Related Experiment Videos
Time-frequency analysis using the matching pursuit algorithm applied to seizures originating from the mesial temporal
P J Franaszczuk1, G K Bergey, P J Durka
1Maryland Epilepsy Center, Department of Neurology, University of Maryland School of Medicine and Medical Center, Baltimore 21201, USA. pfranasz@umaryland.edu
Electroencephalography and Clinical Neurophysiology
|September 19, 1998
Summary
The matching pursuit algorithm effectively analyzes dynamic seizure activity using time-frequency decomposition. This method aids in understanding seizure evolution and comparing patterns across different brain regions.
Area of Science:
- Neuroscience
- Signal Processing
- Medical Technology
Background:
- Analyzing seizure activity is crucial for localization and classification.
- Seizure signals are dynamic and multi-frequency, complicating traditional analysis.
- The matching pursuit algorithm offers continuous time-frequency analysis capabilities.
Purpose of the Study:
- To apply the matching pursuit algorithm to intracranial ictal recordings.
- To evaluate its utility in analyzing dynamic seizure activity.
- To explore its potential for seizure localization and classification.
Main Methods:
- Analysis of mesial temporal onset partial seizures from 9 patients.
- Utilized the matching pursuit algorithm on single-channel intracranial EEG recordings.
- Plotted time-frequency energy distributions and correlated them with EEG data.
Main Results:
- Identified distinct seizure evolution phases: initiation, transitional, organized, and intermittent bursting.
- Organized rhythmic bursting activity showed a predominant frequency of 5.3-8.4 Hz with a <60s decline.
- The matching pursuit method enabled time-frequency decomposition of complete seizures.
Conclusions:
- The matching pursuit method is valuable for time-frequency analysis of dynamic seizure activity.
- It is suitable for non-stationary seizure evolution patterns.
- Facilitates comparison of time-frequency seizure patterns from different brain regions.