Oral Presentation, 22nd International Conference of the Association of Psychology & Psychiatry for Adults & Children (A.P.P.A.C.): Recent Advances in Neuropsychiatric, Psychological and Social Sciences in Psychological Research, 16th – 19th May 2017, Athens, Greece.

1. Interfering with Selective Attention:
Effects of Backward Masking on Saccadic Reaction
Time and Pupil Size.
D. VLASTOS1, M. KYRITSIS1,2, A. PAPAIOANNOU-SPIROULIA1, V.-A. VARELA1,
DEPARTMENT OF PSYCHOLOGY, Ψ RESEARCH CENTER, CITY UNITY COLLEGE1
DEPARTMENT OF COMPUTER INFORMATION SCIENCE, HIGHER COLLEGES OF TECHNOLOGY2

2. Background Information I
 Brain regions related to higher-level processing can be influenced by lower-level operations.
 Both top-down (goal-driven) and predictive signals contribute to bottom-up (stimulus-driven)
information processing.
 The perception of facial expressions may involve both unconscious early visual processing
as well as higher-level operations, in order to make individuals aware of a stimulus and of
its influence.
 Affective processes seem to demonstrate & form perception on sensory pathways whilst, in
terms of vision, possible influences on oculomotor processes may force eyes to
emotionally endogenous & exogenous cues.
(Atkinson & Adolphs, 2005; Domininguez-Borras & Vuilleumier, 2013; Friston, 2012)

3. Background Information II
 Existing studies on selective attention have focused independently on these cues, despite the
fact that attention and awareness are driven by multiple attention gain control systems that
may act simultaneously on visual pathways and also control the acquisition of neural
responses (Pourtois et al., 2013).
 Individuals process emotion-relevant information earlier than neutral, whereas threat-relevant
cues attract attention rapidly in order to prioritize further and accurate processing for the
manifestation of a potential response, even when cognitive resources are sufficient (Öhman et
al., 2001).

4. Eye-Tracking Measures I
Eye-tracking studies have provided variable results when investigating to what extend
emotionally meaningful information:
 impacts on the number of fixations,
 how long such stimuli must stay on screen before they elicit a response,
 what types of stimuli attract the very first saccades and,
 whether individual differences exist in these processes.
 Visual perception can be influenced by affective information both behaviorally and
neurally, whilst emotion-relevant cues are processed more efficiently than neutral; supporting
a negativity bias account, which demonstrates modulation of even early sensory processes
by unpleasant (negative) visual stimuli.

5. Eye-Tracking Measures II
There is evidence to support that emotionally significant information:
 has a greater effect on the number of fixations
 attracts more the first saccades, than neutral information (Humphrey et al., 2012; Niu et al.,
2012).
 Whereas, individuals detect fearful faces faster than other types of stimuli (Domininguez-
Borras & Vuilleumier, 2013; Pourtois et al., 2013; Todd et al., 2012; Yang et al., 2007).
 McSorley & Van Reekum (2013) in an eye-tracking study, used emotional distractor images in
a selective attention task. They found saccadic changes each time a distractor was
represented extrafoveally within the task (compared to neutral images) and concluded that
the process of selection can be biased by motivation.

6. Literature Contradictions
 Pessoa et al. (2002) in an fMRI study supported that the presentation of unattended emotional
stimuli, such as fearful faces, does not elicit a response in the AMG.
 Bishop et al. (2004) hypothesized that the AMG responses to fearful faces vary depending
on the participants’ anxiety levels; Highly anxious participants’ responses were unaffected
by attention, whereas low anxiety participants were more likely to respond to attended than
unattended fearful faces.
 Öhman et al. (2007) in a study where GSR’s were recorded, showed that unattended masked
fearful faces cause a physiological response during a selective attention task; A finding
which later confirmed through PET.
 Szczepanowski et al. (2007), hypothesized that the difference in findings may be due to
participants being able to detect masked fearful faces as early as 17ms.

7. Rationale of the Present Study
 Whether unattended masked fearful faces can activate emotion-relevant areas such as the
Amygdala (AMG) and areas related to cognition, such as the frontoparietal networks of
attention, has been debated for some time.
 One seemingly important factor is exposure time, with studies reporting that some
participants could detect fearful faces as early as 17ms.
 Existing studies on selective attention have focused independently either on exogenous or
endogenous cues due to the difficulties involved in the quantification and measurement of
affective processing. However, different sources of modulatory control act at the same
time on VP’s and also control the acquisition of neural responses (Pourtois et al., 2013)
 Additionally, little is known about the impact of potentially accessible but unattended fearful
stimuli on the process of selective attention in particular (i.e., interference with selection and
recognition).

8. Methods I
 Participants:
The guidelines by Holmqvist et al. (2011) were used to determine the required number of
participants to increase precision of the data given the sampling rate of our tracking equipment.
Thus, 47 student volunteers (13 males and 34 females) aged 19–48 (μ = 24.7, σ = 5.8)
participated in the study. All subjects had normal or corrected-to- normal vision, were in good
health with no past history of psychiatric and neurological disorders and gave their informed
consent prior to inclusion. The experiment was approved by the Cardiff Metropolitan
University Research and Ethics Committee and was in accordance with the Code of Ethics
of the World Medical Association (Declaration of Helsinki). Informed consent was obtained
from all individual participants included in the study.
 Design:
Interference was introduced through extrafoveal backward masked fearful-to-neutral or
neutral-to-neutral faces that appeared on screen with a temporal variability of 17ms, 33ms, or
no exposure to the distractors.

9. Methods II
 Materials and Stimuli:
The experiment was developed in Java in order to
work with the eye tracker’s SDK.
An Eye Tribe tracker (eyetribe.com, Denmark) was
used to record saccadic behavior with a sampling
rate of 6O Hz.
10 fearful and 9 neutral images were used from the
Warsaw Set of Emotional Facial Expression
Pictures (Olszanowski et al., 2015) as task-
irrelevant stimuli, as well as 8 control target
images depicting geometric shapes with no
possible emotional content.
Figure. 1.0:
Possible
Task-Irrelevant
Stimuli.
Figure. 1.1:
List of
Target Stimuli.

10. Methods III
 Experimental Procedure - Task Schematic:
6 BLOCKS X 60
TRIALS PB = 360 TRIALS

11. Method of Analysis
In our repeated measures design experiment we measured:
 3 continuous dependent variables (DVs):
 Total saccadic time from fixation to peripheral target/non-target (RT);
 Selection errors (Errors);
 and pupil size (Pupil).
 2 independent categorical variables (IVs):
 Exposure time of the distractor prior to masking (no interference, 17ms, or 33ms),
 The type of facial stimuli presentation (fearful-to-neutral or neutral-to-neutral face).
 We used omnibus tests to look for main effects and interactions between the variables, and
post-hoc tests to identify which groups showed significant differences.

12. Results I
Our analysis showed that:
1. REGARDING THE REACTION TIME (RT):
 Participants had an increased delayed reaction in
selecting the correct match in trials that had a
masked fearful distractor than in trials with a
masked neutral distractor, a finding which is in
support of the general consensus that masked
fearful faces can modulate the process of cognition,
(e.g., see Öhman, Esteves, & Soares, 1995;
Lundqvist, Juth, & Öhman, 2013).
 However, surprisingly, our results indicate that the
effect of masked fearful faces on reaction time was
stronger at ~17ms than ~33ms. Figure 1. Interaction plot for RT showing an opposite effect at one frame
exposures between fearful-to- neutral and neutral-to-neutral trials. Note,
the y-axis shows the reciprocal of RT. Therefore, the actual values are
flipped (Error bars are 95% CI) (Vlastos, et al., 2016/2017).

13. Results II
2. REGARDING PUPIL DILATION:
Our study provides further evidence to support that
masked fearful faces can affect pupil dilation, an effect
that can be reliably measured using eye- tracking.
 In trials with a masked fearful face distractor, pupil
constriction was reduced when compared to trials with
a masked neutral face distractor.
 Much like with RT, our results indicate that the
constriction response was smaller in trials where the
masked fearful face appeared for ~17ms than in trials
where the masked fearful face appeared for~33ms.
Figure 2. Interaction plot for pupil sizes showing an opposite effect at
one frame exposures between fearful-to-neutral and neutral-to-neutral
trials. Note, eye tracker units are arbitrary and do not reflect any metric
standard (Error bars are 95% CI) (Vlastos, et al., 2016/2017).

14. Results III
3. REGARDING SELECTION ERRORS:
 Finally, we found that there was a significantly higher chance of selection errors in
trials with masked fearful faces than trials with masked neutral faces, a finding
that we anticipated.
 However, the exposure time to the masked distractor, regardless of valence, also
had a significant impact on selection errors, it was in the opposite direction than
expected.
 When not accounting for valence, short exposure times led to more errors than
longer exposure times.

15. Discussion & Future Aims
 Our results suggest that exposure to masked fearful faces in extrafoveal vision for
as low as 17ms is enough to produce a measurable physiological response and
delay in the process of selection.
 The cause for the unexpected results mentioned before, is something we aim to
investigate further in the future.
 However, we suspect that this may have more to do with the process of masking
itself (perhaps due to the movement artifacts sometimes caused when the mask is
applied).

16. Practical Applications I
 Selective attention is fundamental to almost all learning tasks, which allows learners to
focus cognitive resources on vital information (while ignoring unnecessary input), and
by so doing facilitate internal cognitive processes.
 Faces provide an exceptional opportunity to study the generic mechanisms of
unconscious processing.
 Cognitive functions related to social cognition and multisensory integration can clearly
be processed unconsciously. By contrast, the functioning of visual face-specific
mechanisms outside conscious awareness is more limited, and there is already
strong evidence against the unconscious processing of unfamiliar face identity.
Invisible faces constitute a convenient platform for researchers to address new
questions and represent a promising tool in both academic and clinical
domains.

17. Practical Applications II
Development of E-Interventions to promote E-Health:
 Several studies have reported that invisible emotional faces can be used for diagnosis
and for measuring the effectiveness of intervention with patients suffering from
depression (Sheline, Y.I. et al., 2001). Would this or similar methods be effective
for other psychiatric disorders, such as post-traumatic stress disorder (PTSD)?
 When people look at faces, they focus on characteristic features. Can similar eye-
looking patterns be found for invisible faces? In the clinical domain, will autistic
patients, who do not fixate on eyes (or fixate less compared to healthy
participants) in visible faces (Boraston & Blakemore, 2007), also not fixate on
eyes in invisible faces?

18. Practical Applications II
Development of Better Assessment and Evaluation of Forensic Methods:
 Is it possible that, using invisible faces (unconscious processing), we can reveal social
phenomena that cannot be revealed using visible faces (conscious processing)? Can
racist biases be revealed reliably by using invisible faces (Pinto et al., 2012)?
Can invisible faces be used in the future as a lie detector?

19. Practical Applications II
Development of Computational Models and Machine Interfaces to enhance the
“Cognitive Brain”:
 Machine interfaces for cognitive rehabilitation: Neurofeedback applications have
already been proven efficient for several rehabilitation applications on
ADHD. However, it is expected that ADHD patients with low hyperactivity
symptoms but high inattention symptoms will not benefit by this approach, due
to a different functional deficit underlying their symptoms; a finding confirmed
also by eye-tracking studies. Thus, a combination of these methods seems as a
promising strategy that could also be beneficial within educational settings.

20. References
Atkinson AP, Adolphs R (2005). Nonconscious emotions. Visual emotion perception. Mechanisms and
processes. In: Barrett LF, Niedenthal PM, Winkielman P (Eds.) Emotion and consciousness (pp. 150–182).
New York: The Guilford Press.
Dominguez Borras, J., & Vuilleumier, P. (2013). Affective biases in attention and perception.
Friston, K. (2012). Prediction, perception and agency. International Journal of Psychophysiology, 83(2), 248-
252.
Humphrey K, Underwood G, Lambert T (2012). Salience of the lambs: A test of the saliency map hypothesis
with pictures of emotive objects. Journal of Vision 12(1):22–22.
Lang, P. J., Davis, M., & Öhman, A. (2000). Fear and anxiety: animal models and human cognitive
psychophysiology. Journal of affective disorders, 61(3), 137-159.
McSorley, E., & van Reekum, C. M. (2013). The time course of implicit affective picture processing: An eye
movement study. Emotion, 13(4), 769.
Niu Y, Todd RM, Anderson AK (2012). Affective salience can reverse the effects of stimulus-driven salience on
eye movements in complex scenes. Frontiers in Psychology 3:1–11.

21. References
Öhman, A., Carlsson, K., Lundqvist, D., & Ingvar, M. (2007). On the unconscious subcortical origin of human fear.
Physiology & Behavior, 92(1), 180-185.
Pessoa, L., McKenna, M., Gutierrez, E., & Ungerleider, L. G. (2002). Neural processing of emotional faces requires
attention. Proceedings of the National Academy of Sciences, 99(17), 11458-11463.
Pourtois, G., Schettino, A., & Vuilleumier, P. (2013). Brain mechanisms for emotional influences on perception and
attention: what is magic and what is not. Biological psychology, 92(3), 492-512.
Szczepanowski, R., & Pessoa, L. (2007). Fear perception: can objective and subjective awareness measures be
dissociated? Journal of Vision, 7(4), 10.
Todd RM, Cunningham WA, Anderson AK, Thompson E (2012). Affect-biased attention asemotion regulation. Trends in
Cognitive Sciences 16(7):365-372.
Treisman, A., & Gelade, G., (1980). A feature integration theory of attention. Cognitive Psychology, 12, 97-136.
Vlastos, D., Kyritsis, M., Papaioannou-Spiroulia, A., Varela, V., A. (2016/2017). Interfering with Selective Attention: Effects of
Backward Masking on Saccadic Reaction Time and Pupil Size. – Under Review.
Wolfe, J.M., (2013). Approaches to Visual Search: Feature Integration Theory and Guided Search. In Nobre, A.C. &
Kastner. S (Eds.), The Oxford Handbook of Attention. New York, NY: Oxford University Press.

22. End of Presentation
Thank you for your attention!
Any Questions?
For further info, please, do not hesitate to contact me at: d.vlastos@cityu.gr