Inferring seizure frequency from brief EEG recordings
M Brandon Westover1, Matt T Bianchi, Mouhsin Shafi
1MGH Epilepsy Service, Massachusetts General Hospital, and Harvard Medical School, Boston, MA, USA. mwestover@partners.org
Summary
Observing a single seizure during an electroencephalogram (EEG) offers limited insight into epilepsy seizure frequency. This study provides a probabilistic framework to help clinicians interpret this finding and guide treatment decisions.
Area of Science:
- Neurology
- Clinical Neurophysiology
Background:
- Routine electroencephalograms (EEGs) are crucial for epilepsy management.
- A single observed seizure during a brief EEG has unclear implications for underlying seizure frequency.
- Patient-reported seizure frequency is often inaccurate, complicating clinical interpretation.
Purpose of the Study:
- To apply probabilistic inference to model seizure incidence.
- To provide clinicians with guidance for interpreting a single seizure observed during routine EEG.
- To establish bounds on seizure rates based on limited EEG data.
Main Methods:
- Utilized standard concepts of probabilistic inference.
- Developed a simple model of seizure incidence.
- Analyzed the implications of observing a single seizure event.
Main Results:
- Established upper and lower bounds for seizure rates implied by a single EEG seizure.
- Demonstrated that incorporating prior information on expected seizure rates can refine these bounds.
- Provided a framework for more principled decision-making.
Conclusions:
- A single seizure on routine EEG provides a basis for probabilistic estimation of seizure frequency.
- The developed framework aids clinicians in interpreting this finding and making informed treatment decisions.
- This approach offers a more principled method for managing epilepsy based on limited ictal events.
