Mental rotation versus invariant features in object perception from different viewpoints: an fMRI study

Abstract

It has been proposed that object perception can proceed through different routes, which can be situated on a continuum ranging from complete viewpoint-dependency to complete viewpoint-independency, depending on the objects and the task at hand. Although these different routes have been extensively demonstrated on the behavioral level, the corresponding distinction in the underlying neural substrate has not received the same attention. Our goal was to disentangle, on the behavioral and the neurofunctional level, a process associated with extreme viewpoint-dependency, i.e. mental rotation, and a process associated with extreme viewpoint-independency, i.e. the use of viewpoint-invariant, diagnostic features. Two sets of 3-D block figures were created that either differed in handedness (original versus mirrored) or in the angles joining the block components (orthogonal versus skewed). Behavioral measures on a same-different judgment task were predicted to be dependent on viewpoint in the rotation condition (same versus mirrored), but not in the invariance condition (same angles versus different angles). Six subjects participated in an fMRI experiment while presented with both conditions in alternating blocks. Both reaction times and accuracy confirmed the predicted dissociation between the two conditions. Neurofunctional results indicate that all cortical areas activated in the invariance condition were also activated in the rotation condition. Parietal areas were more activated than occipito-temporal areas in the rotation condition, while this pattern was reversed in the invariance condition. Furthermore, some areas were activated uniquely by the rotation condition, probably reflecting the additional processes apparent in the behavioral response patterns.

Motion in the mind's eye: Comparing mental and visual rotation

“…In addition, positing a critical computational role for the inferior parietal cortex would provide some account of why this region is the most consistent area identified with mental rotation in previous studies (e.g., Alivisatos & Petrides, 1997;Barnes et al, 2000;Bonda et al, 1995;Carpenter et al, 1999;M. S. Cohen et al, 1996;Gauthier et al, 2002;Harris et al, 2000;Jagaroo, 2004;Richter et al, 2000;Tagaris et al, 1998;Vanrie et al, 2002;Vingerhoets et al, 2002;Zacks et al, 1999). However, further investigation is warranted before these claims can be substantiated.…”

“…Alivisatos & Petrides, 1997; Barnes et al, 2000;Bonda, Petrides, Frey, & Evans, 1995;Carpenter, Just, Keller, Eddy, & Thulborn, 1999; M. S. Cohen et al, 1996;Gauthier et al, 2002;Harris et al, 2000;Jagaroo, 2004;Richter et al, 2000;Tagaris et al, 1998;Vanrie, Béatse, Wagemans, Sunaert, & Van Hecke, 2002;Vingerhoets, de Lange, Vandemaele, Deblaere, & Achten, 2002;Zacks, Rypma, Gabrieli, Tversky, & Glover, 1999). This region has been associated with processing of higher order motion (Claeys, Lindsey, De Schutter, & Orban, 2003) and lies in close proximity to areas of the human brain sensitive to motion (hV5/MT ; Huk, Dougherty, & Heeger, 2002;Nakamura et al, 2003;Orban et al, 2003;Tootell et al, 1995), suggesting that perhaps mental rotation involves motion imagery that is processed much like visual motion.…”

Using morphs of familiar objects to examine how shape discriminability influences view sensitivity

“…However, there is considerable disagreement about the point at which this view sensitivity dissipates as shape discrimination becomes easier. Many of the studies investigating the view sensitivity of recognition across rotations in depth have tested people's ability to identify novel objects (e.g., Bülthoff & Edelman, 1992;Lawson, 2004b;Lawson & Bülthoff, 2006;Rock & DiVita, 1987;Tarr, 1995;Vanrie, Béatse, Wagemans, Sunaert, & Van Hecke, 2002;Vanrie et al, 2001). However, as compared with stimuli from categories that people distinguish between in everyday, basic-level object recognition, these novel stimuli probably differed in terms of their shape discriminability, level of recognition, and shape properties such as complexity and symmetry, as well as in their reduced familiarity, semantic associations, and nameability.…”

Fixed versus dynamic orientations in environmental learning from ground-level and aerial perspectives

“…This region is commonly associated with unilateral neglect (e.g., Robertson & Marshall, 1993), a deWcit in spatial attention. However, a wealth of neuroimaging studies has linked this region to spatial transformation processes such as mental rotation (e.g., Alivisatos & Petrides, 1997;Barnes et al, 2000;Bonda, Petrides, Frey, & Evans, 1995;Carpenter, Just, Keller, Eddy, & Thulborn, 1999;Cohen et al, 1996;Gauthier et al, 2002;Harris et al, 2000;Richter et al, 2000;Tagaris, 1998;Vanrie, Béatse, Wagemans, Sunaert, & Van Hecke, 2002;Zacks, Rypma, Gabrieli, Tversky, & Glover, 1999). One could imagine that changes in orientation essentially require one to mentally manipulate current information to align it with the reference frame of preceding information.…”