February Fourier Talks 2014

The surgical management of drug-resistant epilepsy offers a unique opportunity to listen in on the human brain at work. When surgical planning requires long-term monitoring with implanted electrodes, clinicians and neuroscientists can study the dynamics of neural networks with unprecedented spatial and temporal resolution as these networks participate in both normal and pathological brain states. However, this requires the interpretation of complex EEG signals that reflect the organized activity of large populations of neurons. One obvious and important way in which population activity is organized is in oscillations, which can occur at different frequencies and can be augmented or suppressed under different physiological conditions. These changes in oscillatory activity can be used to observe the human brain at work as its different functional-anatomic components are recruited in real time to perform everyday tasks. In special circumstances, it is even possible to use these EEG signal changes to decode and reconstruct perceptions and movements that a patient is experiencing, raising the possibility of brain-machine interfaces for individuals with disabilities. Moreover, functional interactions between the different components of large-scale brain networks can also be investigated by measuring correlations and causal relationships between oscillations recorded at different brain sites. The dynamics of these interactions may offer greater insights into how individual brain regions contribute to network activity and thus to overall brain function. This information is not just interesting from a scientific perspective, but also has the potential to provide much needed guidance to clinicians seeking to maximize the resection of tissue responsible for seizures while minimizing the resection of tissue responsible for normal function.