The world around us is filled with a multitude of shapes and colours, all mixed

and oriented in almost any imaginable fashion.

To be able to seek out and recognise objects and information of salience, the
human visual system must be able to distinguish between relevant targets and
those that do not pertain to the task at hand.
Since the development of Feature Integration Theory by Treisman and
Gelade (1980), the last thirty-odd years have seen much research on the
finding the neuroanatomical correlates of visual processing, determining their
functional contributions, and the manner in which they interact to provide an
integrated, meaningful interpretation of visual stimuli.
At present, it is known that three parts of the brain: the Right Posterior Parietal
Cortex, the Left Fontal Eye Fields and the Right Frontal Eye Fields, are
heavily involved in so-called conjunction searches (Lane, Smith, Schenk
and Ellison, 2012), where a target object must be selected from a field of
distractors on the basis of two or more distinct visual features (i.e. colour
and orientation, colour and colour, etc.).
However, a detailed, concrete understanding of how these brain regions
operate and interact is yet to be elucidated.
In this current study, the primary objectives are to review some of the most
recent findings regarding conjunction searches, particularly those by Lane et
al. (2012), extend upon their experimental methodology, and thus evaluate
whether their conclusions about the functions of the aforementioned brain
areas are justified.
The information gained will not only advance understanding of how humans
perceive and make sense of the world via visual faculties, but may have
implications for clinical management of patients with neurological damage to
vision processing pathways.
It has been long known (Howe et al., 2009) that visually searching for a
specific object amidst a potentially confusing array of other items requires
information processing by more than one neuroanatomical location.
Using transcranial magnetic stimulation (TMS), Ashbridge, Walsh and Cowey
(1997) first attested to the role of the right Posterior Parietal Cortex (rPPC) in
facilitating timely conjunction searching.
In particular, temporary inhibition of rPPC function by TMS resulted in
increased reaction times for conjunction searches.
Subsequent experiments by Muggleton, Juan, Cowey and Walsh (2003) and
Kalla, Muggleton, Juan, Cowey and Walsh (2008) also demonstrated the
contributions of the Left and Right Frontal Eye Fields (lFEF and rFEF
respectively), as use of TMS on these areas also elevated reaction times.
Furthermore, direct evidence that these brain regions are effectors of visual
processing
In all, the identities of the principal brain regions associated with conjunction
searching have been resolutely determined.
However, the precise functions of the rPPC, lFEF and rFEF, as well as how
they work in concert, remain a contentious issue.
Prior studies have alluded to the rPPC being recruited in non-primed
conditions (when participants are presented with a novel stimuli field of a

target among distractors), while its activity in primed conditions (when

participants are pre-emptively aware of the targets location, and only the
positions of distractors are altered) is noticeably diminished (Walsh et al.,
1998).
Similar deductions have been made regarding the operation of the frontal eye
fields (Kalla et al., 2008), especially the lFEF (Lane et al., 2012).
It is at this stage in research, and within this context, that the study by Lane et
al. (2012) can be properly considered.
In their investigation, Lane et al. (2012) aimed to compare and delineate the
roles of the rPPC, lFEF, and rFEF in conjunction searches.
Successive visual search task was based around a conjunction of colour and
orientation, and involved selecting a diagonal red line (the target) against a
backdrop of diagonal green lines and red lines angled oppositely (distractors).
If the target was seen, participants pressed a button, and the time from when
the stimulus was presented to when the button was depressed measured as the
reaction time.
Participants were, at times, subjected to TMS inhibition of the rPPC, lFEF, and
rFEF separately.
In addition, the target line and its location would sometimes be repeated from
one visual search to the next (the primed condition). Other times, it only
appeared once (the non-primed condition), with no prior hint as to its
location.
Over the course of the experiment, Lane et. al (2012) observed that in the nonprimed condition, there were increased reaction times no matter what brain
area was inhibited.
However, in the primed condition, applying TMS to the lFEF and rPPC did
not slow down reaction speed, while inhibiting the rFEF did so significantly.
From these results, it was concluded that the lFEF and rPPC and rFEF all
functioned to determine the orientation and location of the target.
However, when target location was already known, only rFEF activity was
seen.
Thus, Lane et. al (2012) stated that the rPPC and lFEF exclusively dealt with
spatial disambiguation (p. 16), while the rFEF not only covered this role by
itself, but also decided whether or not to recruit the rPPC and lFEF.
In favour of this assertion is BOLD fMRI data interpreted by Kristjnsson et
al. (2007), which showed decreased activity within the parietal cortex with
priming, but similar activation levels in the frontal eye fields regardless of
priming status.
Though the study by Lane et al. (2012) initially appears sound and well
supported, there are some points of contention.
All trials were conducted based on a single type of visual conjunction, namely
that of colour and orientation of the target line.
There are many other conjunction combinations, such as of colour and form,
colour and orientation, and colour and size (Wolfe, Cave and Franzel, 1989).

In addition, Walsh et al. (1998) used a similar colour and orientation

conjunction, and were quick to highlight that any changes seen may not be
seen in other conjunction searches.
Furthermore, colour and orientation conjunctions are, according to Wolf et al.
(1989) are handled with ease, even without prior learning.
As such, the possibility that the rPPC and lFEF are needed for more complex
conjunction searches, even with priming, cannot be discounted.
Another pertinent issue is that the study by Kristjnsson et al. (2007) was
centered on a entirely different mode of visual search: namely pop-out
searches, which are inherently faster and easier than conjunction searches
(Howe et al., 2009).
Thus, there is sufficient reason to question the generalizability of the claims
made by Lane et al. (2012).
Given that the experimental methods outlined by Lane et al. (2012) are, for the
most part, well founded, this current study will also make use of the same
procedures, albeit with some modifications (see Methods section).
One important difference shall be in the conjunction combination used.
Wolfe et al. (1990) established that conjunctions of the same feature (e.g. color
and color, orientation and orientation) are not searched efficiently, and thus
present more difficulty than conjunctions of different features.
As such, this experiment will use a color x color conjunction, in an effort to
elicit activity from the lFEF and the rPPC, even in primed conditions.
With reference to the methodology and findings of the aforementioned
research, this present study puts forward the following hypotheses:
(1) On the assumption that priming is efficacious, we predict faster reaction
times in the primed conditions than the non-primed condition, regardless of
whether TMS is applied.
(2) In the non-primed condition, TMS of the rPPC, the lFEF and the rFEF
should increase search times compared to when TMS is not used.
(3) In the primed condition, reactions times should indicate the continued
involvement of the rFEF, as seen in Lane et al. (2012).
(4) In the primed condition, the rPPC and the lFEF should also be involved,
with TMS of these areas increasing reaction times.

Materials and Methods

This study replicated the majority of the methods as used by Lane et al.
(2012), except for the following changes:
Participants
Twenty-five participants (10 male, 15 female), aged 18-45 years, were
selected.
All had normal colour vision (as determined by an Ishihara test)
In addition, visual acuity was normal or corrected to normal, with participants
all scoring at least 20/25 in both eyes on a Snellen eye chart.
At no time were any participants excluded from the study.

Visual Search Task

The color x color conjunction search array was designed as follows.
An 4 x 3 virtual array subtending 32o x 24o of the visual angle was
constructed.
All items were 2o x 2o squares, consisting of two different colours vertically
bisected by a black midline with width 0.2o.
The target squares were coloured red (CIE x,y coordinates = 0.62, 0.36) and
green (CIE x,y coordinates = 0.34, 0.57) on either side, while distractors
contained either blue (CIE x,y coordinates = 0.14, 0.07) and red or green and
red.
Each array contained 12 squares. In 50% of trials, the target was present, along
with 11 distractors. In the remaining trials, only distractors were shown.