Virtual Environments in Neuropsychological Assessment ... - CiteSeerX

GIUSEPPE RIVA (Ed.) Virtual Reality in Neuro-Psycho-Physiology 1997, 1998 © Ios Press: Amsterdam, Netherlands.

Virtual Environments in Neuropsychological Assessment and Rehabilitation F. D. Rose, E. A. Attree and B.M. Brooks Department of Psychology University of East London London E15 4LZ UK Abstract. Brain damage constitutes a major problem for those affected, for their families and friends and for society as a whole. The need for effective rehabilitation strategies is clear. Yet, until the early 1960s, the brain was generally considered to be a somewhat fixed and inflexible organ. In consequence the impairments associated with brain damage were generally regarded as "incurable". Since that time neuroscientists have had reason to change their views dramatically. However, much remains to be done. Progress depends upon a co-ordinated multidisciplinary approach within which assistive technology will be a key player. Within the area of assistive technology, one of the developments which holds particular promise for the field of neurological rehabilitation is the computer technology underlying virtual environments (commonly known as virtual reality). In this chapter we describe the new opportunities offered by virtual reality to pursue several aspects of the rehabilitation process The value of the technology of virtual environments in this context is that it allows us to immerse people with brain damage in relatively realistic interactive environments which, because of their patterns of impairment, would otherwise be unavailable to them.

1. Introduction It has been estimated that the incidence of traumatic brain injury is approximately 250 per 100,000 of the population [1]. Estimates of the incidence of damage to the brain due to strokes are similar [2]. In total brain damage constitutes a major problem for those affected, for their families and friends and for society as a whole. The need for effective rehabilitation strategies is clear. Yet the field of restorative neurology [3] or neurological rehabilitation [4] is relatively new. Until the early 1960s the brain was generally considered to be a somewhat fixed and inflexible organ, largely "hard-wired" before birth and unable to replace nerve cells subsequently lost through accident or disease. In consequence the impairments associated with brain damage were generally regarded as "incurable". Since that time neuroscientists have had reason to change their views dramatically. It is now widely recognised that "........ far from being fixed, unchangeable and static, we now know that the brain is a dynamic and interactive organ, constantly changing in terms


of cellular activity, neural circuitry and transmitter chemistry in response to demands placed upon it." [5] This revision of the way in which the brain is viewed has provided a major impetus to those seeking to develop rehabilitation programmes for people with brain damage. However, much remains to be done. Progress depends upon a co-ordinated multidisciplinary approach within which assistive technology will be a key player. Within the area of assistive technology, one of the developments which holds particular promise for the field of neurological rehabilitation is the computer technology underlying virtual environments (commonly known as virtual reality). The value of the technology of virtual environments in this context is that it allows us to immerse people with brain damage in relatively realistic interactive environments which, because of their patterns of impairment, would otherwise be unavailable to them. This, in turn, allows us new opportunities to pursue several aspects of the rehabilitation process. Some of these are discussed below.

2. Virtual environments in assessing function in people with brain damage The first step in any rehabilitation process is a full appraisal of the patient's current functional profile. The arguments for seeking to assess cognitive function within the context of interaction with virtual environments have been rehearsed on a number of occasions [6,7,8,9,10]. Conventional "paper and pencil" neuropsychological tests of cognitive function have frequently been criticised as lacking "ecological validity" [11]. In other words, it has been argued that the measures of cognitive function (memory, attention etc.) derived from such tests may not give a realistic estimate of that person's probable cognitive capacity in a real world setting. In response to these arguments some psychologists have developed "everyday tests" of cognitive function. Examples include the Rivermead Behavioural Memory Test [12] and the Everyday Attention Test [13]. However, it has been argued that the cost of introducing ecological validity into these measures has been a loss of strict control over the test situation [14]. We have argued that the use of virtual environments provides a potential solution to this problem in allowing patients to interact with life-like everyday environments without entailing any loss of control over the test situation. Additionally, virtual environments can be programmed to help counteract reduced sensory capacity in the patient and allow movement within the environment despite an impaired motor repertoire. There are still further advantages in virtual tests of cognitive function. As well as allowing us to estimate a patientÕs probable cognitive level in a real life simulation, we can separately analyse different elements of that situation in order to identify the precise focus of any cognitive impairment. This approach is illustrated in an investigation of memory reported by Andrews et al. [6]. In the first of two experiments the authors compared recognition memory for objects of five


groups of student participants following exposure to the following five stimulus presentation conditions. · Condition 1 (VR) - Objects presented in interactive virtual rooms. In this condition the objects presented to participants were contained in a virtual house consisting of four interconnected virtual rooms presented on a computer screen. Participants were instructed to move through each room in turn using a joystick. · Condition 2 (no context - NO CON) - Objects presented in static displays (no context). Participants were presented with four static displays on the computer screen. These displays consisted of the objects from the four virtual rooms presented without any background context. Participants were instructed to look at each display in turn. · Condition 3 (context - CON) - Objects presented in the context of static pictures of the virtual rooms. Here objects were contained in a series of four consecutively presented computer pictures showing the views from the doorways in each of the virtual rooms used in condition 1. Participants were again instructed to look at each display in turn. · Condition 4 (context with motor - CON/MOTOR) - Objects presented in the context of static pictures of the virtual rooms but the participants were instructed to move the cursor over each of the objects. The displays of objects in this condition were the same as for condition 2. · Condition 5 (no context with motor - NO CON/MOTOR) - Objects in static displays (no context) with motor task. The displays here were the same as in condition 3 but participants were asked to move the cursor over each of the objects. Object recognition scores are shown in Figure 1.

Figure 1. Object recognition scores (see text for description of the experimental conditions).


Analysis of variance revealed significant differences between the conditions (F(4,99) = 33.9, p