EEG source localization using HD-EEG is a useful noninvasive tool to localize seizure onset and interictal discharges. Minimum number of electrodes needed and the point of measurement on the spike for optimal localization are still unclear.

A total of 37 interictal and 10 ictal analyses from 33 medically intractable epilepsy patients were performed. HD-EEG recording was performed using the 10-10 system with 76 electrodes. LD-EEG including subtemporal electrodes was simulated by subtracting the 10-10 leads. EEG analysis used Curry 8.0 software. Interictal spikes  were averaged and dipole and sLORETA solutions generated at both 50% and 75% of peak amplitude in the rising phase of the spike. The distance between the HD and LD dipoles and sLORETA peak intensities were measured and were considered concordant if the distance of 0-10 mm, partially concordant if 10-50 mm, and discordant if > 50 mm or in a different lobe.

The median distance between HD-EEG and LD-EEG interictal dipoles at 50% of peak was 8.5 mm, and 5.3 mm at 75%. The median distance between sLORETA for HD-EEG and LD-EEG was 3.4 mm at 50% and 1.7 mm at 75% thresholds. Median distance between 50% and 75% measurements for HD-EEG was 6.1 mm and for LD-EEG 6.1 mm.  For 50% and 75% thresholds, the concordance rates for dipole analysis were 56% and 72% (p< 0.001) respectively, and 81% and 72% (p< 0.01) respectively for sLORETA analysis.  

LD-EEG including the sub-temporal chain is comparable to HD-EEG for EEG source imaging. Concordance between LD-EEG and HD-EEG were higher in dipole analyses at 75% of peak compared to 50% but sLORETTA concordance was higher for 50% than 75% analysis.