Page 29 - It' about time: Studying the Encoding of Duration

Chapter 2 durations to shift away from the duration that was presented earlier (Becker & Rasmussen, 2007; Walker et al., 1981). Finally, studies investigating duration discrimination training have shown that training benefits such as increased discrimination sensitivity do not transfer to other non-trained durations (Bartolo & Merchant, 2009; Bueti & Buonomano, 2014; Karmarkar & Buonomano, 2003; Wright et al., 1997). All these studies are consistent with a channel-based model of duration processing in which duration is processed by groups of neurons that selectively respond to specific durations (Hayashi et al., 2015; Heron et al., 2012). Reading out the relative activation of groups of these neurons would allow for an explicit representation of duration that can be used for further processing. The observation that the duration after-effect does not transfer across modalities suggests that duration information is encoded separately for each modality, and combined later on during processing to form a more complete, multimodal representation of duration (Heron et al., 2012, 2013; van Wassenhove, 2009). A relevant question that is currently being investigated is that of the relative position of these duration-selective channels along the visual processing hierarchy (Hayashi et al., 2015; Heron et al., 2013; Li, Yuan, & Huang, 2015). It has been proposed that duration-selective neurons are present in early sensory areas for both auditory and visual information (Heron et al., 2012). Evidence for this claim comes from single cell recording studies in different mammals that have reported duration-selective neurons in both early auditory processing areas such as inferior colliculus and the auditory midbrain (Brand, Urban, & Grothe, 2000; Casseday, Ehrlich, & Covey, 1994; Ehrlich et al., 1997; He et al., 1997), as well as early visual areas such as area 17 & 18 (Duysens et al., 1996; Eriksson, Tompa, & Roland, 2008). One can argue that an early locus for duration processing can be beneficial given that the temporal integration window of neurons is known to increase along the visual processing hierarchy (Hasson, Yang, Vallines, Heeger, & Rubin, 2008). Therefore, encoding duration information at an early stage of visual processing would potentially allow for more precise encoding of the onset and offset of an event, resulting in more accurate duration encoding. Furthermore, several studies on duration perception have shown that adaptation to both temporal and non-temporal visual features can cause changes in the perceived duration of subsequent events, which are restricted to the location at which adaptation took place (Johnston et al., 2006; Ortega et al., 2012; Zhou, Yang, Mao, & Han, 2014). For 28