Brain state forecasting for precise brain stimulation: Current approaches and future perspectives.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Biomedical Engineering

Background:

• Transcranial magnetic stimulation (TMS) offers potential for understanding brain function and treating neurological/psychiatric disorders.
• Low reproducibility in TMS results is a significant challenge, often attributed to individual brain variability.
• Personalized stimulation protocols using neuroimaging data can improve targeting but face theoretical and technical hurdles.

Purpose of the Study:

• To address the challenge of online functional targeting in real-time brain state-dependent TMS.
• To investigate methods for compensating hardware/software delays in electroencephalography (EEG)-triggered TMS.
• To explore the potential of forecasting brain states for accurate, real-time TMS delivery.

Main Methods:

• Reviewing the state-of-the-art in brain state-dependent stimulation.
• Discussing two classes of forecasting methods suitable for EEG time series analysis.
• Examining the evidence for data-driven forecasting in the context of TMS.

Main Results:

• Real-time EEG signal processing and forecasting are necessary to compensate for system delays in online TMS.
• Forecasting brain states allows TMS devices to target predicted, rather than measured, brain activity.
• This approach has shown success in preliminary studies, paving the way for personalized brain stimulation.

Conclusions:

• Forecasting methods are crucial for overcoming limitations in current real-time, brain state-dependent TMS.
• Data-driven forecasting holds significant potential for advancing TMS methodology and understanding brain dynamics.
• The use of big open datasets could further transform personalized brain stimulation treatments.