Navigation In Virtual Space Research Paper

In addition to visual simulation, many VR systems are now able to accurately represent the location of auditory sources using hardware specifically designed for the task. Haptic feedback can also be available to some extent. This involves applying forces to the user’s body to simulate the presence of objects in the environment. However, haptic systems tend to impose considerable constraints on the user’s freedom of movement and are generally only used on particular joints for specific purposes. For example, the thumb and forefinger may be attached to a device that can apply pressure to signal the size of objects being grasped in the VE.

2. Why Use VR?

The advantages of VR depend very much on the use to which it is put; in architecture, for example, entire buildings can be experienced as a simulation without a single brick being laid. In psychology there are numerous benefits. Within a VE the stimuli are completely under the control of the experimenter, and responses can be conveniently recorded. This is especially useful in studies involving large spaces, circumventing the need for large-scale motion tracking and the practical problems of controlling a large environment. Also, in a VE the normal rules of physics need not apply, so that impossible manipulations of space and objects can occur. For example, non-Euclidian geometry can be simulated (e.g., Ruddle, in press); objects can be repositioned and physical dimensions changed instantaneously. Finally, one of the most exciting advantages of VR in neuroscience is that it can be used to simulate movement through space while a subject’s head is kept still for the purposes of functional neuroimaging.

3. ‘Immersion’ And The Correspondence Between Real And Virtual Behavior

A major concern in VR is the issue of immersion. This refers to the degree to which users accept the VE as a space in which they can act, and how convincing a simulation of reality the VE provides. There is evidence that variations in the VR interface have effects on the performance of the user. Chance et al. (1998) examined subjects’ ability to judge direction to target locations in a virtual maze—a task requiring a series of turns and translations to be integrated into a representation of displacement (often referred to as path integration). Their interface had three modes: a motion-tracking, head-mounted display mode in which the subjects actually walked and turned physically (Walk mode); a mode in which movements were made using a joystick (Virtual Turn mode); and a mode in which turns were physically enacted but translations were not (Real Turn mode). The experimenters found significantly better performance in the Walk mode than in the Virtual Turn mode, suggesting that, for the purpose of path integration, immersion is best achieved using real walking to control movement in the VE.

Alternatively, research on imagined movements by Decety and Jeannerod (1995) suggests that, even when the environment is simulated entirely using visual imagery, actions are performed in a similar way to that in a computer-mediated VE. In this study, subjects imagined themselves moving through previously experienced virtual doorways of varying widths, and the duration of each action was related to the door width in the same way as during VR trials. Similarly Ghaem et al. (1997) found a good correspondence between real and imagined walking on a route through Paris in terms of the speed of movement. However, the validity of generalizing between virtual and real spaces remains a major concern. A number of studies have considered this issue. Henry and Furness (1993) found that architects consistently underestimated the size of VEs. Conversely, Ruddle et al. (1997) found subjects’ performance on judgments of distance and direction in a large virtual building to be compatible with performance in the same real building. Witmer et al. (1996) compared route and building layout knowledge in subjects trained in a real building, the same building in VR, and the building learned from photographs and verbal descriptions. They found that when subjects were tested in the real building, virtual exposure gave better performance than verbal pictorial, but neither was as effective as learning in the real building. In another experiment, Witmer and Klein (1998) found that subjects underestimate distances in both real and virtual spaces, but more so in VR. Also, Waller et al. (1998) found that, given long exposure, VE training of an environment eventually surpassed real-world training in tests of route learning. Finally, Klatzky et al. (1998) found a systematic overestimation of turn amplitude in simulated environments, which was not present in the real world.

Evidence from studies of normal subjects is mixed, then, but it appears that spatial cognition in VEs is at least closely related to that in the real world. Clinical studies support this: Brooks et al. (1999) showed that practice in a virtual simulation of a rehabilitation unit transferred to real-world route-finding performance in a 53-year-old patient with anterograde amnesia.

4. The Psychology Of Navigation In Virtual Space

A popular model of the acquisition of spatial information, the landmark–route–survey, or LRS, model holds that spatial information is acquired in stages during exploration. Landmark information is acquired first, followed by information concerning specific routes before, finally, a survey-like spatial representation is formed. (Siegel and White 1975, Thorndyke and Hayes-Roth 1982).

Alternatives to this model have been suggested by experiments using virtual spaces. Aginsky et al. (1997) examined driving behavior in VR and propose that there are two parallel strategies—visually dominated and spatially dominated. The former, they say, relies on way-finding using visual decision points (landmarks, essentially), which are not integrated into a survey representation. In the later strategy, survey knowledge is used from the very start, rather than learned in stages. Colle and Reid (1998) agree with this; in their study of learning metric knowledge in small spaces, they found local survey knowledge was learned rapidly without recourse to landmarks and routes. They propose a dual mode model where local survey knowledge is achieved rapidly, but larger-scale layouts are learned according to the LRS model.

VR has also been used to probe the nature of the formation of the representation of space. Gillner and Mallot (1998) examined the use of information restricted to the views of landmarks and movements local to choice points and found that, at least for some subjects, the maze could be learned from this restricted information alone, and more abstract survey information could be extracted.

5. Strategy And Gender

Virtual studies of navigation have revealed interesting sex differences in navigational strategies. Astur et al. (1998) produced a virtual replication of the Morris water maze: a pool surrounded by landmarks in which a platform that is hidden under the water must be found). They found that males navigated to the platform consistently better than females. In another virtual maze study, Moffat et al. (1998) found that males made fewer errors and completed maze tasks more quickly than females. These findings are typical of the well-known gender differences in spatial cognition (e.g., Silverman and Eals 1992). We note that gender differences in navigation often reflect differences in strategy, as shown by Sandstrom et al. (1998) in another virtual water maze. In their task the location of the goal could be determined either relative to visual landmarks provided around the maze, or relative to the geometry of the room surrounding the maze. Males would use either landmarks or geometry as available whereas females tended to predominantly use the landmarks, thus performing worse when the landmarks were removed. In looking at gender differences, the use of VR itself tends to introduce a confounding factor: males tending to be more familiar with navigation in VR by virtue of having spent more time playing video games. Interestingly, Viaud-Delmon et al. (1998) found no difference in ability to estimate the amplitude of a passive turn while blindfolded, but did find males more likely to recalibrate their estimations after receiving distorted visual feedback while being turned in VR.

6. Neural Aspects Of Navigation In Virtual Space

6.1 Background

Navigation requires many different processes, such as generation of body movements, recognition of landmarks, sense of direction, memory for routes and memory for survey-type knowledge. Thus disentangling the neural correlates of navigation will not be easy. Before considering the uses of VR in this pursuit, we briefly summarize the background to this area.

Electrophysiological recording in awake behaving animals has revealed several types of representation of spatial location in the brain. Neurons in posterior parietal areas have been found to represent the location of stimuli or actions in reference frames related to parts of the body (‘egocentric’ reference frames). For example, the firing rates of neurons in the posterior parietal cortex of monkeys may reflect the location of a stimulus relative to its eye or head or body. They also appear to encode the information necessary to translate between reference frames (e.g., given a location relative to the eye, computing that location relative to the head. See Andersen et al. 1985). Of particular interest with respect to navigation, ‘place cells’ in the rat hippocampus represent the rat’s current location with respect to its environment (an ‘allocentric’ reference frame (e.g., O’Keefe and Nadel 1978). These cells become active whenever the rat is in a particular portion of an open environment, independent of the rat’s orientation. The complementary representation, of heading direction independent of location (‘head-direction cells’), was found in an area close to the hippocampus, the presubiculum.

Consistent with the finding of place cells, and the idea that the hippocampus maintains a survey-type representation or ‘cognitive map’ of the environment (O’Keefe and Nadel 1978), lesions to the hippocampus impair a rat’s ability to navigate to the hidden platform in the Morris water maze (Morris et al. 1982).

The egocentric parietal and allocentric hippocampal systems undoubtedly work together, not least because all sensory information enters the brain in an egocentric form and the read-out of an allocentric map must at some stage be translated into egocentric motor commands such as turning left or right. Indeed, there is evidence that the firing of neurons in area 7a of parietal cortex (the part that projects to parahippocampal cortex) also reflects stimulus locations relative to the monkey’s environment (Snyder et al. 1998). The complementary roles of these two brain regions may also be understood in terms of the time scales over which they support behavior. The hippocampus is involved in long-term memory, whereas the parietal lobe is involved in perception and action over short time scales. It makes sense to use an allocentric representation to store spatial locations over the long term, as the position of the body will have changed before the information is to be used. However, to act on a spatial location (e.g., to reach with the hand or look with the eye) an egocentric representation (e.g., the position relative to the hand or eye) is more useful (see Burgess et al. 1999).

The rat’s place cell representation of space appears to be derived from an appreciation of the relative positions of the walls of its enclosure (O’Keefe and Burgess 1996), presumably represented in the parahippocampal cortices that provide the input to the hippocampus. Functional neuroimaging in humans has also implicated parahippocampal cortex in the perception of space generated by visual stimuli such as enclosing walls (Epstein and Kanwisher 1998). Lesions to the medial temporal lobe (particularly those including parahippocampal cortex) can impair the recognition of landmarks or scenes (see Farrell 1996, Aguirre and D’Esposito 1999). Finally, we should note that navigation to a place can be contrasted with simply learning a sequence of body turns. In rats, learning to go to a place is often learned rather quickly, such that the rat will approach the learned place whether this requires a left or right body turn. However, after several trials of making a particular turn, the turn response will tend to predominate: the rat will make that turn whichever place it leads to. Interestingly, while hippocampal lesions tend to impair place navigation, lesions to the caudate nucleus (part of the basal ganglia) prevent development of a predominant body-turn response (Packard and McGaugh 1996), consistent with a role in storing well learned routes.

6.2 VR And Functional Neuroimaging

The recent advent of functional neuroimaging provides a new opportunity for VR in investigating the neural basis of behavior. This technique involves measuring some metabolic correlate of brain activity while the subject performs various cognitive tasks. The subject’s head must be kept still during this process, which presents a problem for investigating many real-world behaviors, including navigation. However, VR can allow a realistic impression of navigation via a screen in front of the eyes and a keypad or joystick by the side, even while the head is fixed. An obvious limitation of this technique is that it excludes the proprioceptive and vestibular inputs experienced in real navigation.

There have now been several studies combining VR and functional neuroimaging, most making use of the development of realistic first-person perspective computer games. These have provided a cheap and effective means of producing VR displays in modifiable environments rendered at higher frame-rates and with a richer selection of textures than all but the most expensive commercial VR packages. Aguirre et al. (1996) created a simple maze of five dungeon-like corridors, each identified by an object placed at the end. This was used with fMRI to scan subjects while they first explored the maze and then navigated from the ends of four of the corridors to the end of the fifth. Their main finding was greater activation of bilateral parahippocampal areas associated with both exploration and navigation relative to a control condition of moving down an endless corridor. Maguire et al. (1998a) found activation of the right parahippocampus when subjects were scanned using PET as they explored a maze of similarly simple form, containing several rich textures and objects, relative to a control condition of viewing static and moving textures. Interestingly, no such activation was found when subjects explored a featureless maze without rich textures or objects. This difference may arise from neural involvement in encoding of object and texture information within this task, or from the greater level of immersion generated by the extra detail. Aguirre and D’Esposito (1997) also showed the parahippocampus to be involved in recognition of a location within a virtual town and indicating the direction to one location from another. Inferior parietal areas were also activated in these tasks, with relatively more parahippocampal activation in the place recognition task and relatively more parietal activation in the direction task (consistent with a role in translating between allocentric and egocentric reference frames).

Maguire et al. (1998b) directly investigated the brain regions involved in navigation by making subjects find their way between locations within a complex texturerich VR town while in a PET scanner. This town was created to appear as lifelike as possible and to include many different possible routes between any two locations. The right parahippocampus and hippocampus were activated by navigation relative to following a route of arrows through the town. The subject’s accuracy of navigation was found to correlate with activation of the right hippocampus and the right inferior parietal cortex. Again, these results are consistent with the recall of allocentric information from long-term storage in the hippocampus, followed by translation into an egocentric representation of movement direction. Interestingly, speed of virtual movement correlated with activation in the caudate nucleus, and performance of novel detours caused activation of frontal areas (often associated with planning).

A recent study by Gron et al. (2000) used a VR maze in conjunction with fMRI to investigate the neural bases for such differences. Their results are interesting in that they show different brain areas active in men and women as well as different performance. Both groups showed parietal, parahippocampal and right hippocampal activation associated with navigation. However, in men, the left hippocampus was also activated, whereas in women right parietal and right prefrontal areas were activated. The authors suggest that the differences reflect a gender bias in strategy, with women relying more heavily on landmarks, and men using geometric information as well.

Finally, another recent study used VR in combination with electroencephalography of the human brain. One of the most striking signals in the electroencephalogram (EEG) of rats is the appearance of a roughly sinusoidal oscillation of around 6 to 10 Hz (known as the ‘theta’ rhythm, Vanderwolf 1969) whenever the rat is involved in displacement movements of the head. However, this rhythm had not been reported in humans. Kahana et al. (1999) recorded EEG from subdural electrodes in epileptic patients while moving through a virtual space. They reported finding episodes of theta rhythm during this virtual movement, perhaps providing a link between the electrophysiological mechanisms of rats and humans in spatial navigation.

7. Conclusions

The increasing openness and sophistication of commercial VR software and computer games such as Quake III and Unreal open the possibility of further experiments involving good immersion and an ability to make subtle manipulations of the environmental space. This will allow the design of behavioral and functional neuroimaging paradigms that address our processing of large-scale space in ways not possible in reality. Thus, we foresee a bright future for the application of VR in elaborating the neural basis of every day tasks such as spatial navigation.

Bibliography:

1. Aginsky V, Harris C L, Rensink R A, Beusmans J M 1997 Two strategies for learning a route in a driving simulator. Journal of Environmental Psychology 17: 317–31
2. Aguirre G K, D’Esposito M 1997 Environmental knowledge is subserved by separable dorsal ventral neural areas. Journal of Neuroscience 17(7): 2512–18
3. Aguirre G K, D’Esposito M 1999 Topographical disorientation: a synthesis and taxonomy. Brain 122: 1613–28
4. Aguirre G K, Detre J A, Alsop D C, D’Esposito M 1996 The parahippocampus subserves topographical learning in man. Cerebral Cortex 6(6): 823–9
5. Andersen R A, Essick G K, Siegel R M 1985 The encoding of spatial location by posterior parietal neurons. Science 230: 456–8
6. Astur R S, Ortiz M L, Sutherland R J 1998 A characterization of performance by men and women in a virtual Morris water task: A large and reliable sex diff Behavioural Brain Research 93(1–2): 185–90
7. Brooks B M, McNeil J E, Rose F D, Greenwood R J, Attree E A, Leadbetter A G 1999 Route learning in a case of amnesia: The efficacy of training in a virtual environment. Neuropsychological Rehabilitation 9(1): 63–76
8. Burgess N, Jeffery K J, O’Keefe J (eds.) 1999 The Hippocampal and Parietal Foundations of Spatial Cognition. Oxford University Press, Oxford, UK
9. Chance S S, Gaunet F, Beall A C, Loomis J M 1998 Locomotion mode affects the updating of objects encountered during travel: The contribution of vestibular and proprioceptive inputs to path integration. Presence 7: 168–78
10. Colle H A, Reid G B 1998 The room effect: Metric spatial knowledge of local and separated regions. Presence 7: 116–28
11. Decety J, Jeannerod M 1995 Mentally simulated movements in virtual reality: Does Fitts’s law hold in motor imagery? Behavioural Brain Research 72: 127–34
12. Epstein R, Kanwisher N 1998 A cortical representation of the local visual environment. Nature 392(6676): 598–601
13. Farrell M J 1996 Topographical disorientation. Neurocase 2: 509–20
14. Ghaem O, Melle E, Crivello F, Tzourio N, Mazoyer B, Berthoz A, Denis M 1997 Mental navigation along memorized routes activates the hippocampus, precuneus, and insula. Neuroreport 8(3): 739–44
15. Gillner S, Mallot H 1998 Navigation and acquisition of spatial knowledge in a virtual maze. Journal of Cognitive Neuroscience 10: 445–63
16. Gron G, Wunderlich A P, Spitzer M, Tomczak R, Riepe M W 2000 Brain activation during human navigation: Genderdifferent neural networks as substrate of performance. Nature Neuroscience 3: 404–8
17. Henry D, Furness T 1993 Spatial perception in virtual environments: Evaluating an architectural application In: IEEE, 1363, vs. 33–40
18. Kahana M J, Sekuler R, Caplan J B, Kirschen J M, Madsen J R 1999 Human theta oscillations exhibit task dependence during virtual maze navigation. Nature 399: 781–4
19. Klatzky R L, Loomis J M, Beall A C, Chance S S, Golledge R G 1998 Spatial updating of self-position and orientation during real, imagined, and virtual locomotion. Psychological Science 9: 293–8
20. Maguire E A, Burgess N, Donnett J G, O’Keefe J, Frith C D 1998a Knowing where things are parahippocampal involvement in encoding object locations in virtual large-scale space. Journal of Cognitive Neuroscience 10: 61–76
21. Maguire E A, Burgess N, Donnett J G, Frackowiak R S, Frith C D, O’Keefe J 1998b Knowing where and getting there: a human navigation network. Science 280(5365): 921–4
22. Moffat S D, Hampson E, Hatzipantelis M 1998 Navigation in a ‘virtual’ maze: Sex differences and correlation with psychometric measures of spatial ability in humans. E olution and Human Behavior 19: 73–87
23. Morris R G M, Garrud P, Rowlings J N P, O’Keefe J 1982 Place navigation impaired rot, with hippocampal lesions. Nature 297: 681–83
24. O’Keefe J, Nadel L 1978 The Hippocampus as a Cognitive Map. Clarendon Press, Oxford, UK
25. O’Keefe J, Burgess N 1996 Geometric determinants of the place fields of hippocampal neurons. Nature 381(6581): 425–8
26. Packard M G, McGaugh J L 1996 Inactivation of hippocampus or caudate nucleus with lidocaine differentially affects expression of place and response learning. Neurobiology of Learning and Memory 65(1): 65–72
27. Ruddle R A (in press) Navigating overlapping virtual worlds: Arriving in one place and finding that you’re somewhere else. In: Freksa C, Brauer W, Habel C, Wender K F (eds.) Lecture Notes in Computer Science. Springer
28. Ruddle R A, Payne S J, Jones D M 1997 Navigating buildings in ‘desk-top’ virtual environments: Experimental investigations using extended navigational experience. Journal of Experimental Psychology: Applied 3: 143–59
29. Ruddle R A, Payne S J, Jones D M 1998 Navigating large-scale ‘desk-top’ virtual buildings: effects of orientation aids and familiarity. Presence 7: 179–92
30. Sandstrom N J, Kaufman J, Huettel S A 1998 Males and females use different distal cues in a virtual environment navigation task. Cognitive Brain Research 6: 351–60
31. Siegel A W, White S H 1975 The development of spatial representations of large scale environments. In: Reese H (ed.) Advances in Child Development and Behavior. Academic Press, New York, Vol. 10, pp. 10–55
32. Silverman I, Eals M 1992 Sex differences in spatial abilities: Evolutionary theory and data. In: Cosmides L, Tooby J (eds.) The Adapted Mind: Evolutionary Psychology and the Generation of Culture. Oxford University Press, New York, pp. 487–503
33. Snyder L H, Grieve K L, Brotchie P, Andersen R A 1998 Separate bodyand world-referenced representations of visual space in parietal cortex. Nature 394(6696): 887–91
34. Thorndyke P W, Hayes-Roth B 1982 Differences in spatial knowledge acquired from maps and navigation. Cognitive Psychology 14: 560–89
35. Vanderwolf C H 1969 Hippocampal electrical activity and voluntary movement in the rat. EEG. Clinical Neurophysiology 26: 407–18
36. Viaud-Delmon I, Ivanenko Y P, Berthoz A, Jouvent R 1998 Sex, lies and virtual reality. Nature Neuroscience 1(1): 15–16
37. Waller D, Hunt W, Knapp D 1998 The transfer of spatial knowledge in virtual environment training. Presence 7: 129–43
38. Witmer B G, Bailey J H, Knerr B W, Parsons K C 1996 Virtual spaces and real world places: Transfer of route knowledge. International Journal of Human Computer Studies 45: 413–28
39. Witmer B G, Klein P B 1998 Judging perceived and traversed distance in virtual environments. Presence 7: 144–67