Patient-specific seizure prediction based on heart rate variability and recurrence quantification analysis
Fig 3
Dynamics of the most discriminant features.
The image represents the dynamic of the features with the most discriminant power for preictal and interictal segments according to the stepwise regression analysis. Each panel presents dynamic changes in one particular HRV parameter from 25 minutes before seizure onset to 5 minutes after the end of the seizure. Seizure onset and seizure end are represented by the vertical red and green lines respectively. An exemplificative seizure from one patient is reported (Patient 15, Seizure 2). a) meanNN, b) pNN50, c) CosEn: coefficient of sample entropy, d) LAM: laminarity, e) HF: high frequency, f) LF/HF: ratio between low and high frequency.