What do intracerebral electrodes measure?
Introduction
Intracerebral EEG (iEEG) is routinely used for presurgical evaluation in pharmaco-refractory focal epilepsy patients (Kovac et al., 2017, Nagahama et al., 2018, Wasade and Spanaki, 2019). Typically, not all channels in a patient record epileptic activity. Thus, iEEG also provides an invaluable tool for research as the only source of human intracerebral physiological recordings (Engel et al., 2005, Mukamel and Fried, 2012, Frauscher et al., 2018a, von Ellenrieder et al., 2020). Two types of electrodes exist for recording iEEG, cortical grids placed on the surface of the brain, providing a dense covering of the top of the gyri, and intracerebral electrodes that are inserted in the brain resulting in a sparser covering of the gyri but allowing for the sampling of sulci and deeper structures.
In a previous publication we studied the physiologic high frequency activity above 80 Hz, recorded with intracerebral electrodes in a large number of channels selected from different focal epilepsy patients (Frauscher et al., 2018b). Two different intracerebral electrodes had been used to record these signals, commercial Dixi electrodes (Dixi Medical, France) and Montreal Neurological Institute (MNI) home-made (HM) electrodes. We were surprised to find differences in some characteristics of the signal depending on the electrode type, since this had not been observed at frequencies of up to 20 Hz (Frauscher et al., 2018a). This frequency dependent difference between electrodes is unexpected, and prompted us to investigate the issue in more depth.
In this study we aim to gain insight and improve our interpretation of measurements from intracerebral electrodes by using a combination of measured data and numerical simulations. The measured data is available for download through the LORIS platform (Das et al., 2016) as part of the MNI Open iEEG Atlas (https://mni-open-ieegatlas.research.mcgill.ca/). The numerical simulations carried out with the Boundary Elements Method (Brebbia and Dominguez, 1992) are an improvement over an existing model (von Ellenrieder et al., 2012). The Matlab (Mathworks, USA) code is available at Zenodo (https://doi.org//10.5281/zenodo.4044274). We also use these tools to explain the difference in the signals that are recorded with different types of electrodes, and show that differences between typical commercially available electrodes do not affect the sensitivity of iEEG measurements in a significant way.
Section snippets
Characterization of measured signals
We used the high sampling-frequency dataset available for download from the MNI Open iEEG Atlas. This dataset corresponds to recordings from channels with physiologic activity from many different focal epilepsy patients with a duration of 20 min, acquired during non rapid-eye-movement sleep (sleep stages N2 and N3, analyzed often for epileptic interictal activity). We limited our analysis to the recordings from the Montreal Neurological Institute, in order to compare iEEG signals acquired with
Characterization of the signals
The high sampling frequency dataset of the MNI Open iEEG Atlas contains 295 bipolar channels recorded with Dixi electrodes and 247 with HM electrodes, both recorded at the MNI with the same equipment and a sampling rate of 2 kHz. Thirty-five channels from two patients recorded with Dixi electrodes had a different antialiasing filter setting (300 Hz instead of 600 Hz cut-off frequency) and were excluded from the general PDS analysis but kept in the rest of the analyses.
We computed the median PSD
Discussion
Through the use of measurement and simulations we characterized the effect of different parameters in the sensitivity profile and performance of iEEG electrodes, and explained the difference between two intracerebral electrode types.
Declaration of Competing Interest
None of the authors have potential conflicts of interest to be disclosed.
Acknowledgements
This work was supported by the Canadian Institute of Health Research (grant FDN-143208).
References (28)
- et al.
Contact size does not affect high frequency oscillation detection in intracerebral EEG recordings in a rat epilepsy model
Clin Neurophysiol
(2011) - et al.
Influence of contact size on the detection of HFOs in human intracerebral EEG recordings
Clin Neurophysiol
(2013) - et al.
The MNI data-sharing and processing ecosystem
Neuroimage
(2016) - et al.
Invasive epilepsy surgery evaluation
Seizure
(2017) - et al.
Effects of electrode properties on EEG measurements and a related inverse problem
Med Eng Phys
(2000) - et al.
The peri-electrode space is a significant element of the electrode-brain interface in deep brain stimulation: a computational study
Brain Res Bull
(2007) - et al.
Electrode and brain modelling in stereo-EEG
Clin Neurophysiol
(2012) On electrical potentials observed at a distance from intracranial electrode contacts
Clin Neurophysiol
(2010)- et al.
Concerning the observation of an electrical potential at a distance from an intracranial electrode contact
Clin Neurophysiol
(2009) - et al.
Measuring the local electrical conductivity of human brain tissue
J Appl Phys
(2016)