Time processing and predictive coding in Autism Spectrum Disorder

Theisinger, Laura Alena (2022): Time processing and predictive coding in Autism Spectrum Disorder. Dissertation, LMU München: Medizinische Fakultät

Abstract

Attenuated use of prior knowledge (predictive coding; in the context of the Bayesian framework) in some daily activities and more reliance on sensory evidence has often been demonstrated by individuals with Autism-Spectrum-Disorder (ASD)(Lawson, Rees, & Friston, 2014; Pellicano & Burr, 2012). This downweighing of prior beliefs may interfere and cause challenges when it comes to predicting future events and social motives. As impairments in social interaction and communication (American Psychiatric Association, 2013) provide the conditions for establishing a diagnosis of ASD, decreased sensitivity for temporal intervals (Falter, Noreika, Wearden, & Bailey, 2012) has also been associated with ASD, the reason for which is still unknown. Being aware of temporal (minute) changes in their environment has been described as one characteristic in both children and adults with ASD which are often unnoticed by their peers. This has been formally investigated by embedded figures tests (Muth, Hönekopp, & Falter, 2014; A Shah & Firth, 1983), also with typical feature search and conjunction search tasks (Kaldy, Giserman, Carter, & Blaser, 2016; Plaisted, O’Riordan, & Baron-Cohen, 1998). Recently researchers have investigated predictive coding frameworks with time perception in ASD using the classical central tendency effect (Karaminis et al., 2016), a typical perceptual bias coming from integration of prior knowledge of sampled interval range with the sensory input (Jazayeri & Shadlen, 2010; Shi, Church, & Meck, 2013). Often shorter durations are overestimated, and longer duration underestimated, defined as the central tendency effect. Karaminis et al. (2016) found that ASD performed much worse in temporal discrimination tasks than their matched controls, and demonstrated a decreased central tendency in ASD than predicted. Recently it has been shown that the central tendency effect relies upon the volatility of the sequence (Glasauer & Shi, 2018). To date, it is not clear how volatility influences time perception in ASD. Thus, the aim of this dissertation is to investigate the relevance of prior belief generation for interval timing in ASD using a duration reproduction task manipulating prior information on a trial-to-trial basis. We hypothesised that if individuals with ASD show less reliance on prior knowledge, the reproduction in the volatile (random) and involatile (random walk) environments should have less difference in the ASD group than in the typically developed (TD) control group. We found that the TD group adapts to a different environment more quickly and is influenced by the prior belief of that environment accordingly. By contrast, ASD individuals focused more on the sensory input, but were less influenced by prior knowledge of the environment, resulting in less flexibility in coping with the environment.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: autism, predictive coding, prediction errors, adaptability, volatility
• Themengebiete: 600 Technik, Medizin, angewandte Wissenschaften ; 600 Technik, Medizin, angewandte Wissenschaften > 610 Medizin und Gesundheit
• Fakultäten: Medizinische Fakultät
• Sprache der Hochschulschrift: Deutsch
• Datum der mündlichen Prüfung: 27. Oktober 2022
• 1. Bericht­erstatter:in: Koutsouleris, Nikolaos