A physiologically motivated ECoG segmentation method for epileptic seizure onset zone detection

Area of Science:

• Neuroscience
• Medical Engineering

Background:

• Epilepsy diagnosis relies on identifying seizure onset zones.
• Electrocorticography (ECoG) provides high-resolution brain activity data.
• Accurate segmentation of ictal ECoG is crucial for precise localization.

Purpose of the Study:

• To develop a novel segmentation method for multichannel ictal ECoG data.
• To classify segments based on predominant epileptic activity.
• To determine the seizure onset zone using temporal delays in specific frequency bands.

Main Methods:

• Proposed a segmentation method analyzing relative frequency contributions in ECoG data.
• Classified segments exhibiting predominant epileptic activity.
• Determined seizure onset zone by analyzing temporal delays of 4-9Hz epileptic activity across channels.

Main Results:

• The method successfully segmented three seizures from a single focal epilepsy patient.
• Generated segments visually corresponded well to ictal ECoG characteristics.
• The identified seizure onset zone showed high concordance with clinical findings.

Conclusions:

• The novel segmentation method effectively identifies epileptic activity and seizure onset zones.
• This approach offers a promising tool for analyzing ECoG data in focal epilepsy.
• Results support the method's clinical relevance and accuracy.