Combining EEG and eye movement recording in free viewing: Pitfalls and possibilities

Abstract

Co-registration of EEG and eye movement has promise for investigating perceptual processes in free viewing conditions, provided certain methodological challenges can be addressed. Most of these arise from the self-paced character of eye movements in free viewing conditions. Successive eye movements occur within short time intervals. Their evoked activity is likely to distort the EEG signal during fixation. Due to the non-uniform distribution of fixation durations, these distortions are systematic, survive across-trials averaging, and can become a source of confounding. We illustrate this problem with effects of sequential eye movements on the evoked potentials and time-frequency components of EEG and propose a solution based on matching of eye movement characteristics between experimental conditions. The proposal leads to a discussion of which eye movement characteristics are to be matched, depending on the EEG activity of interest. We also compare segmentation of EEG into saccade-related epochs relative to saccade and fixation onsets and discuss the problem of baseline selection and its solution. Further recommendations are given for implementing EEG-eye movement co-registration in free viewing conditions. By resolving some of the methodological problems involved, we aim to facilitate the transition from the traditional stimulus-response paradigm to the study of visual perception in more naturalistic conditions.

EEG Negativity in Fixations Used for Gaze-Based Control: Toward Converting Intentions into Actions with an Eye-Brain-Computer Interface

“…Also, additional experiments and analysis are needed to rule out the possibility that it was an extreme eye position that enabled high effect in Button fixations and to rule out possible effects that the preceding high-amplitude saccade could have on it or (through its effect on the lambda wave) even on the beginning of the baseline (cf. Nikolaev et al, 2016). …”

“…Such contamination cause problems either when the EOG artifacts appear in the time intervals that are the main focus of analysis or when they are present during intervals used for baseline correction (Finke et al, 2016; Nikolaev et al, 2016). In addition, EEG components within the first 200 ms after fixation onset depend on the size of saccade preceding the fixation (Nikolaev et al, 2016). In our study, only the 200–500 ms interval relative to fixation start was used for FRP analysis and classification of fixations to avoid any interference.…”

“…If the EEG markers of intention to act will turn to be sensitive to various factors (properties of fixations, fixated objects, visual context, previous fixations, previous or planned saccades, variations in difficulty of maintaining a fixation intentionally, etc. ; see, e.g., Nikolaev et al, 2016), different classifiers could be used for each specific case.…”

Eye movements in children during reading: a review

“…For the time being, this technique is faced with specific methodological challenges, particularly due to effects of overlap between EEG responses elicited by successive fixations, that can interfere with effects resulting from experimental conditions. However, as more and more solutions are now proposed (Devillez, Guyader, & Guérin-Dugué, 2015;Nikolaev, Meghanathan, & van Leeuwen, 2016), we hope that this technique will shed new light on the issue of the development of reading skill, to better understand how both linguistic processing and eye-movement control change with development.…”

Actes du Symposium international sur la litéracie à l’école = International Symposium for Educational Literacy (SILE/ISEL)

“…For the time being, this technique is faced with specific methodological challenges, particularly due to effects of overlap between EEG responses elicited by successive fixations, that can interfere with effects resulting from experimental conditions. However, as more and more solutions are now proposed (Devillez, Guyader, & Guérin-Dugué, 2015;Nikolaev, Meghanathan, & van Leeuwen, 2016), we hope that this technique will shed new light on the issue of the development of reading skill, to better understand how both linguistic processing and eye-movement control change with development. Résumé : Les dictées réussies de pseudomots dépendent de conversions phonographémiques maitrisées mais aussi de phases de test et d'encodage proches.…”