Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation
695—Social Cognition: Neural Processes and Disorders
Wednesday, November 13, 2013, 8:00 am - 10:15 am
695.05: Adults with autism show atypical, but consistent, patterns of gaze to dynamic social stimuli
*D. P. KENNEDY1, N. GANDHI2, R. ADOLPHS3; 1Psychological and Brain Sci., Indiana Univ., Bloomington, IN; 2Dept. of Bioengineering, UCSD, La Jolla, CA; 3Div. of Biol., Caltech, Pasadena, CA
Abstract Body: One of the most striking features of autism is abnormal attention toward social stimuli, which manifests in various ways from infancy to adulthood (e.g., lack of orienting to one’s name, lack of preference for biological motion, reduced attention to the eyes in faces, etc.). The vast majority of studies have examined visual social attention, but with only a few exceptions research on autism has used highly artificial stimuli and contexts, such as specific stimuli presented in isolation of everything else (e.g., cropped faces), stimuli reduced to basic components or features (e.g., point lights), or degraded stimuli (e.g., “bubbles” faces). The limited amount of research that has used semi-naturalistic stimuli, such as dynamic videos, has revealed important insight into the attentional abnormalities exhibited by those with autism. Here, we utilized this general approach in a sample of 20 adults with autism and 34 control participants, and recorded eye movements at 300 Hz while participants viewed (and listened to) a full episode of a television show (“The Office”; NBC Universal). The entire video (about 22 minutes in total) was divided into 3 parts, and eye tracking calibration and validation were repeated before each ~7 minute clip. Initial analyses were carried out on data from the first video clip alone. Using the Normalized Scanpath Saliency method along with a leave-one-out approach, we found that controls viewed the video more similarly to other controls as compared to autism (p < 0.001), whereas individuals with autism viewed the video more similarly to others with autism as compared to controls (p < 0.01). This pattern was found regardless of the number of people in the scene at any time (0, 1, 2, or 3) (all p < 0.05). Furthermore, at those moments when the gaze pattern of the autism group diverged significantly from the control group, individuals with autism were significantly more likely to fixate on high-contrast areas of the scene, relative to where controls were looking. This finding was replicated using data from the second 7 minute video clip as well. Overall, these results suggest that individuals with autism demonstrate selectivity in their atypical fixated locations and some degree of commonality in what captures their attention. Additional analyses are underway to further elucidate the attentional mechanisms underlying the similarities in gaze patterns within the autism group and differences between groups.
Lay Language Summary: We know that people with autism attend to the world differently than neurotypical controls. These differences emerge early in life and are particularly apparent for social stimuli (e.g., people, faces, voices) and in social contexts. For instance, someone with autism might not orient to one’s name when called or they may look less at people. The vast majority of studies that have examined social attention in autism have used highly artificial stimuli and artificial contexts, such as specific stimuli presented in isolation of everything else (e.g., cropped faces), stimuli reduced to basic components or features (e.g., point lights), or degraded stimuli (e.g., “bubbles” faces). These types of experiments, though essential and provide important insight, result in artificial scenarios that often fail to capture some of the key characteristics of real world social stimuli (e.g., dynamic, multi-modal, complex, subtle, continuous, etc.). Thus, the results of these studies may not generalize to these more complex real-world scenarios. An alternative approach, and the one used here, is to use stimuli that better approximates stimuli encountered in the real world, such as videos of people interacting. The limited amount of research that has used such semi-naturalistic stimuli has already revealed important insight into the attentional abnormalities exhibited by those with autism. In this study, we extended this general approach, and have attempted to reveal new details and mechanisms underlying differences in how individuals with autism view the world. We used an eye-tracker (a device that precisely measures the locations of one’s gaze 300 times a second) while 20 adults with autism and 34 control participants watched and listened to a full episode of the television show, “The Office” (NBC Universal). Using a methods known as Normalized Scanpath Saliency, we were able to show that the autism group did view the video differently from neurotypical controls. Somewhat unexpectedly, however, we also found that individuals with autism viewed the video similarly to others with autism. In other words, it wasn’t just that people with autism all viewed the video idiosyncratically, but rather, that there were shared features of the video that people with autism attended to. But what underlies these differences? We next sought to determine what drove abnormal, but coherent, viewing patterns in individuals with autism. At moments when gaze was different between groups, we found that individuals with autism were significantly more likely to be fixating on high-contrast areas of the scene, relative to where controls were looking. This means that individuals with autism demonstrate selectivity in their fixated locations as well as some degree of commonality in what captures their attention and when. Additional analyses are underway to further elucidate the attentional mechanisms underlying the similarities in patterns of gaze within the autism group and differences in patterns of gaze between groups. Why is this important? We are active participants in the world, and one important task we all have is to seek out information that is relevant to the task at hand; in this case, cues that are relevant for understand a social scenario. If we aren’t able to extract this information from our environment, surely our ability to navigate the social world will be impaired.
Neuroscience 2013 (43rd annual meeting of the Society for Neuroscience)Exit