Blog

Jul 12, 2023

Face familiarity revealed by fixational eye movements and fixation

Scientific Reports volume 12, Article number: 20178 (2022) Cite this article 741 Accesses 1 Citations 1 Altmetric Metrics details Event-related potentials (ERPs) and the oculomotor inhibition (OMI) in

Scientific Reports volume 12, Article number: 20178 (2022) Cite this article

741 Accesses

1 Citations

1 Altmetric

Metrics details

Event-related potentials (ERPs) and the oculomotor inhibition (OMI) in response to visual transients are known to be sensitive to stimulus properties, attention, and expectation. We have recently found that the OMI is also sensitive to face familiarity. In natural vision, stimulation of the visual cortex is generated primarily by saccades, and it has been recently suggested that fixation-related potentials (FRPs) share similar components with the ERPs. Here, we investigated whether FRPs and microsaccade inhibition (OMI) in free viewing are sensitive to face familiarity. Observers freely watched a slideshow of seven unfamiliar and one familiar facial images presented randomly for 4-s periods, with multiple images per identity. We measured the occipital fixation-related N1 relative to the P1 magnitude as well as the associated fixation-triggered OMI. We found that the average N1-P1 was significantly smaller and the OMI was shorter for the familiar face, compared with any of the seven unfamiliar faces. Moreover, the P1 was suppressed across saccades for the familiar but not for the unfamiliar faces. Our results highlight the sensitivity of the occipital FRPs to stimulus properties such as face familiarity and advance our understanding of the integration process across successive saccades in natural vision.

Traditional neurophysiological studies of object perception typically probe the visual system with flashed stimuli, to mimic the saccade-induced transients of natural vision, measuring event-related potentials (ERPs). More recently, fixation-related potentials (FRPs) have been used to study vision in more natural settings of free viewing, demonstrating the advantages as well as the limitations of this method1. In general, these studies show results that are consistent with the event-related measurements; however, none of these studies probed face familiarity. Unlike traditional ERPs that used briefly flashed visual transients presented at the observer’s central visual field, in natural settings the scene is scanned over time via saccades following a peripheral preview.

Accumulating evidence from recent free-viewing studies suggests that the brain’s response following a saccade, termed Fixation-related Potentials (FRP), exhibits electrophysiological components very similar to ERPs. For example, the saccade-related occipital lambda response reflects the same information processing as the classic VEP P12. Recent studies that examined the face-selective activity at the lateral temporo-occipital electrodes, N1703 found a resembling increased negativity for faces in free viewing conditions4,5. More classic findings were replicated in free viewing conditions, such as centro-parietal P300 elicited by target detection in a visual search6 and the N400 priming effect in natural reading7,8. Combining EEG and eye-tracking measurements to study face familiarity enabled us to cross examine eye movement and electrophysiological changes over time, which are influenced by habituation and prior knowledge.

Microsaccades (MS) are miniature saccades, with an average size of < 0.5 dva, generated by neural activity in the superior colliculus (SC)9,10. They occur during fixation, with a rate of one or two per second. Microsaccades, as well as saccades, are known to be inhibited momentarily (Oculomotor-Inhibition, OMI)11,12,13,14,15,16 by stimulus presentation with a later release latency affected by the stimulus properties, attention and expectation. Whereas stimulus saliency, such as high contrast, is known to shorten the inhibition17, prolonged inhibition was found in response to deviants18. Although most studies used flashed stimuli, we have recently found similar inhibitory patterns in free viewing in response to stimulus saliency19.

Faces are considered complex stimuli that are processed holistically in posterior brain regions including the occipital face area (OFA), the fusiform face area (FFA), and the posterior superior temporal sulcus (STS), which form a core system for encoding the visual appearance of faces20. This activity is modulated by top-down feedback from personal knowledge and emotional responses to alter the processing of familiar faces21. Face familiarity is of a major interest in research on concealed memories; the Concealed Information Test (CIT) is used for testing the familiarity of suspects to specific people and objects22. Our recent study found prolonged OMI at fixation for a masked familiar face23, which allowed us to reliably detect identity in a concealed information test24. Next, we will review in more detail the currently known oculomotor and the ERP measures of familiarity.

Most of the experiments examining ocular measures in response to face familiarity used flashed stimuli. Rosenzweig & Bonneh23 measured the oculomotor inhibition of both microsaccades and blinks in response to masked novel and universally learned23 or recently learned24 familiar faces, and found prolonged OMI for the familiar faces in passive viewing. However, few studies have focused on different aspects of gaze fixations; one study claimed that the first two fixations are critical for revealing familiarity25, and other studies suggested that fewer fixations and longer fixation durations are the key for familiarity26,27. In contrast, another study reported more fixations and longer fixation durations for familiar faces with a CIT paradigm28; however, when participants had to memorize faces, familiar faces initially attracted their gaze, but later triggered fewer fixations and shorter fixation durations29. Overall, the eye movement behavior in relation to face familiarity depended on the instructions given by the researcher.

Previous electrophysiological familiarity studies measured the late event-related (ERP) response including the N25030,31 and the P300 contextual response. The face-sensitive N170 component, which reflects the structural encoding of faces prior to person identification32,33 was also tested; however, there were conflicting results. The majority of the studies did not find a familiarity effect30,31,34,35, R. N.36; nevertheless, a few studies found an effect showing either larger N170 magnitudes for familiar faces37,38,39,40,41 or smaller than for unfamiliar ones42,43.

To date, little or no research has focused on earlier visual responses regarding face familiarity. However, a few studies examined the occipital P1 and N1 components using unfamiliar faces as stimuli, suggesting an early coarse processing of faces prior to face identification44,45. Here, we studied the fixation-triggered early posterior components that reflect feature and structure visual processing, and the timing of microsaccades during fixation, as well as saccades, while participants freely viewed large images of both famous and unfamiliar faces presented for several seconds. We expected to find differences in the modulation over time, across successive saccades of the early occipital as well as the oculomotor responses by prior knowledge in the context of familiarity. Thus, the aim of this study was to investigate whether the FRP and OMI in free viewing are sensitive to face familiarity.

A total of sixteen observers were recruited for the experiment: eight females and eight males, aged 21–44. One participant was omitted from the data analysis due to the low quality of the data (more than 50% bad data). All participants had normal or corrected-to-normal vision and were naïve to the purpose of the study. The experiments were approved by the Bar-Ilan Internal Review Board (IRB) Ethics Committee. All participants gave written informed consent, and all the experiments were conducted according to the IRB guidelines.

The study combines eye tracking and electrophysiology recordings in free viewing synchronized by a split trigger to both systems. A wireless 8-channel headset with dry electrodes (Cognionics) was used for the EEG recordings and the Eyelink 1000 plus (SR Research) for eye tracking, both with a sampling rate of 500 Hz. The EEG signal had a built-in reference using channels placed on both ears. Stimuli were displayed using the in-house integrative stimulus presentation and analysis tool (PSY), developed by Y.S. Bonneh, by a 100 Hz calibrated 24-in FHD LCD monitor (Eizo Foris fg2421), at 0.6 m distance. The experiment was administered in dim light. We used a 35 mm lens positioned 0.52 m from the participant’s stabilized head using a chin rest. All recordings were done binocularly, with analyses done on data from the left eye. A standard 9-point calibration was performed before each session.

For the saccade detection, we used an algorithm introduced by Engbert and Kliegl46, which is based on eye movement velocity. Microsaccades were detected as movements exceeding 8 SD of the mean velocity in 2D velocity space, as in23,47. A velocity range of 8°/s–150°/s, an amplitude range of 0.08–1°, and a minimum duration of 9 ms were allowed for the microsaccades. We calculated the Warren Sarle’s bimodality coefficient (BC) of the saccade amplitude data, which is associated with the data skewness and kurtosis using a Matlab function by Hristo Zhivomirov48. BC has a range of 0–1, where values greater than ~ 0.555 (the value for the uniform distribution) indicate bimodal or multimodal data distributions49. Eye tracking epochs were extracted, triggered by saccade (> 1 dva) landing time, as in our previous study19, in a range of − 0.2 s to 0.8 s relative to the fixation onset with some overlap between epochs. This was taken into consideration when computing the microsaccade Reaction Time (msRT). The msRT was calculated for each epoch relative to the fixation onset in a predefined time window, as the latency of the first microsaccade in that window. The first fixation per trial was always ignored to avoid the flash effect on the OMI. The microsaccade RTs (msRT) were averaged across the epochs of each condition within observers and then averaged across observers, with error bars computed across observers on demeaned (within observer) data, with a correction factor (multiplied by √(n/(n − 1))). This method for computing the error bars allows a better representation of within-participant effects (Cousineau & Morey’s method50; see also Bonneh et al.17. The inter-saccade interval, termed the fixation duration, was calculated as the time interval between the current fixation onset and the next fixation onset, including only MS-free fixations.

The EEG data were filtered using a 0.1 Hz high-pass and 30 Hz low-pass cutoffs. FRP epochs were created, as was done for the eye-tracking data, triggered by the saccade (> 1 dva) landing time in a range of − 0.1 s to 0.3 s, relative to the fixation onset to minimize overlapping data between epochs. The first fixation per trial was always ignored to avoid the flash effect. Since our EEG system had only eight channels and is not equipped with EOG electrodes, ICA and deconvolution methods for correcting ocular artifacts were not used. Instead, overlapping data points between proximal saccades were excluded on both epochs triggered by those saccades, as well as epochs with blinks or microsaccades that occurred at less than 200 ms after fixation onset. We focused on the early components at occipital electrodes O1 and O2, which are less prone to be affected by ocular artifacts. We then computed the positive and negative peaks in a predefined time range. The P1 peak was measured using a 50–150 ms time range, and the N1 was measured in a 100–200 ms window with no baseline correction. Finally, we calculated the baseline-corrected peak-to-peak N1 relative to the P1 magnitude (N1-P1). Peak extraction was optimized by setting an individual time range for each observer at around their average peak latency, within the predefined time range, from all the conditions combined. This was done to avoid using a long time-range to overcome the latency differences across observers, which would increase the false peak discoveries. Extreme value artefacts were not allowed using a peak magnitude threshold exceeding ± 50 µVolts. To ensure that we used a similar number of epochs per participant, we used an estimation of an average of 3 saccades per second to include only the first 12 epochs per trial (trial duration = 4 s), in the final analysis.

Usually, the statistical analysis of the variance (One-way ANOVA) and the Tukey multiple comparisons post-hoc tests were performed using Matlab. We first verified that the measured value distributions of different conditions come from normal distributions with equal variance. Another statistical method that was used is the Linear Mix Model (LMM)51. The responses were fitted to a simple model of maximum likelihood with the serial saccade number, or the saccade size used as the predictor variable, and the observer’s variability was set as the random effect. In addition, we computed Pearson's linear correlation coefficient (r2) for the group averages of each plot.

Observers freely watched a slideshow of seven unfamiliar and one familiar world leader’s facial images presented randomly for four-second periods, with multiple different images per person (see Fig. 1). All images were chromatic with 600 × 800 minimum resolution, taken from the internet and were radially cropped around the head with a radius of 10 dva. The use of 5 different images per identity was done to reduce the effects of low image attributes and facial expressions from a specific image. In addition, we verified that the familiar face images were not appreciably different from the non-familiar face images regarding the low-level properties of luminance and RMS contrast. A task was not used, and the observers were just instructed to freely inspect the images.

Stimuli and procedure. Observers freely inspected large facial images, presented in random order for four seconds each. In total, eight different identities were chosen of seven unfamiliar and one familiar world leader with five different images per identity. Note: the above images are of the authors who gave an informed consent to publish the images in an online open-access publication. They are used here for illustration only and were not used in the experiment.

We used static displays of faces, presented for four-second periods, while the observer viewed them freely without a task (see Fig. 1, “Stimuli and procedure”). Fixation-related potentials (FRPs), triggered by saccade landing time, were calculated, with a primary focus on early occipital components as well as Oculomotor-Inhibition (OMI), computed here as the fixation duration (MS-free fixations only) or the timing of the first microsaccade following the fixation onset, excluding corrective microsaccades occurring proximate to the preceding saccade landing (see the “Methods”).

The main familiarity effect was found at the posterior right hemisphere via the early occipital responses measured at the O2 electrode (see Fig. 2a for a view of the EEG channel locations). The FRP waveforms in Fig. 2a (left and right panels) were baseline corrected to the mean value of the data points preceding the fixation-inducing saccade spike potential (− 100 to − 75 ms). Then, we computed the peak magnitude (see the “Methods”) for the Lambda response (P1) peaking around ~ 90 ms, and the N1 peaking around ~ 140 ms after fixation onset (see Fig. 2b,c). We then computed a baseline-corrected N1 by subtracting the P1 magnitude (peak-to-peak) per fixation-related epoch (see the “Methods”), which allowed us to test the combined FRP N1 and the P1 response. We hypothesized that the N1, which is often associated with a bottom-up prediction-error (PE), would reflect less PE for familiar faces with smaller magnitudes due to prior visual knowledge, which facilitates oculomotor dynamics and visual processing.

Basic FRP results. (a) Scalp EEG channel topography and O1, O2 FRP results for all the unfamiliar faces (thin lines) and one familiar face (the thick blue line) averaged across observers (N = 15). The vertical dashed line denotes the fixation onset. (b) Results for the O2-FRP P1 peak latency (left), magnitude (center), and P1 averaged across observers’ magnitudes for each of the identities (right) with error bars calculated over the demeaned data showing a significantly smaller magnitude for the familiar identities (F(7112) = 3.11, p < 0.005, One-way ANOVA). (c) Results for O2-FRP N1 peak latency (left), magnitude (center), and N1 averaged across observers’ magnitudes for each of the identities (right) with error bars calculated over the demeaned data showing a non-significant smaller magnitude for the familiar identities (p < 0.6).

The baseline-corrected N1 (N1-P1) was calculated relative to the P1 magnitude for each epoch, with a mean value of 18.8 µVolt ± 9 SD. The N1-P1 magnitudes were calculated for all conditions (8 different face identities) averaged across observers (N = 15) with error bars calculated over the demeaned data. We examined the importance of defining a threshold discriminating between fixation-inducing saccades and smaller microsaccades. We tested the familiarity effect by taking all saccade sizes (> 0.08 dva) as triggers for FRP. The results are shown in Fig. 3a. A significantly smaller N1-P1 magnitude was found for the familiar identity, compared with each of the unfamiliar identities, p < 0.015 (F(7112) = 2.62, One-way ANOVA). This effect was much smaller than the effect induced by larger (> 1 dva) saccades (p < 0.0005 (F(7112) = 4.04, One-way ANOVA, see Fig. 3b). A multiple comparisons test yielded three out of seven significantly different groups from the familiar identity, with an illustration of the confidence intervals. To account for the individual contribution to the results, a detailed observer scatter plot with a different color for each participant and a dot for an unfamiliar identity N1-P1 magnitude, compared with the familiar one, indicated that most of the dots are above the diagonal, signifying a larger magnitude for the unfamiliar one (see Fig. 3c).

The baseline corrected N1 familiarity effect. (a) N1-P1 magnitude for each of the 8 identities averaged across observers with error bars calculated over the demeaned data using epochs triggered by saccades with sizes ranging from 0.08 to 20 dva. The results show a significantly smaller magnitude for the familiar identity (p < 0.015, One-way ANOVA) with post-hoc multiple comparisons tests yielding two groups that significantly differ from the familiar identity, with 95% confidence intervals. (b) The same as (a) but with saccade sizes ranging from 1 to 20 dva. The results show a smaller magnitude for the familiar identity (p < 0.0005, One-way ANOVA), with three groups that significantly differ from the familiar identity (post-hoc tests). (c) A detailed observer scatter plot with a different color per participant and 7 dots showing the unfamiliar identities; the N1-P1 magnitude was compared with the familiar identity, showing that most of the dots are above the diagonal, signifying a larger magnitude for the unfamiliar identity. The dot size was reduced in crowded areas for better visibility. (d) The FRP N1-P1 familiarity effect (familiar denoted in blue and all the unfamiliar combined in red), was tested (Paired t-test) by systematically changing the threshold for the minimum saccade size taken as a trigger for the FRP. The p-values are plotted in light gray.

Finally, we investigated the effect of systematically changing the threshold for the minimum size of the fixation-inducing saccades, on the FRP N1-P1 from 0.2 to 2 dva. The results appear in Fig. 3d. Importantly, we found that the N1-P1 value was not specific to the 1 dva threshold for the inducing saccade size, and it changed gradually, with the strongest effect (p-value) found in the range of 0.7–1.4 dva. Note that increasing the threshold also results in a reduced number of epochs and statistical power, whereas decreasing the threshold may add epochs induced by corrective saccades, which by themselves, do not produce a familiarity effect (e.g. 0.08–0.5 dva). The 1 dva threshold seems optimal, although the results remain robust and are in a much wider range.

We tested whether the adaptation of the occipital activity over successive saccades affects the FRP signal differently for the two categories. We found that the average P1-magnitude as a function of the serial saccade number was attenuated across successive saccades for the familiar identity via the O1 and O2 electrodes (p = 0.04, p = 0.0042, respectively. Linear Mixed Model), but not for the unfamiliar identity, with all identities combined (Fig. 4a and Fig. 4d). We then calculated the P1 (via O2) linear fit slopes separately for each of the identities and found that the average across observer negative slope was the largest for the familiar identity (Fig. 4b). A paired t test comparison was performed between the familiar slope values and the combined unfamiliar ones, showing nearly a significant difference (p = 0.06; see Fig. 4c), with a calculated medium effect size (ES = 0.72, Cohen’s d) and the area under the ROC curve (AUC = 0.66). The baseline-corrected N1 also shows a similar adaptation for the familiar identity (Fig. 4e) as expected, because it was calculated relative to P1, peak-to-peak.

P1 adaptation effect. (a) O2 FRP P1 magnitude as a function of the serial saccade number showing an adaptation effect across the first seven saccades, for the familiar (p = 0.0042, Linear Mixed Model) but not for the unfamiliar identities combined. (b) Slopes of the O2 FRP P1 magnitude linear fit across successive saccades for each of the identities. (c) A paired t test comparison between the familiar identities and the combined unfamiliar P1 slopes showing an almost significant difference (p = 0.062). (d) O1 FRP P1 magnitude as a function of the serial saccade number showing an adaptation effect across the first seven saccades, for the familiar (p = 0.04, Linear Mixed Model) but not for the unfamiliar identities combined. (e) O2 FRP N1-P1 magnitude as a function of the serial saccade number showing an adaptation effect across the first seven saccades, for the familiar (p = 0.02, Linear Mixed Model) but not for the unfamiliar identities combined. In (a), (d) and (e), the data were averaged on observers and error bars were computed on the demeaned data.

A recent study by Rosenzweig & Bonneh23 showed prolonged OMI for familiar faces briefly presented and masked. We wanted to determine whether microsaccade latencies after a saccade are also increased when a famous face is freely inspected for several seconds relative to an unfamiliar one.

The results for the OMI are shown in Fig. 5. Fixation duration as well as the saccade size histograms for all observers and conditions combined are shown in Fig. 5a,b. The mean fixation duration was 350 ms ± 175 SD. The saccade size distribution reflects a bimodal distribution of saccades and microsaccades (see Fig. 5b, right panel) with a calculation of the bimodality coefficient yielding a significance, BC = 0.6, suggesting bimodality (see the “Methods”). The mean microsaccade size was 0.42 dva ± 0.24 SD. We computed the msRT (microsaccade reaction time, see the “Methods”) as the first microsaccade (< 0.6 dva) occurrence from 150 to 600 ms after the fixation onset, ignoring the corrective microsaccades that may occur immediately after the saccade. Figure 5c shows the average msRT across observers for each of the identities, showing decreased OMI for the familiar identity, which is opposite to the results obtained with flashed faces23. Significance was assessed using one-way ANOVA, F(7112) = 2.84 and p < 0.0093. The post-hoc Tukey method tests indicated that the familiar identity was significantly different only from one other unfamiliar group. The observer was denoted by color, in the scatter plot (Fig. 5d), with a dot per identity, compared with the familiar identity. Interestingly, it was shown that most observers had longer msRTs for the unfamiliar identity (the dots above the symmetry line).

The Oculomotor Inhibition (OMI) familiarity effect. (a) A fixation duration histogram with a mean duration of 350.5 ms ± 175.3 SD. (b) A saccade size histogram (left) with a mean size of 2.96 dva ± 3 SD and a Bimodality-Coefficient of 0.6, suggesting that the saccade size have bimodality distributions. Magnification of the region between 0 and 2.5 dva reveals a second peak (right). (c) The OMI (msRT 150–600 ms) results for each of the 8 different face identities averaged across observers (N = 15) showing a significantly shorter msRT for the familiar (p < 0.0093, One-way ANOVA, 95% confidence-intervals). (d) A detailed scatter plot with a different color per participant and a dot for an unfamiliar identity’s msRT, compared with the familiar. Note that most of the dots are above the diagonal, signifying a longer OMI for the unfamiliar identity. The dot size was reduced in crowded areas for better visibility. (e) A scatter plot with a dot for an unfamiliar identity; each of the observers showed a positive correlation (R = 0.31, p = 0.001, Pearson correlation) between the msRT and N1-P1 magnitude after demeaning within observer (see the “Results”). (f,g) The fixation duration after large saccades (> 0.08 dva in (f) and > 1 dva in (g)), for each of the 8 identities, averaged across observers using a 200–600 ms duration range and MS-free fixations only in (g). Like msRT, the fixation duration is shorter for the familiar, but the results were insignificant. (h) The same as in (e) but for fixation duration. The results show a nonsignificant relation (R = 0.17, p = 0.078).

We found that the distinction between small and larger saccades was also critical for the OMI effect. When examining the intervals between all saccades (the fixation duration) with sizes from 0.08 and from 1 dva (Fig. 5f,g, respectively), the familiar had a shorter duration; however, there was no significant familiarity effect on the fixation duration. Finally, we examined the possible correlation between the FRP and OMI measures (Fig. 5e,h). To control for irrelevant factors affecting the N1-P1 and OMI values, the data were demeaned within observers and within saccade size bins (1 dva). Then the grand average from all observers and saccade sizes was added. For this analysis we also excluded fixations triggered by saccade sizes > 8 dva. The msRT/fixation-duration and the N1-P1, grouped by unfamiliar identity and each of the observers, show a positive correlation (R = 0.31, p = 0.001, and R = 0.17, p = 0.078, Pearson’s correlation).

Previous FRP studies reported that the Lambda response (P1) amplitude increases with larger saccades1,52,53. This suggests that our finding of a decreased N1-P1 peak-to-peak magnitude for the familiar identity could result from smaller saccades that reduce the P1 amplitude. We therefore examined the effect of saccade size on the FRP. Figure 6 shows the effect of inducing the saccade size on the FRP and the relationship between the saccade size and familiarity. The saccade size (> 1 dva) did not differ between familiar and unfamiliar identities when averaged across observers (Fig. 6a), or when plotted for each observer in a scatter plot (Fig. 6b). A significant positive relation (r2 = 0.41, Pearson correlation) of the P1 magnitude and the saccade size (p = 0.0016, LMM) is plotted in Fig. 6c, which is consistent with previous studies. Figure 6d shows that the N1-P1 magnitude was also positively correlated with saccade size (r2 = 0.36, Pearson correlation; p = 0.012, LMM), because it was calculated relative to the P1 magnitude (peak-to-peak). Finally, the corrective microsaccade latencies show a negative correlation with saccade size (r2 = 0.85, Pearson correlation; p = 0.00001, LMM); thus, larger saccades induced faster microsaccade reaction times due to the lower peripheral preview acuity (see Fig. 6e).

Saccade size effect. (a) Mean saccade size per identity, showing no difference between the familiar and unfamiliar identities. (b) A detailed observer scatter plot comparing the average saccade size for the familiar identities with each of the unfamiliar ones, for each observer (a different color). As shown, saccade size did not differ with familiarity, indicated by the balanced distribution of the dots along the diagonal line. (c) P1-magnitudes as a function of inducing the saccade size, averaged across observers, showing a significant positive relation (p = 0.0016, LMM, see the “Methods”). (d) FRP N1-P1 magnitude was also positively correlated with saccade size (r2 = 0.36, Pearson correlation; p = 0.012, LMM). (e) Microsaccade latency (msRT) shows a negative correlation with the saccade size (r2 = 0.85, Pearson correlation; p = 0.00001, LMM).

We investigated how familiarity in passive free viewing affected the early occipital responses that occur prior to face identification, and the effect on the latency of microsaccades, which is known to be affected by attention46,54,55 and expectations18,56,57. Our results revealed a smaller fixation-related N1-P1 magnitude and a shorter microsaccade inhibition (OMI) for the familiar faces, suggesting a lower prediction error while freely viewing familiar faces. Interestingly, the N1-P1 FRP effect of familiarity appeared only at the right hemisphere, which is more specialized, according to some studies on face processing58,59.

We also examined the possible effect of adaptation and saccade size on the FRP and found that whereas saccade size did not differ between conditions and thus affected the FRP for different categories in the same way, the P1 magnitude was consistently adapted for the familiar faces and might have significantly contributed to the FRP familiarity effect. Next, we will discuss the three main factors that could affect the FRP results: (1) Prediction Error, PE, (2) Adaptation, and (3) Saccade size.

We propose here that long-term familiarity provides a type of priming for the viewed image across saccades, revealed by a shorter OMI (see Fig. 5) and smaller occipital P1 and N1 amplitudes (see Fig. 3). These effects could be attributed to the effect of prior knowledge or long-term (over several seconds) priming by exposure to different photos having the same identity. Rare visual or auditory events are known to induce a longer OMI. For example, prolonged microsaccade inhibition has been reported for oddballs in a sequence, a rare blue patch among frequent red patches18, and for auditory deviants57,60,61. More preliminary evidence of prolonged inhibition was found for high-contrast patches among low-contrast patches62 and for temporal oddballs via unpredicted intervals11,12. The ERP Mismatch-Negativity (MMN) is a well-known electrophysiological marker for oddball response. It reflects an automatic change detection mechanism in the auditory domain63,Brain Generators Implicated in the Processing of Auditory Stimulus Deviance: A Topographic Event‐Related Potential Study, 199064,65; it is manifested by a larger N1 negativity, peaking at about 170 ms at the temporal electrodes. Some studies suggest that a correlate of this prediction error generator also exists in the visual domain for visual mismatches termed vMMN66. Infrequent color patterns67, low spatial frequency gratings68, and face and house deviant orientations69 elicited a posterior vMMN. Even facial expressions and emotions were used in an oddball paradigm and elicited vMMN peaking at 100–200 ms70,71,72.

All the above are examples of larger PE, measured by longer OMI and larger N1 amplitudes. Priming by repetition, on the other hand, has an effect that is opposite that of PE, manifested by attenuated responses similar to adaptation, and behavioral facilitation. Previous N170 ERP studies of face familiarity reported priming effects via smaller N170 amplitudes. For example, a study that used moony faces and priming by the same photo or a different photo, but with the same identity, found attenuated N170 amplitudes for the familiar identity, suggesting a top-down feedback effect, whereas a stronger priming effect was found with the same photo42. Short-term, within a second, priming effects, indicated by smaller N170 amplitudes, were reported in a face memory recall task for repeated stimuli, using inverted and contrast-reversed unknown faces44. Another study found identity-specific priming effects via attenuated P1 and N170 using morphed faces and argued that they may be due to low-level visual similarities73. A very recent study by Buonocore, Dimigen et al. found a reduction of the fixation-related N170, following an extra-foveal face preview, and contended that it is due to prediction5.

In our recent paper, fixation-related OMI was measured for low-level stimuli in free viewing19. We found a resembling effect for OMI across saccades in the form of shortening of the OMI by repetition priming. Although this facilitation effect was found for gratings with the same spatial frequency, we believe that eye movement enhancement may occur due to familiarity or because different images have the same identity. Finally, we suggest that the N1 and OMI priming effects we found for the familiar identify did not occur independent of adaptation.

Our results indicated that the occipital P1 for the familiar identity was attenuated across saccades (Fig. 4). At least half of the participants showed a larger P1 adaptation for the familiar than for the unfamiliar identity. This attenuation could be interpreted as a simple reduction of attention for the familiar identity, because P1 reflects low-level features activity; this could also explain the shorter saccadic inhibition because attended stimuli induce longer OMI. However, it could also be related to prediction and priming. Adaptation and priming are difficult to separate; they can co-exist in the current settings and discriminating between the two requires further research. The process of attenuation over successive saccades is reminiscent of the repetition suppression phenomenon, which was found for face category and identity via fMRI and EEG. Image-invariant adaptation for familiar faces, but not for unfamiliar faces, was found in the face-selective regions of the medial temporal lobe, MTL74. An fMRI study found a repetition suppression effect only for famous faces, whereas the opposite was found for unfamiliar ones (R.75. More ERP studies found N170 adaptation to face category76 as well as to face identity77. A later identity-specific adaptation on the ERP was also found over superior occipito-temporal sites at around 200-280 ms78.

There is a known asymmetry of the ventral visual cortex for face processing in the literature. It is evident from lesions to the left/right hemispheres that affect word or face processing58,59. In addition, the N170 also shows lateralization and is more prominent in the right hemisphere for faces than are words3,79. In the current study, adaptation of the P1 FRP for the familiar identity appeared on both the O1 and O2 electrodes, placed at the posterior sites of the two hemispheres (see Fig. 4a,d). However, the N1-P1 FRP effect was found only through O2 (see Fig. 3). This can indicate that adaptation is not the main factor for the N1-P1 FRP effect.

In our study we found that the occipital P1 magnitude increased as a function of the amplitude of the saccade that precedes the fixation (see Fig. 6) as previously reported1,52,53. This suggests that the effect of decreased N1-P1 magnitude for the familiar identity could result from differences in the saccade amplitudes for different identities because the N1-P1 negative peak value was baseline corrected by the preceding P1 peak value, peak-to-peak (see the “Methods”). For this reason, we measured the saccade size for each identity and found no significant difference between different identities (Fig. 6a,b). The results also showed that the microsaccade timing after the saccade had a negative relationship with saccade size, meaning that after a large saccade, the corrective saccade was more immediate. In our OMI measurements we ignored the corrective saccades by including only microsaccade timings over 150 ms post-fixation onset.

Recent studies provide evidence that saccades and microsaccades share a common neurophysiological basis10 and perform similar functions80. This raises the question of whether the distinction we make between saccades with sizes > 1 dva as triggers for fixation-related responses is indeed important and whether the threshold we use is critical for generating the familiarity effect. We first noted that the main theme in the current study as well as in our previous study19 is that each saccade generates a transient stimulus to the visual system, such as the flashed stimuli in the ERP and OMI studies. Microsaccades also generate a transient visual stimulation, but their magnitude is smaller (see Fig. 6d,e, assuming that the occipital FRP magnitude will become smaller below 1 dva, not shown). The use of a 1 dva threshold in the current study was initially derived from a popular definition of microsaccades (e.g.81,82,83) corresponding to the size of the foveola, although other studies use smaller thresholds, e.g. 0.5 dva84,85. Overall, when considering all saccades as fixation-inducing, the FRP familiarity effect was still significant but degraded (compare Fig. 3a,b), whereas the OMI effect became insignificant (Fig. 5f,g). See FRP & OMI familiarity effect pars in Results.

Previous familiarity studies that used eye tracking measurements detected fewer fixations with longer fixation durations when a familiar face was viewed28 and fewer areas of interest on the face when the participants were instructed to conceal their knowledge of familiarity26. With a memory task, familiar faces were less explored perhaps because they are easier to remember29. In our recent study with different eye measurements, a longer OMI for the familiar was found for microsaccades and for blinks when face stimuli were briefly presented and masked23. Here, we also used images with one familiar identity out of eight, as in our recent study, but with a longer presentation duration. We took care of the unequal sampling of the familiar and the unfamiliar faces by comparing the results for the familiar face to each of the equally sampled unfamiliar faces separately, having a chance level of 1/8 (12.5%) to find a difference. We expected to find similar results of longer OMI for the familiar in free viewing, assuming that familiarity would raise associations, episodic-memories, and emotions, thus delaying the next saccade/microsaccade. However, we found the opposite from the expected, shorter saccadic inhibition for the familiar identities, which might be related, as discussed before, to scanning enhancement due to priming/adaptation and feedback from other areas associated with prior knowledge.

When observers free viewed a set of familiar and unfamiliar faces for a few seconds per face, the fixation-related potentials (FRPs) showed a decreased N1-P1 magnitude for the familiar faces at the right occipital electrode and a shorter OMI for the familiar, compared with unfamiliar identities, both indicative of a smaller prediction error. The P1 magnitude for the familiar identity had been suppressed across successive saccades, implying priming or adaptation. In the current study, adaptation of the P1 magnitude for the familiar identity appeared on both hemispheres; however, the decreased N1-P1 magnitude effect was observable only through the right occipital electrode, O2; this suggests that predictive structural face features, reflected by N1 rather than adaptation of low-level features, reflected by P1, is the main factor driving this effect. Overall, the results indicate the sensitivity of the occipital FRP and the OMI in free viewing in relation to face familiarity; this could be used as a novel physiological measure for studying hidden memories.

The experimental datasets generated during the current study will be available from the corresponding author upon reasonable request.

Nikolaev, A. R., Meghanathan, R. N. & van Leeuwen, C. Combining EEG and eye movement recording in free viewing: Pitfalls and possibilities. Brain Cogn. 107, 55–83. https://doi.org/10.1016/j.bandc.2016.06.004 (2016).

Article PubMed Google Scholar

Kazai, K. & Yagi, A. Comparison between the lambda response of eye-fixation-related potentials and the P100 component of pattern-reversal visual evoked potentials. Cogn. Affect. Behav. Neurosci. 3(1), 46–56. https://doi.org/10.3758/CABN.3.1.46 (2003).

Article PubMed Google Scholar

Bentin, S., Allison, T., Puce, A., Perez, E. & McCarthy, G. Electrophysiological studies of face perception in humans. J. Cogn. Neurosci. 8(6), 551–565. https://doi.org/10.1162/jocn.1996.8.6.551 (1996).

Article PubMed PubMed Central Google Scholar

Auerbach-Asch, C. R., Bein, O. & Deouell, P. L. Y. Face selective neural activity : Comparison between fixed and free viewing. Brain Topogr. https://doi.org/10.1007/s10548-020-00764-7 (2019).

Article Google Scholar

Buonocore, A., Dimigen, O. & Melcher, D. Post-saccadic face processing is modulated by pre-saccadic preview: Evidence from fixation-related potentials. J. Neurosci. 40(11), 2305–2313. https://doi.org/10.1523/JNEUROSCI.0861-19.2020 (2020).

Article CAS PubMed PubMed Central Google Scholar

Hiebel, H. et al. Target probability modulates fixation-related potentials in visual search. Biol. Psychol. 138(May), 199–210. https://doi.org/10.1016/j.biopsycho.2018.09.007 (2018).

Article PubMed PubMed Central Google Scholar

Dimigen, O., Sommer, W., Hohlfeld, A., Jacobs, A. M. & Kliegl, R. Coregistration of eye movements and EEG in natural reading: Analyses and review. J. Exp. Psychol. Gen. 140(4), 552–572. https://doi.org/10.1037/a0023885 (2011).

Article PubMed Google Scholar

Niefind, F. & Dimigen, O. Dissociating parafoveal preview benefit and parafovea-on-fovea effects during reading: A combined eye tracking and EEG study. Psychophysiology https://doi.org/10.1111/psyp.12765 (2016).

Article PubMed Google Scholar

Hafed, Z. M., Goffart, L. & Krauzlis, R. J. A neural mechanism for microsaccade generation in the primate superior colliculus. Science 323(5916), 940–943. https://doi.org/10.1126/science.1166112 (2009).

Article ADS CAS PubMed PubMed Central Google Scholar

Hafed, Z. M. & Krauzlis, R. J. Similarity of superior colliculus involvement in microsaccade and saccade generation. J. Neurophysiol. 107(7), 1904–1916. https://doi.org/10.1152/jn.01125.2011 (2012).

Article PubMed PubMed Central Google Scholar

Bonneh, Y. S., Adini, Y. & Polat, U. Contrast sensitivity revealed by spontaneous eyeblinks: Evidence for a common mechanism of oculomotor inhibition. J. Vis. 16(7), 1 (2016).

Article PubMed Google Scholar

Bonneh, Y. S., Polat, U. & Adini, Y. The buildup of temporal anticipation revealed by microsaccades and eye-blinks. J. Vis. 16(12), 935. https://doi.org/10.1167/16.12.935 (2016).

Article Google Scholar

Reingold, E. M. & Stampe, D. M. Saccadic inhibition in complex visual tasks. Curr. Oculomot. Res. https://doi.org/10.1007/978-1-4757-3054-8_35 (1999).

Article Google Scholar

Reingold, E. M. & Stampe, D. M. Saccadic inhibition in voluntary and reflexive saccades. J. Cogn. Neurosci. 14(3), 371–388. https://doi.org/10.1162/089892902317361903 (2002).

Article PubMed Google Scholar

White, A. L. & Rolfs, M. Oculomotor inhibition covaries with conscious detection. J. Neurophysiol. 116(3), 1507–1521. https://doi.org/10.1152/jn.00268.2016 (2016).

Article PubMed PubMed Central Google Scholar

Ziv, I. & Bonneh, Y. S. Oculomotor inhibition during smooth pursuit and its dependence on contrast sensitivity. J. Vis. 21(2), 1–20. https://doi.org/10.1167/jov.21.2.12 (2021).

Article Google Scholar

Bonneh, Y. S., Adini, Y. & Polat, U. Contrast sensitivity revealed by microsaccades. J. Vis. 15(9), 11. https://doi.org/10.1167/15.9.11 (2015).

Article PubMed Google Scholar

Valsecchi, M., Betta, E. & Turatto, M. Visual oddballs induce prolonged microsaccadic inhibition. Exp. Brain Res. 177(2), 196–208. https://doi.org/10.1007/s00221-006-0665-6 (2007).

Article PubMed Google Scholar

Kadosh, O. & Bonneh, Y. S. Fixation - related saccadic inhibition in free viewing in response to stimulus saliency. Sci. Rep. https://doi.org/10.1038/s41598-022-10605-1 (2022).

Article PubMed PubMed Central Google Scholar

Haxby, J. V., Hoffman, E. A. & Gobbini, M. I. The distributed human neural system for face perception. Trends Cogn. Sci. 4(6), 223–233 (2000).

Article CAS PubMed Google Scholar

Gobbini, M. I. & Haxby, J. V. Neural systems for recognition of familiar faces. Neuropsychologia 45(1), 32–41. https://doi.org/10.1016/j.neuropsychologia.2006.04.015 (2007).

Article PubMed Google Scholar

Rosenfeld, J. P. P300 in detecting concealed information and deception: A review. Psychophysiology 57(7), 1–12. https://doi.org/10.1111/psyp.13362 (2020).

Article Google Scholar

Rosenzweig, G. & Bonneh, Y. S. Familiarity revealed by involuntary eye movements on the fringe of awareness. Sci. Rep. 9(1), 1–12. https://doi.org/10.1038/s41598-019-39889-6 (2019).

Article CAS Google Scholar

Rosenzweig, G. & Bonneh, Y. S. Concealed information revealed by involuntary eye movements on the fringe of awareness in a mock terror experiment. Sci. Rep. 10(1), 1–15. https://doi.org/10.1038/s41598-020-71487-9 (2020).

Article ADS CAS Google Scholar

Schwedes, C. & Wentura, D. The relevance of the first two eye fixations for recognition memory processes. Memory https://doi.org/10.1080/09658211.2019.1567789 (2019).

Article PubMed Google Scholar

Millen, A. & Hancock, P. Eye see through you! Eye tracking unmasks concealed face recognition despite countermeasures. Cogn. Res. https://doi.org/10.1186/s41235-019-0169-0 (2019).

Article Google Scholar

Peth, J., Kim, J. & Gamer, M. Fixations and eye-blinks allow for detecting concealed crime related memories. Int. J. Psychophysiol. https://doi.org/10.1016/j.ijpsycho.2013.03.003 (2013).

Article PubMed Google Scholar

Nahari, T., Lancry-Dayan, O. C., Ben-Shakhar, G. & Pertzov, Y. Detecting concealed familiarity using eye movements: The role of task demands. Cogn. Res. https://doi.org/10.1186/s41235-019-0162-7 (2019).

Article Google Scholar

Lancry-Dayan, O. C., Nahari, T., Ben-Shakhar, G. & Pertzov, Y. Do you know him? Gaze dynamics toward familiar faces on a concealed information test. J. Appl. Res. Mem. Cogn. 7(2), 291–302. https://doi.org/10.1016/J.JARMAC.2018.01.011 (2018).

Article Google Scholar

Bentin, S. & Deouell, L. Y. Structural encoding and identification in face processing: ERP evidence for separate mechanisms. Cogn. Neuropsychol. 17(1–3), 35–55. https://doi.org/10.1080/026432900380472 (2000).

Article CAS PubMed Google Scholar

Gosling, A. & Eimer, M. An event-related brain potential study of explicit face recognition. Neuropsychologia 49(9), 2736–2745. https://doi.org/10.1016/j.neuropsychologia.2011.05.025 (2011).

Article PubMed Google Scholar

Huang, W. et al. Revisiting the earliest electrophysiological correlate of familiar face recognition. Int. J. Psychophysiol. 120(January), 42–53. https://doi.org/10.1016/j.ijpsycho.2017.07.001 (2017).

Article PubMed Google Scholar

Sagiv, N. & Bentin, S. Structural encoding of human and schematic faces: Holistic and part-based processes. J. Cogn. Neurosci. 13(7), 937–951. https://doi.org/10.1162/089892901753165854 (2001).

Article CAS PubMed Google Scholar

Anaki, D., Zion-Golumbic, E. & Bentin, S. Electrophysiological neural mechanisms for detection, configural analysis and recognition of faces. Neuroimage 37(4), 1407–1416. https://doi.org/10.1016/j.neuroimage.2007.05.054 (2007).

Article PubMed Google Scholar

Eimer, M. Event-related brain potentials distinguish processing stages involved in face perception and recognition. Clin. Neurophysiol. 111(4), 694–705. https://doi.org/10.1016/S1388-2457(99)00285-0 (2000).

Article CAS PubMed Google Scholar

Henson, R. N. et al. Electrophysiological and haemodynamic correlates of face perception, recognition and priming. Cereb. Cortex 13(7), 793–805. https://doi.org/10.1093/cercor/13.7.793 (2003).

Article CAS PubMed Google Scholar

Barragan-Jason, G., Cauchoix, M. & Barbeau, E. J. The neural speed of familiar face recognition. Neuropsychologia 75, 390–401. https://doi.org/10.1016/j.neuropsychologia.2015.06.017 (2015).

Article CAS PubMed Google Scholar

Caharel, S., Courtay, N., Bernard, C., Lalonde, R. & Rebaï, M. Familiarity and emotional expression influence an early stage of face processing: An electrophysiological study. Brain Cogn. 59(1), 96–100. https://doi.org/10.1016/j.bandc.2005.05.005 (2005).

Article PubMed Google Scholar

Caharel, S., Fiori, N., Bernard, C., Lalonde, R. & Rebaï, M. The effects of inversion and eye displacements of familiar and unknown faces on early and late-stage ERPs. Int. J. Psychophysiol. 62(1), 141–151. https://doi.org/10.1016/j.ijpsycho.2006.03.002 (2006).

Article PubMed Google Scholar

Caharel, S. et al. ERPs associated with familiarity and degree of familiarity during face recognition. Int. J. Neurosci. 112(12), 1499–1512. https://doi.org/10.1080/00207450290158368 (2002).

Article PubMed Google Scholar

Keyes, H., Brady, N., Reilly, R. B. & Foxe, J. J. My face or yours? Event-related potential correlates of self-face processing. Brain Cogn. 72(2), 244–254. https://doi.org/10.1016/j.bandc.2009.09.006 (2010).

Article PubMed Google Scholar

Jemel, B., Pisani, M., Calabria, M., Crommelinck, M. & Bruyer, R. Is the N170 for faces cognitively penetrable? Evidence from repetition priming of Mooney faces of familiar and unfamiliar persons. Cogn. Brain Res. 17(2), 431–446. https://doi.org/10.1016/S0926-6410(03)00145-9 (2003).

Article Google Scholar

Marzi, T. & Viggiano, M. P. Interplay between familiarity and orientation in face processing: An ERP study. Int. J. Psychophysiol. 65(3), 182–192. https://doi.org/10.1016/j.ijpsycho.2007.04.003 (2007).

Article PubMed Google Scholar

Itier, R. J. & Taylor, M. J. Inversion and contrast polarity reversal affect both encoding and recognition processes of unfamiliar faces: A repetition study using ERPs. Neuroimage 15(2), 353–372. https://doi.org/10.1006/nimg.2001.0982 (2002).

Article PubMed Google Scholar

Nakashima, T. et al. Early ERP components differentially extract facial features: Evidence for spatial frequency-and-contrast detectors. Neurosci. Res. 62(4), 225–235. https://doi.org/10.1016/j.neures.2008.08.009 (2008).

Article PubMed Google Scholar

Engbert, R. & Kliegl, R. Microsaccades uncover the orientation of covert attention. Vis. Res. 43(9), 1035–1045. https://doi.org/10.1016/S0042-6989(03)00084-1 (2003).

Article PubMed Google Scholar

Yablonski, M., Polat, U., Bonneh, Y. S. & Ben-Shachar, M. Microsaccades are sensitive to word structure: A novel approach to study language processing. Sci. Rep. 7(1), 3999. https://doi.org/10.1038/s41598-017-04391-4 (2017).

Article ADS CAS PubMed PubMed Central Google Scholar

Hristo Zhivomirov. Bimodality coefficient calculation with matlab. MATLAB central file exchange. https://www.mathworks.com/matlabcentral/fileexchange/84933-bimodality-coefficient-calculation-with-matlab (Accessed 7 November 2021).

Pfister, R., Schwarz, K. A., Janczyk, M., Dale, R. & Freeman, J. B. Good things peak in pairs: A note on the bimodality coefficient. Front. Psychol. 4(1), 83–97. https://doi.org/10.3389/fpsyg.2013.00700 (2013).

Article Google Scholar

Morey, R. D. Confidence intervals from normalized data: A correction to Cousineau (2005). Tutor. Quant. Methods Psychol. 4(2), 61–64. https://doi.org/10.20982/tqmp.04.2.p061 (2008).

Article Google Scholar

Hohenstein, S., Matuschek, H. & Kliegl, R. Linked linear mixed models: A joint analysis of fixation locations and fixation durations in natural reading. Psychon. Bull. Rev. 24(3), 637–651. https://doi.org/10.3758/s13423-016-1138-y (2017).

Article PubMed Google Scholar

Dimigen, O., Valsecchi, M., Sommer, W. & Kliegl, R. Human microsaccade-related visual brain responses. J. Neurosci. 29(39), 12321–12331. https://doi.org/10.1523/JNEUROSCI.0911-09.2009 (2009).

Article CAS PubMed PubMed Central Google Scholar

Ries, A. J., Slayback, D. & Touryan, J. The fixation-related lambda response: Effects of saccade magnitude, spatial frequency and ocular artifact removal. Int. J. Psychophysiol. 134(August), 1–8. https://doi.org/10.1016/j.ijpsycho.2018.09.004 (2018).

Article PubMed Google Scholar

Corbetta, M. et al. A common network of functional areas for attention and eye movements. Neuron 21(4), 761–773. https://doi.org/10.1016/S0896-6273(00)80593-0 (1998).

Article CAS PubMed Google Scholar

Hafed, Z. M. & Clark, J. J. Microsaccades as an overt measure of covert attention shifts. Vis. Res. 42(22), 2533–2545. https://doi.org/10.1016/S0042-6989(02)00263-8 (2002).

Article PubMed Google Scholar

Abeles, D., Amit, R., Tal-Perry, N., Carrasco, M. & Yuval-Greenberg, S. Oculomotor inhibition precedes temporally expected auditory targets. Nat. Commun. 11(1), 1–12. https://doi.org/10.1038/s41467-020-17158-9 (2020).

Article CAS Google Scholar

Widmann, A., Engbert, R. & Schroger, E. Microsaccadic responses indicate fast categorization of sounds: A novel approach to study auditory cognition. J. Neurosci. 34(33), 11152–11158. https://doi.org/10.1523/JNEUROSCI.1568-14.2014 (2014).

Article CAS PubMed PubMed Central Google Scholar

Behrmann, M. & Plaut, D. C. Bilateral hemispheric processing of words and faces: Evidence from word impairments in prosopagnosia and face impairments in pure alexia. Cereb. Cortex 24(4), 1102–1118. https://doi.org/10.1093/cercor/bhs390 (2014).

Article PubMed Google Scholar

Yovel, G., Tambini, A. & Brandman, T. The asymmetry of the fusiform face area is a stable individual characteristic that underlies the left-visual-field superiority for faces. Neuropsychologia 46(13), 3061–3068. https://doi.org/10.1016/j.neuropsychologia.2008.06.017 (2008).

Article PubMed Google Scholar

Kadosh, O. & Bonneh, Y. S. Involuntary oculomotor inhibition markers of saliency and deviance in response to auditory sequences. J. Vis. 22(5), 8. https://doi.org/10.1167/jov.22.5.8 (2022).

Article PubMed PubMed Central Google Scholar

Valsecchi, M. & Turatto, M. Microsaccadic responses in a bimodal oddball task. Psychol. Res. 73(1), 23–33. https://doi.org/10.1007/s00426-008-0142-x (2009).

Article PubMed Google Scholar

Bonneh, Y. S. et al. Microsaccade latency uncovers stimulus predictability: Faster and longer inhibition for unpredicted stimuli. J. Vis. 13(9), 1342–1342. https://doi.org/10.1167/13.9.1342 (2013).

Article Google Scholar

Duncan-Johnson, C. C. & Donchin, E. On quantifying surprise: The variation of event-related potentials with subjective probability. Psychophysiology 14(5), 456–467. https://doi.org/10.1111/j.1469-8986.1977.tb01312.x (1977).

Article CAS PubMed Google Scholar

Jääskeläinen, I. P. et al. Human posterior auditory cortex gates novel sounds to consciousness. Proc. Natl. Acad. Sci. U.S.A. 101(17), 6809–6814. https://doi.org/10.1073/pnas.0303760101 (2004).

Article ADS PubMed PubMed Central Google Scholar

Näätänen, R., Gaillard, A. W. K. & Mäntysalo, S. Early selective-attention effect on evoked potential reinterpreted. Acta Physiol. 42(4), 313–329. https://doi.org/10.1016/0001-6918(78)90006-9 (1978).

Article Google Scholar

Stefanics, G., Stefanics, G., Kremláček, J. & Czigler, I. Visual mismatch negativity: A predictive coding view. Front. Hum. Neurosci. 8(September), 1–19. https://doi.org/10.3389/fnhum.2014.00666 (2014).

Article Google Scholar

Czigler, I., Balázs, L. & Pató, L. G. Visual change detection: Event-related potentials are dependent on stimulus location in humans. Neurosci. Lett. 364(3), 149–153. https://doi.org/10.1016/j.neulet.2004.04.048 (2004).

Article CAS PubMed Google Scholar

Cleary, K. M., Donkers, F. C. L., Evans, A. M. & Belger, A. Investigating developmental changes in sensory processing: Visual mismatch response in healthy children. Front. Hum. Neurosci. 7(December), 1–13. https://doi.org/10.3389/fnhum.2013.00922 (2013).

Article Google Scholar

Zhang, Y. et al. Perceptual expertise impacts preattentive processing of visual simple feature: A visual mismatch negativity study. NeuroReport 29(5), 341–346. https://doi.org/10.1097/WNR.0000000000000947 (2018).

Article PubMed Google Scholar

Astikainen, P., Cong, F., Ristaniemi, T. & Hietanen, J. K. Event-related potentials to unattended changes in facial expressions: Detection of regularity violations or encoding of emotions?. Front. Hum. Neurosci. 7(September), 1–10. https://doi.org/10.3389/fnhum.2013.00557 (2013).

Article Google Scholar

Gayle, L. C., Gal, D. E. & Kieffaber, P. D. Measuring affective reactivity in individuals with autism spectrum personality traits using the visual mismatch negativity event-related brain potential. Front. Hum. Neurosci. 6(December), 1–7. https://doi.org/10.3389/fnhum.2012.00334 (2012).

Article Google Scholar

Kreegipuu, K. et al. vMMN for schematic faces: Automatic detection of change in emotional expression. Front. Hum. Neurosci. 7(October), 1–11. https://doi.org/10.3389/fnhum.2013.00714 (2013).

Article Google Scholar

Walther, C., Schweinberger, S. R., Kaiser, D. & Kovács, G. Neural correlates of priming and adaptation in familiar face perception. Cortex 49(7), 1963–1977. https://doi.org/10.1016/j.cortex.2012.08.012 (2013).

Article PubMed Google Scholar

Weibert, K. et al. An image-invariant neural response to familiar faces in the human medial temporal lobe. Cortex 84, 34–42. https://doi.org/10.1016/j.cortex.2016.08.014 (2016).

Article PubMed Google Scholar

Henson, R., Shallice, T. & Dolan, R. Neuroimaging evidence for dissociable forms of repetition priming. Science 287(5456), 1269–1272. https://doi.org/10.1126/science.287.5456.1269 (2000).

Article ADS CAS PubMed Google Scholar

Kloth, N., Schweinberger, S. R. & Kovács, G. Neural correlates of generic versus gender-specific face adaptation. J. Cogn. Neurosci. 22(10), 2345–2356. https://doi.org/10.1162/jocn.2009.21329 (2010).

Article PubMed Google Scholar

Amihai, I., Deouell, L. Y. & Bentin, S. Neural adaptation is related to face repetition irrespective of identity: A reappraisal of the N170 effect. Exp. Brain Res. 209(2), 193–204. https://doi.org/10.1007/s00221-011-2546-x (2011).

Article PubMed PubMed Central Google Scholar

Hills, P. J., Elward, R. L. & Lewis, M. B. Cross-modal face identity aftereffects and their relation to priming. J. Exp. Psychol. Hum. Percept. Perform. 36(4), 876–891. https://doi.org/10.1037/a0018731 (2010).

Article PubMed Google Scholar

Maurer, U., Rossion, B. & McCandliss, B. D. Category specificity in early perception: Face and word N170 responses differ in both lateralization and habituation properties. Front. Hum. Neurosci. 2(December), 1–7. https://doi.org/10.3389/neuro.09.018.2008 (2008).

Article Google Scholar

Ko, H. K., Poletti, M. & Rucci, M. Microsaccades precisely relocate gaze in a high visual acuity task. Nat. Neurosci. 13(12), 1549–1554. https://doi.org/10.1038/nn.2663 (2010).

Article CAS PubMed PubMed Central Google Scholar

Badde, S., Myers, C. F., Yuval-Greenberg, S. & Carrasco, M. Oculomotor freezing reflects tactile temporal expectation and aids tactile perception. Nat. Commun. 11(1), 1–9. https://doi.org/10.1038/s41467-020-17160-1 (2020).

Article CAS Google Scholar

Martinez-Conde, S., Otero-Millan, J. & Macknik, S. L. The impact of microsaccades on vision: Towards a unified theory of saccadic function. Nat. Rev. Neurosci. 14(2), 83–96. https://doi.org/10.1038/nrn3405 (2013).

Article CAS PubMed Google Scholar

Rolfs, M. Microsaccades: Small steps on a long way. Vis. Res. 49(20), 2415–2441. https://doi.org/10.1016/j.visres.2009.08.010 (2009).

Article PubMed Google Scholar

Bellet, M. E., Bellet, J., Nienborg, H., Hafed, Z. M. & Berens, P. Human-level saccade detection performance using deep neural networks. J. Neurophysiol. 121(2), 646–661. https://doi.org/10.1152/jn.00601.2018 (2019).

Article PubMed Google Scholar

Shelchkova, N., Tang, C. & Poletti, M. Task-driven visual exploration at the foveal scale. Proc. Natl. Acad. Sci. U.S.A. 116(12), 5811–5818. https://doi.org/10.1073/pnas.1812222116 (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

Download references

School of Optometry and Vision Science, Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel

Oren Kadosh & Yoram Bonneh

The Leslie and Susan Gonda Multidisciplinary Brain Research Center, Bar-Ilan University, Ramat-Gan, Israel

Yoram Bonneh

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

O.K. and Y.S.B. designed the experiments. O.K. collected the data. O.K. and Y.S.B. developed the software used for running the experiments and the data analysis. O.K. analyzed the data and wrote the manuscript, Y.S.B. reviewed it.

Correspondence to Yoram Bonneh.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Kadosh, O., Bonneh, Y. Face familiarity revealed by fixational eye movements and fixation-related potentials in free viewing. Sci Rep 12, 20178 (2022). https://doi.org/10.1038/s41598-022-24603-w

Download citation

Received: 29 May 2022

Accepted: 17 November 2022

Published: 23 November 2022

DOI: https://doi.org/10.1038/s41598-022-24603-w

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.