struck AF, Jing J, Westover MB, Zafar S. Manuscript in preparation.

    Presents Argus, a multi-scale convolutional and bidirectional LSTM network with an attention mechanism for automated seizure detection. Trained and evaluated on the Temple University Seizure Corpus (TUSZ v1.5.2) using a 22-channel TCP bipolar montage at 200 Hz. The architecture is designed with clinical deployment in mind, prioritizing low false positive rates in continuous long-term monitoring settings.

Posters and conference abstracts Google Scholar