Search

Children’s vision screening children commonly uses optotype-based visual acuity or instrument-based methods measuring amblyogenic risk factors (ARFs). To compare the performance of the Spot Vision Screener (SVS) (PediaVision, Welch Allyn, NY) and a nurse-administered visual acuity screen (NVAS) in identifying ARFs and decreased visual acuity. A prospective, cross-sectional population-based study of preschool children in South-East Queensland, Australia. Eligible participants had both forms of screening by trained community nurses. All children with an abnormal result by either method as well as a cohort of randomly selected children who passed both assessments were assessed at a tertiary paediatric ophthalmology clinic. Over a 10 month period, 2237 children (mean age; 64.4 ± 4.0 months) were screened from 38 schools. 6.4% of children failed SVS and 8.3% failed NVAS (with 3.8% overlap, failing both). The positive predictive value (PPV) in identifying either ARFs and/or reduced VA for the SVS and NVAS was 70.4% (95% Confidence Interval (CI): 61.6%–78.2%) and 60.5% (95% CI: 52.6%–67.9%) respectively. Highest PPV to detect either ARFs and/or reduced VA was achieved by a ‘hybrid’ method by combining failed NVAS and failed SVS: 91.0% (95% CI: 82.4 to 96.3) but this would risk children with sight impairment being missed in the community. To our knowledge, this is the first population-based study providing detailed comparative measures of diagnostic accuracy for NVAS and SVS in preschool children. One in ten preschool children failed one or both screens. A number of children who required ophthalmic intervention were missed if only one screening method was utilized.

Multistable visual phenomena, wherein unchanging sensory input elicits in an observer, perceptual fluctuations, have been instrumental in unravelling the neural correlates of conscious perception. Such paradigms, when combined with single unit recordings in macaques trained to report their perception, have allowed neurophysiologists to elucidate, if the cells in various regions of the brain are correlated with subjective experience or respond to the invariant retinal input. Results obtained from such an approach has so far revealed that the proportion of feature selective cells which fire in concordance with perception, increase as one progresses in the ventral visual pathway, with this fraction being up to 90% in the temporal lobe. The next station in the ventral stream of vision is the lateral prefrontal cortex (LPFC), which has reciprocal connectivity with the inferotemporal cortex and displays responses which are selective for complex visual stimuli. However, it’s not clear if this feature selective neural activity is just the result of sensory input or is related to subjective perception. Utilizing the task of binocular flash suppression (BFS), a psychophysical paradigm capable of dissociating perception from the retinal message, we probed the neural responses in the LPFC. The results revealed a robust perceptual modulation of both the spiking activity as well as high frequency gamma oscillations in this region of the brain. Even though single unit activity is robustly modulated according to perceptual content, a measure of effective functional connectivity between pairs of neurons, such as correlated variability could be revealing of interactions among neuronal populations during visual ambiguity. We therefore computed the spike count correlations across pairs of simultaneously recorded neurons during subjective visual perception. Interestingly, such interneuronal correlations among single units which preferred the same stimulus were close to zero during incongruent visual input, thus reflecting a modulation of the correlation structure during visual perception. Simulations with biophysically realistic networks suggested that the source of decorrelation was an active suppression of input fluctuations. This suggests that such a decorrelated state might be critical for representation of conscious content during visual conflict. These results together provide credence to the ‘frontal lobe hypothesis’ proposed by Crick and Koch, which suggested that the planning stages of the brain must have explicit access to the conscious visual percept so as to direct motor output. Such access is essential, if the LPFC needs to carry out one of its major function which is of cognitive control. Interestingly, when a control related signal, namely the modulation pattern of the beta band oscillations in the LPFC was analyzed, its modulation pattern was unchanged not only across monocular and incongruent visual stimulation but also during perceptual dominance and suppression. This suggests that a signal which is related to control processes is unaffected by local conscious or unconscious neural processing. Lastly, we observed an enormous diversity among the patterns of single unit activity recorded in the LPFC and the neurons which displayed visual preference were just a minority. In order to elucidate, if there were any other patterns of activity which were related to the task, we clustered the neuronal responses using a non-negative matrix factorization (NNMF) method. This revealed five sequential dominant response patterns (or components) whose peaks were temporally distributed across various phases of the trial. A majority of the units with firing profiles similar to the patterns obtained, maintained their responses across monocular or incongruent stimulation suggesting that visual conflict did not affect their spiking modulation. Interestingly, an assessment of the effective functional connectivity across the pairs of neurons belonging to different temporally distributed components revealed that such correlated variability was maximum among units which were temporally coincident. However, we observed successive decorrelation as the pairs of units were chosen from temporally separated populations. This suggests a computational principle mediating a representation of sequential patterns of activity in the LPFC. Together, the results presented in this thesis suggest a role for the LPFC in representation of conscious content. At the same time, we find that such a role of this region is coexistent with other major functions typically attributed to this area, such as cognitive control or temporal encoding of task events through sequential neural activity.

The neural correlates of conscious visual perception are commonly studied in paradigms of perceptual multistability that/nallow multiple perceptual interpretations during unchanged sensory stimulation. What is the source of this multistability in/nthe content of perception? From a theoretical perspective, a fine balance between deterministic and stochastic forces has/nbeen suggested to underlie the spontaneous, intrinsically driven perceptual transitions observed during multistable/nperception. Deterministic forces are represented by adaptation of feature-selective neuronal populations encoding the/ncompeting percepts while stochastic forces are modeled as noise-driven processes. Here, we used a unified neuronal/ncompetition model to study the dynamics of adaptation and noise processes in binocular flash suppression (BFS), a form of/nexternally induced perceptual suppression, and compare it with the dynamics of intrinsically driven alternations in binocular/nrivalry (BR). For the first time, we use electrophysiological, biologically relevant data to constrain a model of perceptual/nrivalry. Specifically, we show that the mean population discharge pattern of a perceptually modulated neuronal population/ndetected in electrophysiological recordings in the lateral prefrontal cortex (LPFC) during BFS, constrains the dynamical/nrange of externally induced perceptual transitions to a region around the bifurcation separating a noise-driven attractor/nregime from an adaptation-driven oscillatory regime. Most interestingly, the dynamical range of intrinsically driven/nperceptual transitions during BR is located in the noise-driven attractor regime, where it overlaps with BFS. Our results/nsuggest that the neurodynamical mechanisms of externally induced and spontaneously generated perceptual alternations/noverlap in a narrow, noise-driven region just before a bifurcation where the system becomes adaptation-driven.

Privacy, its violations and techniques to bypass privacy violation have grabbed the centre-stage of both academia and industry in recent months. Corporations worldwide have become conscious of the implications of privacy violation and its impact on them and to other stakeholders. Moreover, nations across the world are coming out with privacy protecting legislations to prevent data privacy violations. Such legislations however expose organizations to the issues of intentional or unintentional violation of privacy data. A violation by either malicious external hackers or by internal employees can expose the organizations to costly litigations. In this paper, we propose PRIVDAM; a data mining based intelligent architecture of a Privacy Violation Detection and Monitoring system whose purpose is to detect possible privacy violations and to prevent them in the future. Experimental evaluations show that our approach is scalable and robust and that it can detect privacy violations or chances of violations quite accurately. Please contact the author for full text at rajanish@iima.ac.in