The Effect of Perspective Based Change Blindness with Regard to the Results of Tracy et al (2005). H. Allisson R. Seiler Indiana University [email protected] Beenish Chaudry and Toshi Uchido Dept. of Computer Science Indiana University [email protected] & [email protected] Jacob Koruth Dept. of Telecommunications Indiana University [email protected] (in partial fulfillment of the requirements for COGS-Q551)For almost two centuries, multistable images have been seen as “a unique window into the involvement and interplay of critical underlying processes in the visual system” (Long and Toppino, 2004, p.748—see, also, for a review of the experimental history and alternating views of perceptual ambiguity) including the transition from sensory to higher cognitive functions (Blake and Logothesis, 2002). Recent work utilizing imaging techniques, both electrophysical and fMRI (Lumer et al. 1998; Inui et al. 2002; Leopold et al. 2002; Müller et al. 2005; Haynes and Rees, 2005, Kamitani and Tong, 2005, Slotnick and Yantis, 2005, and Tracy et al. 2005), have both suggested several areas within the frontal, parietal, temporal, and striate cortex as the origins of multistability. Tracy et al’s (2005) recent work implicated the medial frontal gyrus (BA6) as having a primary role in the reversal of an ambiguous stimulus (Necker Cube), as it may serve as an intermediary between knowledge of body position, visual perspective relative to the object, and object orientation and the motor output centers. Other work by Inui et al (2000) did examine specific neural substrates regarding depth perception and the Necker Cube and suggested that prefrontal regions associated with motor and mental imagery plays a role in the perception of visually presented object. It was felt by Tracy and colleagues that a kind of specific somatosensory information is processed by the medial frontal cortex and is key to the creation of the three dimensional mental image of the cube. They suggested that the alternating perceptions of the cube are due to a “fictitious change in view position relative to the object (Tracy et al, 2005, p 280)” may be due to this area attempting to resolve the ambiguity. Tracy and colleagues did not find any significant activation of this area compared to either suppression tasks (where the subjects held one of the faces of the Necker cube fixed forward or back) or to rest. If this medial frontal gyrus is serving as a kind of reference for body position relative to viewpoint, which is consistent with work by Leopold and Logothesis (1999) and Sterzer, Russ, Priebisch, & Kleinschmidt (2002), then unattended perspective changes might cause increased activation in BA6, especially during suppression tasks where subjects are instructed to hold the ambiguous figure constant. Additionally, it is likely that spontaneous reversals of the cube will occur despite conscious effort on the part of the subject to hold the cube stable. We propose to replicate Tracy and colleagues fMRI findings, with three key differences other than the use of a different fMRI scanner, to ensure the robustness of their findings. One area of concern is that Slotnick and Yantis (2005) also found activation in the medial frontal gyrus (BA6) in their fMRI - 1 -Changes in Perspective & Tracy et al’s (2005) Results study, which they attributed to the act of subjects in their study pushing a button—motor movement. In Tracy et al’s (2005) work, during active reversal tasks, when significant activity in BA6 was observed, subjects had been pointing to whichever face of the Necker cube they perceived as being forward, also motor movement. Additional concerns are: there was no corresponding pointing motor movement during the suppression tasks, as Tracy and colleagues found it initiated spontaneous reversals, nor was there an objective way to measure the number of spontaneous reversals during suppression, only the subjective reporting of the subjects during debriefing; it should be noted that Tracy et al (2005) did have a control condition where the subject simply alternated periods of pointing and fixation (rest). We feel that by using a bitebar, which also helps to minimize head motion, that has a sensor that can be activated by pressure of the teeth, we can count reversals during both active reversal and suppression tasks; this allows us to have a consistent measure of what the subject perceives across all visual stimuli in both replication and experimental trials. Similar to the methodology utilized by Halgren et al (2003) in their MEG study, we will measure cortical activity from the subject activating the sensor only, with no visual stimuli. We can then subtract that motor activity from the experimental data leaving only that activity related to the perception and processing of the stimuli; thus we will answer conclusively if the significant activity seen by Tracy et al (2005) in the medial frontal gyrus is related to ambiguous figure perception. Additionally, we will examine how unattended perspective changes, created by the use of invisible stimuli via the phenomenon of change blindness (Maier et al., 2003; Simmons & Ambinder, 2005; Haynes & Rees, 2005), impacts both cortical activity in the medial frontal gyrus (BA6) and the number of spontaneous reversals perceived by the subjects. It is predicted that both activity and the number of spontaneous reversals will increase, especially during the suppression tasks of the ambiguous figure as it is probable that the changes of orientation of the cube will be processed in the visual cortex (Haynes and Rees, 2005; Leopold & Wilke, 2005) and cause the somatosensory information regarding the subject’s visual perspective to be updated. As the experimental conditions of our fMRI experiment depends on presenting skewed stimuli interlaced with upright ones, it is important to know at what rate we can present these disparate images without the subject reporting that two images have been seen. Additionally, the very nature of presenting a changing stimulus, even if the two images are unchanged, within a trial may itself cause shifts in attention and evoke responses in areas involved in motor preparation (Rizzolatti, Riggio, and Sheliga, 1994, as cited in Tracy et al, 2005). However, Leopold et al (2002) found when inducing patterns were presented intermittently—specifically when blank periods were introduced between