Poster group
Details of individual items:
poster
Preferential looking (PL) tests reveal sensitivity to motion direction onlyafter about 8 weeks of age. However, the ability to elicit optokineticnystagmus (OKN) indicates that mechanisms sensitive to motion direction arepresent in the infants' visual system from birth. It has been proposedthat PL responses depends on cortical motion processing while OKN can bemediate at least in part by subcortical nuclei. We have used motioncoherence thresholds to test the differential sensitivity of these twosystems in the age range 9-25 weeks.Infants viewed random dot displays containing a 'signal' of dots moving at9 deg/sec with a limited lifetime of 4 frames (80 msec), and a variablepercentage of 'noise' dots which were randomly plotted on each frame. Thesame infants were tested using two different responses. For PL, signaldots were in oscillatory motion; on each trial one side of the display wasdivided into three strips with the central strip moving in counterphase tothose above and below. For OKN, all signal dots moved uniformly leftwardsor rightwards. A 'blind' observer used the infants' eye movements to judgethe side of stimulus in PL, and the direction of motion in OKN. Thecoherence ( % signal dots) was adjusted from trial to trial by a staircaserule based on the correctness of the observer's judgment, to give anestimate of motion coherence threshold for the two types of response.Mean coherence thresholds for PL improved gradually between 8-25 weeks. Incontrast, coherence thresholds for OKN were lower at all ages and showed noage trend. We conclude that even after the emergence of directional mechanisms thatcan by used to detect and fixate motion-defined targets, optokineticresponses depend on a distinct, relatively mature directional mechanismwhich is probably subcortical. Coherence sensitivity and its age trendmay provide a method to tease apart the contributions of these twomechanisms to other responses, such as the naso-temporal direction ofmonocular OKN which has been attributed to cortical processing. Visualmotion processing is not subserved by a single system, but by multiplemechanisms which are associated with different visual tasks and which canbe dissociated in development.Supported by MRC programme grant G7908507
poster
The asymmetry of optokinetic nystagmus (OKN) in young infants is well known.For horizontal OKN, the asymmetry is only apparent with monocular viewing,and consists of a stronger response to nasalwards stimulus motion; whilevertical OKN is asymmetric with binocular viewing, and is stronger forupwards motion. The experiments described here examined whether directionalasymmetries are confined to the control of OKN, or are present when theinfant's response does not involve smooth eye movements.Infants viewed pairs of patterns displayed side-by-side on a video monitor.The patterns moved in opposite directions, but were otherwise identical, andthe experiment used preferential-looking (PL) to determine which directionof motion the infants preferred. Most of the experiments used 90% contrastrandom-dot patterns (RDPs), but some used 0.25 c/deg gratings. In both casespattern speed was 18 deg/sec. The infant subjects were aged 5-25 weeks.With horizontally-moving RDPs, and monocular viewing, infants showed asignificant preference for nasalwards motion, mirroring the asymmetry ofOKN. However, unlike the OKN asymmetry, the PL asymmetry was not present inthe youngest (5-weeks-old) infants, but first appeared at about 7 weeks,which agrees with previous results that sensitivity to direction emerges atthis age. A sub-group of the infants were also tested for stereopsis. The PLmotion asymmetry emerges before, and persists for a long time after, theonset of stereopsis; it is still present at 25 weeks.With the RDP stimulus, the vertical PL asymmetry was relatively weak; it wasonly evident as a significant inverse correlation between the preferreddirection of motion and the degree of OKN asymmetry. However, testing withlow-contrast gratings (10 & 20%) produced a large PL asymmetry - a strongpreference for downwards motion - which was in the opposite direction to theasymmetry of vertical OKN.The differences between the PL and OKN asymmetries implies that there are atleast two distinct populations of directional mechanisms: one, perhapssubcortical, is responsible for the control of OKN, while the other(cortical?) subserves more general behaviour. Both populations showdirectional asymmetries during development.
poster
There is strong evidence that primate pre-striate visual cortex isfunctionally divided into two streams of processing; the dorsal stream thattransmits information to the parietal lobe, and the ventral stream leadingto the temporal lobe. Observers thresholds in a global motion coherencetask are thought to reflect a functional limit of the extra-striate dorsalstream. An analogous measure of global processing by the ventral stream canbe obtained from subjects performance in a form coherence task (Braddick etal, 1999, Perception, 28S). This is supported by fMRI measurementssuggesting that concentric coherent form patterns activate extra-striateregions distinct from those areas differentially activated by motioncoherence. We have used the form coherence task to test young infants ability to detect a coherent form embedded in a background of randomlyoriented line segments.Infants viewed a computer-generated stimulus comprising a field of short,black line segments against a light background. On one side of the displayline segments within a circular region were oriented tangentially toconcentric circles; the remaining line segments were randomly oriented. Thestimulus was re-plotted every 800 msec with the coherent form verticallydisplaced; a field of randomly oriented line segments was present during theinter-displacement interval (133 msec). An adult viewing the stimulus seesa coherent circular form moving up and down on one side of the display.Infants looking behaviour was observed in a forced-choice preferentiallooking paradigm, in which the side containing the coherent form wasrandomly chosen from trial to trial. Three age groups were tested: 12-14,15-17 and 18-20 week olds. The results show that infants from each agegroup tested displayed significant preferential looking, in favour of theside of the display containing the coherent form. This suggests that, by atleast 12 weeks of age, the ventral pathway of the infant visual system isdeveloped to the point where it can group together spatially separateelements in the visual array on the basis that they are related locally andglobally to create a coherent form.
poster
This study examined the quantitative characteristics of infant saccadic eyemovements in infants from 5 to 26 weeks of age. The Ômain sequence' is alinear relation between saccade velocity and amplitude. The main sequencerelation is generated by the brainstem areas that control oculomotor nerves.In this study the main sequence was examined when infants made a saccadefrom a central stimulus to a peripheral stimulus.Infants ranging in age from 5 to 26 weeks (total N 209, approximately 30participants per age) had their fixation drawn to a blinking dot in thecenter of the television screen. A more complex central stimulus waspresented when fixation was on the center of the screen. After a delay, adistractor stimulus was presented in the periphery. At that time thecentral stimulus either remained on (addition trials) or was turned off(replacement trials). The delay between the onset of the central stimulusand the distractor stimulus was defined by time (i.e., 0 sec or 2 sec delay)or by heart rate changes known to be associated with attention (i.e.,sustained attention: heart rate deceleration; inattention: return of heartrate to prestimulus level). The electrooculogram was used to measuresaccadic eye movements to the peripheral stimulus. The fastest velocity ofthe saccade and the total amplitude of the saccade were quantified.The main sequence relation between saccade velocity and amplitude was foundin this study. That is, the velocity of the saccade was positivelycorrelated with the amplitude of the saccade. The beta weight describingthe linear relation between velocity and amplitude showed a decline from 5to 8 to 11 weeks of age, but did not change significantly from 11 to 14, 20,or 26 weeks of age. For the infants from 5 to 14 weeks of age, the betaweights describing the main sequence relation were not substantiallydifferent for experimental conditions in which attention was engaged andconditions in which attention was unengaged. For the 20 and 26 week oldinfants, if the infant was attending to the central stimulus the mainsequence relation for the saccade toward the peripheral stimulus wasdisrupted relative to inattentive fixation. These results suggest that thereis a developmental change in the brainstem areas controlling saccades in thefirst part of this age range (5 to 11 weeks), and changes at later ages (20and 26 weeks) in the attention-related cortical areas affecting eyemovements.
poster
To explore the possibility of hemispheric asymmetries in infant encoding of colour, 36 infants (mean age: 20.9 weeks, SD: 4.3) were familiarized with 250 msec exposures of a display consisting of two star-shaped stimuli in different colours, 8 degrees in the left or right visual field. Recognition for the two colours was then assessed in terms of degree of visual preference for a novel-coloured stimulus paired respectively with either of the familiar colours. All colours were counterbalanced. There was a marginally significant novelty preference for the right visual field (RVF) colours but a more robust familiarity preference for the left visual field (LVF) colours, implying better encoding of the colours by the left cerebral hemisphere (LHS) relative to the right (RHS) - a finding at odds with that for adult response to colour. Alternative interpretations of the results include the possibility of a developmental shift in the lateralization of response to colour or of hemispheric differences in the response to repetitive stimulation.