25994
Internal Noise and Global Motion Pooling and Their Relationship with Autistic Traits in Typically Developed Adults.

Friday, May 12, 2017: 5:00 PM-6:30 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
E. Orchard and J. van Boxtel, Monash Biomedical Imaging, Monash University, Melbourne, Australia
Background:

People with autism spectrum disorder (ASD) display abnormalities in motion processing. Motion perception abnormalities in ASD have been linked to high levels of internal noise (unreliable neuronal fluctuations), and deficits in global motion pooling (integrating local motion into a global percept). There are two types of internal noise: additive (‘baseline’ that is constant across different amounts of external, i.e. stimulus, noise); and multiplicative (proportional to external noise). Previously, internal noise in ASD has been investigated through additive internal noise, comparing ASD and control groups. However, multiplicative noise has not yet been investigated in terms of ASD and neither type of internal noise has been investigated in terms of traits of autism in typically developed populations.

Objectives:

To investigate how additive and multiplicative internal noise and global motion pooling vary across number of autistic traits in typically developed adults.

Methods:

To achieve these aims, we employed a visual motion integration task with a typically developed population, measuring autistic traits with the Autism Spectrum Questionnaire (AQ). Forty-five adults (Mage= 22.18, SDage= 4.96) indicated average direction (left or right) of 200 dots across eight external noise levels (varied with increased directional noise). Additive and multiplicative internal noise and global motion pooling were calculated from task accuracy and consistency across external noise levels, using equivalent noise analysis (ENA), within a double-pass paradigm. ENA compares human performance to an ‘ideal observer’ with known levels of additive internal noise and pooling (Lu & Dosher, 2008). An equivalent noise function is then fit to these data to estimate internal noise and pooling (Manning et al., 2014). This method of analysis has been used within populations of both ASD (Manning et al., 2015) and typically developed individuals (Burgess & Colborne, 1988).

Results:

A Mann-Whitney U test revealed no significant difference in additive noise between those with high (Mean Rank = 22.45, n = 20) and low (Mean Rank = 20.64, n = 22) AQ scores, U = 201.00, p= .56, two-tailed. No significant correlation was found between multiplicative noise and AQ score: r= -0.03, p= .84, or between global motion pooling and AQ score: r= 0.12, p= .448. Bayesian statistics indicate these values give moderate support for the null hypothesis, with BF10=0.12, and 0.16, for multiplicative noise and global motion pooling, respectively.

Conclusions:

Our results suggest that differences in internal noise and global motion pooling may not be present across a sub-clinical spectrum of autistic traits. In support of Manning et al. (2015), who found no difference in internal noise between ASD and control groups using similar methods, our results may be tentatively extrapolated to clinical ASD populations. We offer support for the absence of a difference in internal noise and global motion pooling between clinical ASD and typically developed populations.