Selective role of lingual/parahippocampal gyrus and retrosplenial complex in spatial memory across viewpoint changes relative to the environmental reference frame

Abstract

Remembering object locations across different views is a fundamental competence for keeping oriented in large-scale space. Here we investigated such ability by comparing encoding and retrieval of locations across viewpoint changes relative to different spatial frames of reference. We acquired functional magnetic resonance images while subjects detected target displacements across consecutive views of a familiar virtual room, reporting changes in the target absolute position in the room (stable environmental frame), changes in its position relative to a set of movable objects (unstable object-based frame), and changes relative to their point of view (control viewer-centered frame). Behavioral costs were higher for the stable environmental frame, and a cortical network including the lingual/parahippocampal gyrus (LPHG) and the retrosplenial complex (RSC) selectively encoded spatial locations relative to this frame. Several regions, including the dorsal fronto-parietal cortex and the LPHG, were modulated by the amount of experienced viewpoint change, but only the RSC was selectively modulated by the amount of viewpoint change relative to the environmental frame, thus showing a special role in coding one's own position and heading in familiar environments.

Introduction

Human beings have an outstanding capacity to seamlessly recognize scenes and to remember spatial locations despite intervening changes in point of view, such as those occurring when we walk or are passively transported through a familiar environment. This ability, which is fundamental to spatial orientation, can be experimentally measured by asking observers to study a scene from a given viewpoint, and then, after a true, imagined, or virtual self displacement, to detect whether a given object has been moved, or to point to a memorized location [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. In principle, people could simply match the studied and the test view through some process of mental rotation. However, memory across viewpoint changes involves some more specific process: people perform better when they (physically or virtually) move in the environment around a stationary object array, than when they are still and an object array rotates around them [5], [6], [7], [18], [19], [20], [21], although the two situations appear identical from a geometrical standpoint. This and other lines of evidence [22], [23], [24] show that comparing different views of a scene involves some form of mental self-rotation, which is distinct from the process of mentally rotating objects.

In terms of spatial reference frames, mental rotation of objects requires transforming an object-based frame relative to a fixed egocentric reference, while compensating for occurred viewpoint changes requires transforming an egocentric frame relative to an external (i.e., allocentric) reference (a “perspective transformation”: [21]). In principle, any set of objects can be used to establish an allocentric reference frame relative to which to evaluate a viewpoint change, and the operations involved in encoding spatial locations and then in updating memorized locations after a perceived viewpoint change may not depend on the particular set of objects chosen. However, human neuroimaging studies have shown not only that perceptual coding in egocentric and allocentric reference frames have distinguishable neural signatures (reviewed in [25]), but also that two different forms of allocentric representations can be distinguished: object-based reference frames, encoding spatial locations relative to arbitrary objects, and environmental reference frames, encoding spatial locations relative to some fixed features of the environment [21], [25]. A particular set of cortical regions are selectively activated when a spatial judgment is referenced to enduring environmental features but not to unstable objects [26]. These regions include portions of the medial temporal cortex and adjacent antero-medial occipital lobe (fusiform, lingual and posterior parahippocampal gyrus), the retrosplenial cortex and the precuneus. Their selectivity for environmental referencing depends on the stability of the spatial location of environmental features over time and not on their perceptual features or orienting value [25]. Intracerebral recordings within the posterior medial temporal lobe recently showed that this kind of allocentric selectivity constitutes a separate, later processing stage relative to early visual scene processing [27].

On the basis of the distinction between environmental and object-based allocentric reference frames, we suggest that perspective transformations may be preferentially associated with the former. Indirect evidence for this idea comes from neuroimaging studies which have shown that the cortical regions selective for environmental reference frames are involved in recognizing a scene across different views (see [28] for a recent review) and in perspective vs. object-based transformations [29]. However, previous studies requiring to memorize and recall object locations across viewpoint changes have used either stable environmental features [16], [29] or unstable configurations of objects [18], [20]. In the present event-related functional magnetic resonance imaging (fMRI) study, we adapted a viewpoint-change paradigm often used in behavioral research [7], [13] and compared perspective transformations relative to object-based vs. environmental reference frames, under the hypothesis that the neural circuit described above is actively exploited either during memory encoding or when updating memorized locations after a viewpoint change, but only if the established reference frame includes stable environmental features.

We asked observers to encode the spatial location of a target object in a virtual room with respect either to stable, familiar features of the scene (room frame), or to an arbitrary, unstable object set (objects frame). Observers then experienced a certain amount of viewpoint change, with “views” defined either relative to the room or to the object set, and asked them to judge whether the target was in the same spatial location as before, with “same” locations again defined relative to either the room or the object set. Importantly, the manipulation of the amount of viewpoint change allowed us to test a further prediction: regions actively involved in perspective transformations relative to the environmental reference frame should be selectively modulated by the amount of experienced viewpoint change, only when this is defined in room-based, but not in object-relative coordinates.