To identify differences in the broad-band visual evoked electroencephalographic responses to intermittent photic stimulation (IPS) between patients with and without photosensitive epilepsy (PSE).

IPS evokes time-locked steady-state visual evoked potentials (ssVEPs), or ‘photic driving’. Although only the fundamental ‘following’ response is clinically assessed, the posterior cortex also responds non-linearly, producing harmonics and subharmonics. Frequencies composing the photo-paroxysmal response (PPR) often differ markedly from the IPS frequency, for reasons largely unexplored. In this exploratory study, we examined broadband responses to multiple IPS frequencies, hypothesizing that PPR-positive subjects would show distinct response profiles from PPR-negative subjects even outside PPR-provoking frequencies.

EEG data from 15 subjects (five PSE, five idiopathic generalized epilepsy [IGE] without PPR, five non-IGE controls) were analyzed. Power spectra from 10-s occipital epochs across 13 stimulation frequencies (2, 3, 8–16, 21, 24 Hz) were computed. We developed a novel metric, the Weighted Adaptive Driving Index (WADI), which quantified the proximity of three largest amplitude response peaks to ‘frequency mix (fundamental, harmonics, subharmonics). It was adapted to three stimulation frequency ranges: low (1-8 Hz), middle (8-16 Hz) and high (>16 Hz) that was weighted by peak rank. Scores range from 0 (decoupled) to 1.75 (well-ordered).

Mean (SD) WADI scores were significantly lower in PPR-positive IGE (0.46 ± 0.13) compared to both PPR-negative IGE (0.83 ± 0.03; p<0.0003) and non-IGE controls (0.95 ± 0.22; p<0.003). No significant difference was found between PPR-negative IGE and non-IGE subjects. A WADI threshold of 0.7 was 100% sensitive and specific in discriminating PPR-positive and PPR-negative patients.