Decoding human spontaneous spiking activity in medial temporal lobe from scalp EEG

Area of Science:

• Neuroscience
• Biomedical Engineering

Background:

• Directly linking scalp electroencephalography (EEG) signals to deep brain nuclei activity in humans is challenging.
• Simultaneous recordings of scalp EEG and deep brain unit activity are crucial for understanding brain function.

Purpose of the Study:

• To model the relationship between scalp EEG spectral features and neuronal firing activity in the human amygdala and hippocampus.
• To determine if EEG can predict spontaneous neuronal firing rates in deep brain structures.

Main Methods:

• Analysis of simultaneous overnight scalp EEG and deep brain unit recordings from patients undergoing pre-surgical monitoring.
• Linear regression modeling to correlate EEG spectral features with average unit firing activity during wakefulness and non-REM sleep.
• Investigated region specificity and contribution of short- and long-term firing rate fluctuations.

Main Results:

• EEG spectral features significantly predicted average unit firing activity in both the amygdala and hippocampus (Pearson r > 0.2, p << 0.001).
• Model accuracy was influenced by both short- and long-term fluctuations in neuronal firing rates.
• Demonstrated region-specific relationships between EEG and deep brain activity.

Conclusions:

• Scalp EEG frequency modulations can reliably estimate changes in neuronal firing rates in deep brain structures like the amygdala and hippocampus.
• This research opens new avenues for non-invasive monitoring and therapeutic interventions in neurological and psychiatric disorders.
• Establishes a predictive link between non-invasive EEG and invasive neuronal activity, enhancing brain monitoring capabilities.