Atypical gaze patterns in autistic adults are heterogeneous across but reliable within individuals

Abstract

Background Across behavioral studies, autistic individuals show greater variability than typically developing individuals. However, it remains unknown to what extent this variability arises from heterogeneity across individuals, or from unreliability within individuals. Here, we focus on eye tracking, which provides rich dependent measures that have been used extensively in studies of autism. Autistic individuals have an atypical gaze onto both static visual images and dynamic videos that could be leveraged for diagnostic purposes if the above open question could be addressed. Methods We tested three competing hypotheses: (1) that gaze patterns of autistic individuals are less reliable or noisier than those of controls, (2) that atypical gaze patterns are individually reliable but heterogeneous across autistic individuals, or (3) that atypical gaze patterns are individually reliable and also homogeneous among autistic individuals. We collected desktop-based eye tracking data from two different full-length television sitcom episodes, at two independent sites (Caltech and Indiana University), in a total of over 150 adult participants (N = 48 autistic individuals with IQ in the normal range, 105 controls) and quantified gaze onto features of the videos using automated computer vision-based feature extraction. Results We found support for the second of these hypotheses. Autistic people and controls showed equivalently reliable gaze onto specific features of videos, such as faces, so much so that individuals could be identified significantly above chance using a fingerprinting approach from video epochs as short as 2 min. However, classification of participants into diagnostic groups based on their eye tracking data failed to produce clear group classifications, due to heterogeneity in the autistic group. Limitations Three limitations are the relatively small sample size, assessment across only two videos (from the same television series), and the absence of other dependent measures (e.g., neuroimaging or genetics) that might have revealed individual-level variability that was not evident with eye tracking. Future studies should expand to larger samples across longer longitudinal epochs, an aim that is now becoming feasible with Internet- and phone-based eye tracking. Conclusions These findings pave the way for the investigation of autism subtypes, and for elucidating the specific visual features that best discriminate gaze patterns—directions that will also combine with and inform neuroimaging and genetic studies of this complex disorder.

Detection of idiosyncratic gaze fingerprint signatures in humans

“…These accuracy rates are also well above the accuracy in identifying twin pairs reported by Kennedy et al, 14 (i.e. 29%) or in movie stimuli reported by Keles et al, 12 (i.e. 30%).…”

“…In this work we discovered that identification of individuals is possible simply by using gaze patterns sampled from 3-second viewing epochs on a large and diverse range of static visual stimuli depicting complex natural scenes. By sampling gaze patterns across 700 stimuli, we replicably show that we can gaze fingerprint individuals at levels much higher than chance (50-65% vs 1-2% chance levels), and at rates much higher than prior research identifying twin pairs (29%) 14 or identifying target individuals with a small number of movie clips (~30%) 12 . These higher accuracy rates are achieved by averaging gaze similarity values across large sets of stimuli and we find that for some semantic features with relatively less numbers of stimuli to average over, that accuracy drops.…”

“…In subsequent work, Keles and colleagues 12 showed autistic and non-autistic participants two different movie clips and measured summary gaze features such as the duration of looking at specific objects (e.g., faces, body parts, non-social content), as well how deviant their gaze heatmaps were from normative reference gaze heatmaps. Distances were computed between these multivariate gaze vectors for each individual and movie and identity was predicted based on the gaze vector with the minimal pairwise distance to the target individual.…”

“…While prior work has shown that individual gaze patterns can be stable and systematic [6][7][8][9][10][11] , such work does not address whether the 'target' individual can be accurately identified when compared to gaze patterns of many other 'distractor' individuals. As far as we know, only two prior studies have attempted to address this question 12,14 . These studies suggest that some, though not all, individuals may be gaze fingerprintable and that such gaze fingerprints are potentially driven in part by genetic similarity.…”

“…Many factors outside of the individual have been theorized to explain variance in gaze patterns, such as context, task, novelty, and characteristics of the stimuli themselves [1][2][3][4][5][6][7] . However, because gaze patterns can also be quite stable and systematic [6][7][8][9][10][11][12] , this suggests that there must also be a strong within-individual factor that explains variance in gaze patterns. Recent twin studies have revealed heritable common genetic variation to be one such powerful within-individual mechanism constraining gaze patterns [13][14][15] .…”