Related Experiment Video
Updated: May 16, 2026
09:57
Electromagnetic Source Imaging in Presurgical Evaluation of Children with Drug-Resistant Epilepsy
Published on: September 20, 2024
A statistical method for predicting seizure onset zones from human single-neuron recordings
André B Valdez1, Erin N Hickman, David M Treiman
1Department of Neurology, Barrow Neurological Institute, Phoenix, AZ, USA.
Journal of Neural Engineering
|December 11, 2012
Summary
Logistic regression models accurately predict seizure onset zones (SOZs) using single-neuron recordings from depth electrodes. This method enhances epilepsy diagnosis and surgical planning for patients.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Medical Technology
Background:
- Depth-electrode recordings are crucial for localizing epileptogenic foci in epilepsy patients.
- Improving the diagnostic accuracy of these recordings can significantly aid clinical decision-making.
Purpose of the Study:
- To apply logistic regression models to single-neuron recordings from depth electrodes.
- To predict seizure onset zones (SOZs) and enhance the diagnostic value of electrophysiological data.
Main Methods:
- Data collected from 17 epilepsy patients.
- Developed logistic regression models using burst interspike interval (ISI) statistics.
- Calculated the odds of SOZs in the hippocampus, amygdala, and ventromedial prefrontal cortex.
Main Results:
- A single-unit increase in burst ISI ratio increased the likelihood of SOZs in the left hippocampus (12x) and right amygdala (14.5x).
- Models achieved 85% average sensitivity for bilateral hippocampus SOZs.
- Performance was comparable to electroencephalography (EEG).
Conclusions:
- Logistic regression models integrated with single-neuron recordings can predict likely SOZs.
- This approach offers an automated, clinically valuable tool for epilepsy surgery evaluation.
