Comparison of saccadic peak velocities between electro-oculography, video-oculography and scleral search coil

Peak velocity of saccadic eye movements is a crucial motor parameter in clinical neurology and oculomotor research. However, it may differ depending on the technique recording eye movements. Electrooculography (EOG) is easy and comfortable to handle but its signal is noisy and it may overestimate peak velocity of upwards saccades (Yee et al. 1985). Videooculography (VOG) and the magnetic Scleral Search Coil System (SSC) both show high spatial and temporal resolution but also have some technical flaws. VOG does not work when eyes are closed. SSC is invasive, expensive and decreases peak velocity compared to VOG (Frens and van der Geest 2002). For the first time, we recorded saccades and compared them intraindividually using all three recording techniques with open and closed eyes.

We compared these techniques in two sessions, either recording EOG and SSC or EOG and VOG (EyeLinkII). 10 subjects participated in this study (mean 24.5 years, range 19 to 42). Saccadic eye movements were recorded during four conditions: (i) reflexive saccades, (ii) self-paced saccades towards targets of a stationary array, (iii) imagining this array with open eyes and (iv) with closed eyes. These conditions were chosen as it is known that velocity decreases in this order (Sprenger et al. 2009) suggesting that there may be interactions using different techniques.

EOG recordings revealed systematically higher peak velocities than both other techniques: with open eyes by about 10%, with closed eyes by 27%. However, the difference is only significant for condition (ii) (p<0.048). Saccadic peak velocities obtained by VOG recordings were higher by about 12.8% (self-paced, imagined targets: 20%) than those by SSC recordings (p=0.001). Generally, eye movements towards a visible target were faster than those to an imagined target with open or closed eyes, irrespective of the recording technique (p<0.000).

Unlike previously published, higher peak velocities using EOG were only found with self-paced but not reflexive saccades. Thus, EOG is a valid method for analysing eye movements not only with open eyes but also when they are closed. Additionally we were able to compare SSC and VOG: our results are in accord with a related study (Frens and van der Geest 2002) which showed slower peak velocities of reflexive saccades (5%) using SSC when compared to VOG recordings. We extend this finding to a general reduction of peak velocity by SSC – also without visible stimuli or closed eyes. In conclusion, both SSC and VOG are highly accurate recording methods but have their applicability limits (duration, eye closure). EOG tends to overestimate peak velocities but only with self-paced saccades. Moreover, EOG provides reliable saccade velocities even with closed eyes. This should be taken into consideration when saccade velocity data of different studies are compared which are recorded by different techniques.