11,522 Publications · 65,312 Citations

We assessed heart rate (HR) around the clinician-marked onset of the first ictal run in 21 patients with stereo-electroencephalography (SEEG) in the HUP dataset (OpenNeuro ds004100). For each seizure, we compared median HR in windows before and after onset with a baseline window 120 to 60 seconds before onset. Mean HR was 71.8 bpm at baseline. HR increased by 18.3 bpm in the first 30 seconds after onset (95% bootstrap CI 10.7 to 27.2; 18/21 patients positive; Holm-adjusted p < 0.001) and by 32.9 bpm in seconds 30–60 (CI 20.1 to 45.7; 18/21; Holm-adjusted p = 0.0002). In the 30 seconds before onset, the change was +1.8 bpm (CI −1.4 to +5.9; 9/21; Holm-adjusted p = 0.41). The control window 60 to 30 seconds before onset changed by +1.2 bpm (Holm-adjusted p = 0.15). HR first exceeded baseline +10 bpm in 20/21 seizures, at a median of 3.75 seconds after onset; 8 crossings occurred before onset. Thus, in this sample, a group-level HR rise was detected after the marked intracranial onset, while no group-level rise was detected in the preceding 30 seconds. Early individual crossings could reflect early autonomic involvement, late onset marking, or noise. An ECG detector would mostly flag seizures after intracranial onset in these data. The analysis covered 21 of about 36 SEEG patients, used one seizure per patient, and has additional limitations described in the full text. The source files were fetched from an unversioned OpenNeuro S3 mirror URL, so their exact dataset release is uncertain.