Seizure Detection from ECG Data
This visualization shows heart rate variability (RR intervals) extracted from ECG data, with seizure events marked in red.
Study Background
This dataset is from a study examining epileptic seizure detection using heart rate variability from ambulatory ECG recordings. The research demonstrates that autonomic changes in heart rate patterns can be used to detect seizures with reasonable accuracy.
Reference: Li J, Nurse ES, Grayden DB, Cook MJ, Karoly PJ. Epileptic seizure detection using heart rate variability from ambulatory ECG: A pseudoprospective study. J Neural Eng. 2025.
ECG Data Visualization
Select a patient to view their RR interval data with seizure markers. The red highlighted regions indicate seizure events.
About the Data
The dataset includes ECG recordings from 65 patients (47 in training set, 18 in test set). For each patient, we have:
- RR intervals: The time between consecutive heartbeats
- Seizure times: Start and end timestamps of seizure events
- Study times: Start and end timestamps of the recording period
Key Findings
Seizure Detection
The machine learning model achieved 72% sensitivity and 68% specificity on unseen test data.
Patient Factors
Detection performed better for patients with focal epilepsy and those who experienced significant heart rate changes during seizures.
Clinical Relevance
ECG-based seizure detection offers a less obtrusive alternative to EEG monitoring for long-term seizure tracking.