Neurotech · biomedical signals
SeizureX
Applied research to assist epileptic-seizure detection from EEG signals.
Problem
Reviewing EEG for seizures is slow and specialist-heavy. Computational support should respect clinical judgment, not replace it blindly.
Approach
Deep-learning models on biomedical signals, with emphasis on preprocessing, the time window, and an interpretable output for whoever reviews the study.
Outcome
A research prototype for assisted detection. We do not publish sensitivity or specificity figures: the relevant outcome is a reproducible pipeline, ready to evaluate on a clinical partner's real data.
Stack
Python · PyTorch · EEG · time series