An Integrative Approach of Visual Functions in ASD

Background:

Visual perception is atypical in children with ASD at many stages from oculomotor functions to cerebral processing. ASD children present more ophthalmological disturbances, faster saccades and atypical early brain responses to social and non-social stimuli. A number of studies have recently suggested that sensory processing might play a crucial role in social disturbances. However, few studies have investigated the relationship between impairements in fixation, saccades and the underlying brain responses recorded by EEG.

Objectives:

The current analysis aims at investigating whether impairments in high-level perception are the consequence of low-level sensory dysfunctions and to quantify their respective influence.

Methods:

After ophthalmological examination, thirteen children with ASD with normal vision completed several tasks conceived to investigate visual perception at different stages. Children were aged from 42 to 130 months (mean age ± standard deviation: 78.4 ± 29.2 months). Because several children presented an intellectual delay, all tasks were passive, and no explicit instruction was given to the participants. The tasks included the recording of visual evoked potentials to pattern reversal stimulation and to emotional faces. Moreover, two eye tracking tasks were designed to investigate i) fixation and ii) saccade metrics. For the ERPs studies, latency of the N170 and amplitudes of N75 and P1 for pattern-reversal stimulation and P1 and N170 for faces were measured in each child. For the oculomotor tasks, we focused on number of fixations and fixation duration in the fixation task, and the time to reach the target and the precision of the saccades in the prosaccade task. These parameters were entered in a principal component analysis.

Results:

The principal component analysis revealed that two components explained about 52% of the variability between the thirteen children. The first component was summarised mainly by the number of fixations, and the amplitudes of early visual evoked potentials (N75, P1 and N170) regardless of them being evoked by a social or non-social stimulation. The second component had larger weights on N170 amplitude and latency as well as saccade precisions. None of the components correlated with clinical scores such as ADOS or CARS. Future analyses will look at whether weighting the atypicalities found at each processing step influence the next step of processing at the individual level, therefore providing an individualised approach to bridge between clinical and research findings.

Conclusions:

These data highlights a strong link between oculomotor functions and visual evoked potentials in ASD, regardless of the stimuli being social or not. Moreover, the latency and amplitude of the N170 seems to depend on the ability of the child to reach the point of interest with accuracy, and fixate on it. This analysis strategy allows to determine the weights of low-level atypicalities on social difficulties. It further demonstrates the importance of controlling for sensory impairments (from ophthalmologic pathologies to oculomotor behaviour) in research looking at higher functions in ASD.