F82 EEG voltage display of focal epileptiform transients: Applications to correlation with gross generator anatomy

Abstract

EEG voltage display for examination of focal interictal epileptiform transients (FIET) was developed to improve reliability of differentiation of EEG spikes/sharp waves (SSW) from FIET not predictive of epilepsy (Matsuo 2017, http://www.AESnet.org). Electrode array consisted of 23 head-surface placements, 19 10–20 System placements, and 2 south-hemisphere pairs, zygomatic (ZY) and mastoid (MS). Each of 23 head-surface electrodes served as common reference (CR) against 23 inputs for 529 (23 × 23) CR derivations, which were grouped by CR and reformatted into 23 polygraphic channel-overlays (PGCO). Operation of comprehensive CR replacement (PGCO-CCRR) utilized Persyst 13 edit functions. Solid angle theorem applies to interpretation of PGCO-CCRR. Objective of this study was to explore broader applications of PGCO-CCRR to examination of FIET with waveform complexity beyond 4 major SSW components (base-peak-trough-wave: B-P-T-W), as previously investigated in 121 FIET. Digital EEG sample pages (typically 10 to 20-s segment in anterior-posterior serial bipolar derivations) were reviewed and 2 groups of EEG features were sought. (1) Rare FIET suggestive of non-vertical orientation of equivalent current dipole secondary to cerebral anatomy that deviates from spherical model at major brain fissures, and (2) waveform complexity suggestive of contribution of multiple generators, commonly seen in periodic discharges (PD), particularly focal PD exhibiting traveling wave features. FIET of interest were prepared in 3 display alternatives, (1) anterior-posterior serial bipolar derivations, (2) common average reference (CAR) derivations, and (3) PGCO-CCRR. Time scale and filter settings were carefully controlled so that accuracy of cursor placement could be assured among 3 displays, when made into composite for visual inspection. Display gain was independently adjusted for best image quality. Major (B-P-T-W) and minor (notches and/or turns) FIET components identified in conventional displays were readily correlated in PGCO-CCRR. Minor latency differences could be appreciated in PGCO-CCRR by inspection alone as phase-shift, while conventional polygraphic examination required multiple latency measurements for confirmation. PGCO-CCRR provided one-step alternative for evaluation of non-vertical dipole source configuration. Some PD revealed previously noted pre-base shift (Matsuo 2012, http://www.AESnet.org). PGCO-CCRR could reliably demonstrate subtle features that are relevant for source localization and neuroanatomical correlation. PGCO-CCRR may be more reliable in EEG screening for benign epilepsy with Rolandic spikes. Inclusion of south-hemisphere electrode coverage is important. Increased temporal sampling frequency can be expected to improve display resolution, while increased spatial sampling may not, unless signal magnitude is favorable against background.