Authors :
Andrei Barborica, PhD – University of Bucharest; Ioana Mindruta, MD, PhD – Neurology – Emergency University Hospital Bucharest; Christian Bénar, PhD – Inst. Neurosci. Syst. – Aix Marseille University; Victor Lopez-Madrona, PhD – Inst. Neurosci. Syst. – Aix Marseille University; Irina Oane, MD, PhD – Neurology – Emergency University Hospital Bucharest; Felicia Mihai, MSc – Physics / Biomedical Engineering – University of Bucharest; Cristian Donos, PhD – Physics / Biomedical Engineering – University of Bucharest; Constantin Pistol, MSc – Physics / Biomedical Engineering – University of Bucharest
Rationale: Scalp EEG provides information about putative sources in the brain. Relatively few studies to date studied correspondence between intracranial activity and their visibility on scalp recordings. Single-pulse intracranial stimulation has been used as ground truth to evaluate the performance of distributed source localization algorithms (Mikulan et al. Sci. Data. 2020;28:127). We report the use of repetitive high-frequency intracranial stimulation aiming at evaluating the visibility of the deep signals on the scalp, depending on their frequency content and distance between stimulation site and scalp EEG.
Methods: Seven patients undergoing presurgical evaluation for drug-resistant epilepsy using stereo-EEG method had up to 40 scalp electrodes placed simultaneously with intracranial electrodes for up to 3 days. Scalp responses to high-frequency intracranial stimulation (constant current 0.25-3 mA, biphasic 1-ms pulses having alternating polarity, 43.2 Hz, train duration 5s) were recorded using two XLTek Quantum 128 systems (Natus, Middleton, WI), for EEG/SEEG, having independent references. Raw data was acquired at 4096 Hz and systems were synchronized between them and with the cortical stimulator (Guideline 4000 - FHC, Bowdoin, ME) via digital trigger signals sent from the stimulator. We have analyzed the frequency content of the scalp recordings by calculating the mean frequency of the artefactual components, propagating through volume conduction, which are distinct from tissue response components for alternating polarity stimulation
(Barborica et al. Hum Brain Mapp. 2022;43:1657), as a function of the distance between stimulation site and scalp electrodes.
Results: A total of 1406 scalp responses to 42 intracranial stimulations in 3 patients were analyzed (Figure 1). The spectrum of the responses to individual stimulations, presented in Figure 1A, and of the artefactual components only, shown in Figure 1B, exhibited a per-sensor variability. However, a group-level analysis showed that this variability, captured by the mean frequency of the artefactual components is not significantly correlated with the distance between scalp sensors and stimulation site, linear regression fit: r=-0.02, p< 0.42 (Figure 1C).