Assessing successful reproduction of egocentric and allocentric spatial representations using virtual reality

Abstract

Background. Currently in psychological rehabilitation the necessity of developing innovative methods for testing cognitive dysfunctions with via the modern sophisticated technology is becoming increasingly important. One of the urgent requests is associated with developing the methods of diagnostics and correction of spatial representations disorders, which are manifested by decreasing accuracy of spatial representations of the environment in particular. Objective. To study this issue the method for evaluating the accuracy of spatial information using which the ability to memorize the three-dimensional complex scenes was developed. It was assumed that the accuracy of reproduction would differ significantly depending on the coordinate (egocentric or allocentric) system of mental reconstruction processing. Design. The library of virtual objects and six unique virtual scenes were created. Each scene of seven objects was shown to the participants within the interval for 25 seconds. Thirty six subjects (aged from 18 to 26) participated in the experiment. They were told to memorize the objects and their locations, and then to reproduce the memorized scene using the given viewpoint of the scene. Three viewpoints were chosen: the «front» (to reproduce the scene from the egocentric position); the «left» and the» above» (to reproduce the memorized scene from on the left and above imaginary allocentric positions, respectively). To perform the task the participants chose objects from the library of virtual objects using the flystick 2 and placed them in virtual space in accordance with the memorized scene. The object locations in virtual space were recorded. Moreover, the accuracy of egocentric and allocentric representations in terms of measurements, topology and depth parameters were calculated. Conclusion. The results show that the egocentric representations (the «front» viewpoint) were more accurate for all parameters in comparison with the allocentric representations (the «left» and the «above» viewpoints), and the “above” representations were more accurate compared with the “left” ones. The topological accuracy was much better than the measurements and depth accuracy. Regardless of the viewpoints, the topological space parameters are stored in memory much more accurately than the depth parameters, which, in turn, are reproduced more accurately than metric parameters. It was also shown that the accuracy of spatial representations differs for different allocentric viewpoints: the «above» view is reproduced much more accurately than the «left» view. The method developed made it possible to reveal the features of encoding spatial information in ER and AP blocks in terms of measurements, topology and depth parameters. It can be used in clinical rehabilitation to test impairments in the perception of space, and also violations of short-term memory. The results obtained allow refining the existing models of encoding spatial information.

Keywords

• short-term memory
• accuracy of spatial representation cording
• egocentric and allocentric systems
• virtual reality
• VR
• CAVE