Localization of deep brain activity with scalp and subdural EEG

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Deep and subcortical brain regions are crucial for cognitive functions.
• Comparing the capabilities of electroencephalography (EEG) and electrocorticography (ECoG) for deep brain source localization remains an open question.
• Existing research highlights EEG's potential for deep source detection, while ECoG offers superior spatial resolution but limited coverage.

Purpose of the Study:

• To investigate and compare the accuracy of electroencephalography (EEG) and electrocorticography (ECoG) in localizing deep brain activity sources.
• To evaluate the effectiveness of independent component analysis (ICA) for source localization using multimodal EEG and ECoG data.
• To assess the clinical significance of differences in deep source localization accuracy between EEG and ECoG.

Main Methods:

• Recorded simultaneous scalp EEG, subdural ECoG, and depth EEG (hippocampus/insula) from 4 epilepsy patients during quiet wakefulness.
• Applied independent component analysis (ICA) to isolate neural sources in theta, alpha, and beta frequency bands.
• Utilized dipole modeling to determine source locations of correlating components and compared their proximity to depth electrodes.

Main Results:

• Components from both subdural ECoG and scalp EEG showed significant correlations with depth electrode contacts.
• Dipole modeling revealed that sources localized using subdural electrodes were approximately 70% closer to depth electrodes (~2cm improvement) compared to scalp EEG.
• Despite the proximity advantage of ECoG, scalp EEG demonstrated comparable deep source localization accuracy to ECoG, challenging the notion of a considerable practical difference.

Conclusions:

• Scalp EEG offers a deep source localization accuracy comparable to ECoG, particularly when considering the practical limitations and proximity of implanted ECoG electrodes.
• The study provides novel multimodal data and experimental evidence comparing invasive and non-invasive brain activity measurements for deep cortical structure localization.
• The findings suggest that scalp EEG, despite its lower spatial resolution, can be a viable tool for investigating deep brain activity, potentially reducing the need for invasive procedures in some clinical contexts.