Quantitative approaches to guide epilepsy surgery from intracranial EEG

Area of Science:

• Computational neuroscience
• Epileptology
• Medical device development

Background:

• Epilepsy surgery outcomes are limited by subjective interpretation of intracranial EEG (iEEG).
• Quantitative iEEG analytics are being developed to objectively guide surgical decisions.
• Existing computational methods often lack generalizability and are sensitive to data variations.

Purpose of the Study:

• To critically review computational methods for identifying the epileptogenic zone from iEEG data.
• To highlight common methodological challenges in quantitative iEEG analysis.
• To propose solutions and share a centralized, high-quality, multi-center iEEG dataset.

Main Methods:

• Literature review of computational methods for epileptogenic zone identification using iEEG.
• Analysis of common pitfalls including data limitations and method generalizability.
• Development and sharing of a centralized, well-annotated, multi-center iEEG dataset (>100 patients).

Main Results:

• Identified shared methodological challenges across various quantitative iEEG analysis techniques.
• Highlighted the critical need for open-source, high-quality data to improve method rigor.
• Provided a centralized dataset to facilitate larger, more robust studies.

Conclusions:

• Addressing methodological challenges and data limitations is crucial for advancing quantitative iEEG analysis.
• The shared dataset aims to support the development and validation of more generalizable tools.
• A roadmap is proposed to guide computational epilepsy research towards clinical trials and improved patient outcomes.