Abstract
The time-frequency (TF) version of Renyi entropy, which measures the information content and complexity of a signal, is used here as a feature in the classification of the newborn heart rate variability (HRV) as either corresponding to seizure or non-seizure. The newborn HRV is initially mapped to the TF domain using the modified B distribution (MBD). The time-frequency distribution (TFD) of HRV is post-processed before the Renyi entropy is computed. This post-processing method uses an image processing technique called component linking to identify the true HRV components and localize them in the TF plane. The results obtained so far show that the HRV corresponding to non-seizure can be discriminated from those corresponding to seizure using TF-based Renyi entropy with 78.57% sensitivity and 83.33 % specificity.
Original language | English |
---|---|
Title of host publication | 2007 9th International Symposium on Signal Processing and its Applications, ISSPA 2007, Proceedings |
DOIs | |
Publication status | Published - 2007 |
Event | 2007 9th International Symposium on Signal Processing and its Applications, ISSPA 2007 - Sharjah, United Arab Emirates Duration: Feb 12 2007 → Feb 15 2007 |
Other
Other | 2007 9th International Symposium on Signal Processing and its Applications, ISSPA 2007 |
---|---|
Country | United Arab Emirates |
City | Sharjah |
Period | 2/12/07 → 2/15/07 |
ASJC Scopus subject areas
- Signal Processing