Decoding Intracranial EEG With Machine Learning: A Systematic Review.

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Technology

Background:

• Intracranial electroencephalography (iEEG) offers high-resolution brain activity data.
• Understanding neural codes is crucial for diagnosing and treating neurological disorders.
• Limited clinician awareness exists regarding machine learning (ML) applications in iEEG for neurosurgery.

Approach:

• A systematic literature review identified 107 articles on artificial intelligence applications in iEEG data.
• Machine learning techniques were analyzed for their merits, limitations, and clinical relevance.
• Applications were categorized into seizure analysis, motor tasks, cognitive assessment, and sleep staging.

Key Points:

• Supervised learning algorithms are the most frequently employed ML techniques in iEEG studies.
• Publicly available time-series datasets are commonly utilized in ML-based iEEG research.
• ML shows promise in decoding complex neural signals for improved diagnostic and therapeutic insights.

Conclusions:

• Machine learning offers significant potential for advancing neurosurgical applications by enhancing iEEG data interpretation.
• Further research and development are needed to bridge the gap between ML techniques and clinical practice in neurosurgery.
• Clinical translation of ML in neurosurgery can improve patient outcomes through better seizure analysis, motor task assessment, cognitive evaluation, and sleep staging.