IN THE PAST few years, virtual reality (VR)

Virtual Reality and Neuropsychology: Upgrading the Current Tools Background: Virtual reality (VR) is an evolving technology that has been applied in various aspects of medicine, including the treatment of phobia disorders, pain distraction interventions, surgical training, and medical education. These applications have served to demonstrate the various assets offered through the use of VR. Objective: To provide a background and rationale for the application of VR to neuropsychological assessment. Methods: A brief introduction to VR technology and a review of current ongoing neuropsychological research that integrates the use of this technology. Conclusions: VR offers numerous assets that may enhance current neuropsychological assessment protocols and address many of the limitations faced by our traditional methods. Key words: neuropsychological tests, technology, virtual reality Maria T. Schultheis, PhD Clinical Research Scientist Neuropsychology & Neuroscience Laboratory Kessler Medical Rehabilitation Research & Education Corporation Department of Physical Medicine & Rehabilitation University of Medicine & Dentistry of New Jersey-New Jersey Medical School West Orange, New Jersey Jessica Himelstein, MA Postdoctoral Fellow Department of Physical Medicine & Rehabilitation University of Medicine & Dentistry of New Jersey-New Jersey Medical School West Orange, New Jersey Albert A. Rizzo, PhD Assistant Research Professor Integrated Media Systems Center School of Gerontology University of Southern California Los Angeles IN THE PAST few years, virtual reality (VR) has undergone a transition that has taken it out of the realm of expensive toy and into that of functional technology. Continuing advances in VR technology, along with concomitant system cost reductions, have supported the development of more usable, useful, and accessible VR systems. These systems have the potential to uniquely target a wide range of physical, cognitive, and behavioral human issues and research questions. Neuropsychology is one discipline that may clearly benefit from the application of VR technology. Address correspondence and requests for reprints to Maria T. Schultheis, PhD, Neuropsychology and Neuroscience Laboratory, Kessler Medical Rehabilitation Research & Education Corporation, 1199 Pleasant Valley Way, West Orange, NJ 07052. Telephone: (973) 324 3528; Fax: (973)243 6984. E-mail: mschultheis@kmrrec.org. Supported in part by grant #HD08589 01 from the National Institute of Child Health and Human Development, grant #H133G000073 from the National Institute on Disability and Rehabilitation Research, and by the Integrated Media Systems Center, a National Science Foundation Engineering Research Center, Cooperative Agreement No. EEC-9529152. J Head Trauma Rehabil 2002;17(5):378 394 c 2002 Lippincott Williams & Wilkins, Inc. 378

Virtual Reality and Neuropsychology 379 The application of VR in neuropsychology is so distinctively important because it represents more than a simple linear extension of existing computer technology for human use. VR offers the potential to deliver systematic human testing and training environments that allow for the precise control of complex dynamic three-dimensional (3D) stimulus presentations, where sophisticated behavioral recording is possible. When combining these assets within the context of functionally relevant, ecologically valid virtual environments, a fundamental advancement emerges in how human behavior can be addressed in many scientific disciplines. In the broadest sense neuropsychology can be defined as an applied science that evaluates how specific activities in the brain are expressed in observable behaviors. 1 Effective neuropsychological assessment is often a key component in both the treatment and the scientific analysis of many central nervous system disorders and can serve a number of functions. These include the use of psychometric evaluation tools for diagnostic purposes, the provision of normative data for comparison with impaired cognitive and functional abilities, the specification of cognitive strengths and weaknesses to help inform the design of rehabilitative strategies, and provision of metrics to assess treatment efficacy or plot neurodegenerative decline. Neuropsychological assessment also serves to create data for the scientific understanding of brain functioning through the examination of measurable sequelae that occur after brain damage or dysfunction. To date, most neuropsychological research has focused on increasing our understanding of component cognitive processes, such as attention, executive functions, memory, and language spatial abilities. More recently, understanding the role of cognitive processes during functionally relevant tasks (i.e., instrumental activities of daily living) has evolved as an area of focus within neuropsychological research. VR offers the potential to integrate these two aspects of neuropsychology and provide an innovative approach to understanding brain behavior relationships. In view of the promising benefits that VR can offer neuropsychological assessment, treatment, and research, this article seeks to (1) provide a brief introduction to VR technology and the potential benefits of the application of VR to the field of neuropsychology, (2) provide a brief overview of recent studies that have successfully applied VR to the assessment of component cognitive abilities and functional evaluations, and (3) discuss important considerations in the application of VR. WHAT IS VR TECHNOLOGY? VR can be viewed as an advanced form of human computer interface that allows the user to interact with and become immersed in a computer-generated environment in a naturalistic fashion. 2 By analogy, much like an aircraft simulator serves to test and train piloting ability, computer-generated interactive virtual environments (VEs) can be designed to assess and rehabilitate cognitive and functional abilities by means of exposure to simulated real world and/or analog tasks. Interaction in three dimensions is a key characteristic that distinguishes a VR experience from other technologies. The believability of the virtual experience (or sense of presence ) is fostered by the use of such specialized technology as head-mounted displays (HMDs), tracking systems, earphones, gesture-sensing gloves, and haptic-feedback devices. An HMD is an image display system worn on the head (like a diving mask) that remains optically coupled to the user s eyes as he or she turns and moves. A tracking system senses the position and orientation of the user s head (and HMD) and reports that information to a computer that updates (in real

380 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 Fig 1. Person in a virtual environment wearing an HMD and head tracker. time) the images for display in the HMD. In most cases full-color stereo image pairs are produced, and earphones may deliver relevant 3D sound. The combination of an HMD and tracking system allows the computer to generate images and sounds in any computermodeled (virtual) scene that corresponds to what the user would see and hear from their current position if the scene were real (Figure 1). The user may walk and turn around to survey a virtual landscape or inspect a virtual object by moving toward it and peering around its sides or back. Although HMDs are most commonly associated with VR, other methods incorporating 3D projection walls and rooms (known as CAVES), as well as basic flat-screen computer systems, have been used to create interactive scenarios. Methods for navigation and interaction such as data gloves, joysticks, and some high-end force feedback mechanisms that can provide haptic feedback are also available and can be used to further enhance realism and the suspension of disbelief required to generate the sense of presence within a VR environment. Although VR remains a developing technology, it is not a new addition to the medical field. Various studies have served to document the successful integration of VR to myriad aspects of medicine, including surgical training, 3 education of patients and medical students, 4 and treatment of sensory mobility deficits. 5 Treatment of psychological dysfunction, including anxiety disorders and phobia, has been successfully reported by numerous researchers who have applied the use of VR for treating fear of flying and fear of heights. 6,7 VR has also been applied to the treatment of posttraumatic stress disorder (PTSD), 8 eating disorders, 9 and pain management. 10 Results from these studies have served to underscore some of the specific assets afforded

Virtual Reality and Neuropsychology 381 through VR applications and advantages over other more traditional assessment protocols. WHAT ARE THE ADVANTAGES OF APPLYING VR? Earlier discussions of VR technology in other application areas suggest a number of assets for its use. 11,12 Among these are several advantages that are of direct relevance to neuropsychological research, assessment, and treatment. One key benefit of applying VR protocols in the assessment of cognitive skills and abilities is its more naturalistic or real-life testing environment, which may serve to address one of the most commonly noted limitations of traditional neuropsychological measures. 13 That is, VEs may be especially well suited to improve ecological validity, or the degree of relevance or similarity that a test has relative to the real world. The complexity of stimulus challenges found in naturalistic settings could be delivered while still maintaining the experimental control required for rigorous scientific analysis. Results would have greater clinical relevance and could have direct implications for the development of more effective functional rehabilitation approaches. Full control over stimulus presentation and response measurement is another asset offered through the use of VR technology. Clinicians or researchers can specify the creation of highly controllable VR environments. Specifically, the use of VR protocols permits such factors as the number, speed, and/or order of stimulus presentation to be effortlessly manipulated while maintaining an objective means of data collection on relevant target responses. In addition, the relative flexibility of stimulus presentation in a VE can allow gradual increments of difficulty and challenge to be presented to the individual. This could allow for individualization of assessment and treatment, while still maintaining consistency in desired outcome measures. As such, VR technology could be used to supplement existing neuropsychological assessment procedures that traditionally rely mainly on penciland-paper tests and behavioral observation, potentially leading to improvements in psychometric reliability and validity and promoting the independent replication of research findings needed for scientific progress in this field. Another unique asset of VR for neuropsychological assessment is the opportunity to evaluate cognitive functions within dynamic interactions and environments. That is, although traditional measures can provide information regarding component cognitive functions and potentially predict how deficits may translate into the real world, VR offers a medium to examine this relationship directly and allow for the evaluation of complex cognitive behaviors. For example, although traditional measures may separately assess different systems of memory functioning (i.e., visual memory, verbal memory) within a VE, clinicians and researchers can evaluate how these two systems may or may not be used concomitantly by individuals during real-life tasks. Furthermore, such evaluations could serve to better identify the particular compensatory strategies initiated by individuals during these tasks, potentially resulting in more applicable and usable neuropsychological recommendations. In summary, the various assets offered by VR, including increased ecological validity, objectivity, and assessment in interactive and functionally relevant scenarios could support a number of opportunities for applications in both the clinical and research domains of neuropsychology. In recent years, a growing number of researchers have begun exploring the use of VR technology for applications designed to target neuropsychological assessment with populations having central nervous system dysfunction. Although the breadth of this clinical literature pales by comparison with VR research in the testing and training

382 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 area with normal populations, the initial efforts using VR designed for impaired clinical groups are encouraging. Summaries of some relevant studies that have applied VR to the evaluation of specific cognitive functions are provided in the following. COGNITIVE FUNCTIONS Executive functions Alexander Luria 14 defined executive functioning as those functions that are involved in the planning, regulation, and verification of an action. This definition has been expanded to include a set of behavioral competencies which include planning, sequencing, the ability to sustain attention, resistance to interference, utilization of feedback, the ability to coordinate simultaneous activity, cognitive flexibility and the ability to deal with novelty. 15 Given the complexity of this cognitive construct, the use of traditional psychometric methods for the assessment and rehabilitation of executive functions has been questioned. 16,17 One group of researchers has developed an HMD VE system specifically designed for the assessment and rehabilitation of this process in persons with traumatic brain injury (TBI), multiple sclerosis (MS), and stroke. 17,18 Using a standard tool of neuropsychological assessment as a model (Wisconsin Card Sorting Test WCST), these researchers have created a virtual building that requires the person to use environmental clues in the selection of appropriate choices (doorways) to navigate from room to room in the structure. The doorway choices can vary according to the categories of shape, color, and number of portholes, and the person is required to refer to the previous doorway for clues as to the appropriate next choice. When the administrator surreptitiously changes the choice criterion, the person being tested is then required to shift cognitive set, analyze clues, and devise and initiate a new choice strategy. The researchers concluded that the cognitive requirements of this task were similar to those included within the WCST. Using this VE system, Pugnetti et al 19 compared a mixed group of neurological patients (MS, stroke, and TBI) with healthy controls on performance on both the WCST and the VE executive function system. The results of this study were concordant with previous anecdotal observations by family members regarding the patient s everyday behavior. Although the psychometric properties of the VE task were comparable to the WCST in terms of gross differentiation of patients and controls, weak correlations between the two methods suggested that they were measuring different functions. In this regard, the VE task was seen to specify impairments earlier in the test sequence compared with the analog pencil-and-paper test. The authors suggest that...this finding depends on the more complex (and complete) cognitive demands of the VE setting at the beginning of the test when perceptuomotor, visuospatial (orientation), memory, and conceptual aspects of the task need to be fully integrated into an efficient routine. 19p.160 The detection of these integrative difficulties is of vital importance when predicting real-world capabilities from test results. The results of this study also included a discriminant function analysis that suggested the VR test was more effective at correctly classifying groups of healthy controls and patients with neurological damage than the WCST. Further support for the enhanced ecological validity of VR assessment can be seen in a detailed single-subject case study of a stroke patient using this system. In this report, 20 results indicated that the VE was more accurate than the WCST in specifying executive function deficits in a highly educated patient 2 years after a stroke. The VE was successful in detecting deficits that had been reported to be limiting the patient s everyday performance yet were

Virtual Reality and Neuropsychology 383 missed using traditional neuropsychological tests. Most recently, Elkind et al 21 devised a flatscreen VR analog of the WCST, similar in concept to the design of Pugnetti et al, 22 but with enhanced graphics and a more detailed environment. Although Pugnetti et al s scenario involved navigation through a house, the Elkind et al abstract reasoning scenario consisted of a beach scene, including four beach umbrellas each containing one of four stimuli. The four stimuli consisted of common objects differing in color and number. For the task, the participant was presented with a new object and was required to match the object to one of the four stimuli. Feedback regarding the accuracy of match performance was provided with each new item. To complete the test successfully, the participant had to learn the matching principle being used to determine the correct stimuli choice and use that principle to guide future choices. Correlational analysis of healthy control performance on Elkind et al s VR analog and the WCST indicated that participants generalized the strategies learned through administration of one measure and applied these techniques to completion of the second measure. Excluding one score on the WCST (perseverative errors), all performance scales on the WCST and VR analog were directly related, with correlations ranging from r =.33 (P <.01) to r =.58 (P <.001). The investigators interpreted this finding as evidence of learning transfer between the WCST and the VE. Although the VR task was more difficult for the participants than the WCST, subjective report indicated higher motivation levels for the VR task. Thus, use of the VR analog seems to alleviate the issues of poor motivation and ceiling effects present during paper-and-pencil cognitive testing. McGeorge et al 23 addressed the planning component of executive functioning. The authors observed the drawbacks of using traditional neuropsychological tests as measures of planning, including the lack of long-term continuous planning requirements, the absence of competing subtasks, and the poor relationship to real-life situations. Limitations of realworld assessments of planning skills include safety concerns and an inability to control for environmental factors that may confound performance. Rationale for the use of a VE included alleviation of most safety concerns, minimization of restrictions present in the real world because of physical impairments, the manipulation or elimination of environmental distracters, and improved relevance to real-life situations. Specifically, McGeorge et al s 23 work involved participants with TBI who were reported by family members to have planning deficits. Their performance was compared with that of healthy controls on both virtual and actual vocationally oriented planning tasks. The tasks were administered within a simulated work setting, where participants were given 20 minutes to complete a maximum number of work-related errands. Examples of errands included collecting office equipment from a cupboard, reading materials, addressing envelopes. TBI participants performed significantly worse than the healthy controls in both environments. The authors concluded that support for the enhanced ecological validity of VR assessment of planning skills was provided by the high correlation between number of errands completed in the real-world and virtual environments (r =.79, P <.01). The three aforementioned VEs seem to offer advantages for the valid and reliable assessment of executive functions relative to the more traditional paper-and-pencil tests typically used by neuropsychologists. Attention processes Sohlberg and Mateer 24 presented a clinical model of attention that outlines various levels of this basic cognitive process, including focused attention, sustained attention,

384 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 selective attention, alternating attention, and divided attention. The specific asset offered by VE technology for the assessment of attentional subprocesses is the provision of a controlled stimulus environment in which cognitive challenges can be presented along with the precise delivery and control of distracting auditory and visual stimuli. Efforts to assess attention are supported by the widespread occurrence of attentional deficits in numerous conditions across the lifespan, including attention deficit hyperactivity disorder (ADHD), acquired brain injury (ABI), and as a feature of various neurodegenerative disorders (e.g., Alzheimer s disease, vascular dementia). Although some investigators have highlighted the advantages that VR offers for the assessment of attention functions, 25,26 few studies have been conducted using VR to investigate this cognitive process. One group, Rizzo et al, 27 has designed a VR classroom environment to begin to examine the different aspects of attention in children with ADHD. The Virtual Classroom is an HMD VR system that consists of a standard rectangular classroom environment containing desks, a female teacher, a blackboard across the front wall, a side wall with a large window looking out onto a playground, and a street with moving vehicles, and, on each end of the opposite wall, a pair of doorways through which activity occurs. Within this scenario, children s attention is assessed while a series of typical classroom distracters (e.g., ambient classroom noise, activity occurring outside the window) are systematically controlled and manipulated within the VE. The child sits at a virtual desk within the virtual classroom, and attention can be measured in terms of reaction time performance and error profiles on a variety of attention challenge tasks that are delivered visually using the blackboard or auditorily by the teacher s voice. In the user-centered design phase, 20 nondiagnosed children (ages 6 12) were tested on basic selective and alternating attention tasks. Feedback pertaining to aesthetics and usability of the VE was solicited and incorporated into the iterative design-evaluate-redesign cycle. Overall results indicated little difficulty in adapting to use of the HMD, no self-reported occurrence of side effects determined by posttest interviews using the Simulator Sickness Questionnaire, 28 and excellent performance on the stimulus tracking challenges. After this phase, a clinical trial that compared 9 physician-referred ADHD boys (age 6 12) with 10 nondiagnosed children. The attention testing involved a continuous performance task delivered on the blackboard that required the participants to hit a response button whenever they saw the letter X preceded by the letter A. Two 10-minute conditions were presented to participants: one without distraction and one with distractions (pure audio, pure visual, and mixed audiovisual). VR performance was compared with results from standard neuropsychological testing. Preliminary findings from this study revealed significant difference in omission and commission error performance between ADHD children and nondiagnosed children, with the children with ADHD making more omission errors in the distracting condition. In addition, the number of movements made was also found to be significantly greater in children with ADHD relative to nondiagnosed children (tracked from head, arm, and leg). These initial findings suggest that performance in the Virtual Classroom could serve to help discriminate and diagnose the symptoms of ADHD among children. The VR Classroom may have potential as an efficient, cost-effective, and scalable tool for measuring attention performance beyond what exists using traditional methods. Investigation into performance on VE divided attention tasks by individuals with TBI and healthy controls has been conducted by Lengenfelder et al. 29 The VE consisted of a

Virtual Reality and Neuropsychology 385 driving course, where the primary task was to drive a car and the secondary task required the correct identification of a four-digit number presented during the driving course. The numbers were presented in a varied location (always in same place versus random placement) on the windshield of the vehicle (i.e., computer screen) and varied presentation rate (2.4 versus 0.6 ms) to differentiate between simple and complex divided attention. The simple attention condition required the subject to drive through the course (primary task) and identify the four-digit number (secondary task) that repeatedly appeared in the center of the screen at a slow rate. By contrast, the complex condition required the subject to drive through the course (primary task) and identify the four-digit number (secondary task) that appeared randomly on screen at a very fast rate. Performance measures included driving speed and correct number of stimuli identification. Pilot data found no differences in speed management, across both simple and complex tasks, between the two groups and reported that both groups drove faster when stimuli presentation was at a faster rate. The findings also suggested that the participants with a TBI had greater difficulty completing the secondary task than healthy controls. Early evidence for the validity of this VE as a measure of divided attention was provided by means of the correlation between performance on the VR task and on traditional neuropsychological measures of divided attention. Spearman s correlations calculated between the VR performance index of divided attention and neuropsychological measures of divided attention yielded correlations that ranged from r =.40 (P =.44) to r =.94 (P =.01). These correlations indicated that the more errors that were made on the VR divided attention task, the lower the number of correct responses on neuropsychological measures of divided attention. Thus the Virtual Classroom of Rizzo et al 27 and the driving course of Lengenfelder et al 29 provide two viable options for the use of VR to assess attentional functions in varied clinical and healthy populations. Visuospatial processes Spatial ability is commonly broken down and defined in terms of the tests used to measure its various subcomponents (e.g., spatial perception, orientation, visualization, mental rotation). However, Carpenter and Just 30 present a general definition of this process as the ability to generate a mental representation of a two or three-dimensional structure and then assessing its properties or performing a transformation of the representation (p. 221). VR has been applied to examine spatial processes and their generalization to everyday functions. For example, flat-screen VE systems have been used in studies of human place learning ability. 31 34 Flat-screen, joystick-controlled VEs modeled after the Morris Water Maze task 35 have been designed to allow a person to navigate an environment in search of a hidden platform. Studies using this VE have reported significant effects in the areas of aging, CNS dysfunction, and gender differences. 31,34 One recent study examining performance of this task in individuals with TBI demonstrated that these individuals have significantly impaired VE place learning abilities, which correlated with self-reported frequency of wayfinding problems in everyday life. 36 Other flat-screen VE scenarios have also shown promising results for training spatial orientation and navigation skills with physically disabled children whose inability to independently explore environments may engender developmental impairments in cognitive mapping abilities (i.e., the ability to visualize a route in the mind s eye). 37 In a series of four studies reviewed in Stanton et al, 37 children with physical disabilities were allowed to independently explore VEs modeled after

386 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 school environments. Transfer of spatial learning was shown to generalize to the actual school that was modeled, and this learning improved with practice. Also, flexibility in the cognitive representation, or mapping, of the environment was inferred, because the children were able to accurately indicate the direction of objects that were not in their line of sight. Another group that developed a visuospatial VE scenario has reported similar findings. 38 Successful learning and transfer of object locations were found using a flatscreen classroom VE with unimpaired children. In this study the spatial performance equaled real-world training, and successful transfer to the real environment was reported. Similar spatial learning using a flat-screen VE training system targeting functional spatial knowledge has also been reported for adolescents with developmental disabilities in a supermarket search-and-navigation scenario. 39 Taking a more component-based approach to addressing spatial ability, Rizzo et al 40 have developed a suite of projection-based 3D holographic VR-delivered applications targeting mental rotation, depth perception, manual movement stability, 3D field dependence (3D rod and frame test), 3D manual tracking, and visual field specific reaction time. An initial study with this system specifically targeted mental rotation (MR), a dynamic imagery process that involves turning something over in one s mind. 41 In the VE system, MR was assessed by means of a manual spatial rotation task that required subjects to manipulate virtual block configurations. Subjects were presented with a target block configuration, and the speed and efficiency of their movements to superimpose a replica design on the target was measured and recorded. An initial feasibility study with adult healthy controls investigated self-reported side effect occurrence, learning on the VE MR task, transfer of training from the VE to a pencil-and-paper MR test, 42 associations between VE performance and other standard tests of cognitive performance, and gender differences. A number of encouraging findings emerged, including minimal side effect occurrence, good psychometric properties of the VE test, provocative relationships with standard neuropsychological tests, a lack of gender differences compared with the penciland-paper measures, training improvement, and significant transfer of training. 27 The full VE system (all visuospatial scenarios cited earlier) is currently being tested with normal elderly subjects and comparisons of these data with central nervous system impaired populations (Alzheimer s disease, TBI) is planned. Given the myriad number of abilities included under the rubric of visuospatial functions, as well as the intuitive match between VR and assessment of visuospatial functions, more research has been conducted in this area relative to the use of VR for assessment of other cognitive functions. Memory functions The assessment and rehabilitation of memory disorders has received considerable attention in the neuropsychological literature. This has resulted in the better specification of memory processes, explanation of underlying central nervous system mechanisms, and a greater effort to determine the problems that occur in everyday life because of clinical memory disorders. This has also led to the acceptance of memory as a multicomponent system, including sensory memory, short-term memory, and long-term memory, 43 and working memory. 44 VR technology could be of value for improving the specificity of memory assessment for more effective diagnosis. Targeting of memory processes in VEs is closely related to VR work in the area of spatial orientation and way-finding. To this end, research examining memory for objects and spatial layout contained in a VE has been

Virtual Reality and Neuropsychology 387 underway since the mid-1990s at the University of East London (UEL). The UEL group has focused on specifying the types of memory that may be enhanced during a fourroom house navigation task. 45 This approach uses a flat-screen system with a joystick interface, which allows one subject to navigate the house (active condition), whereas a yoked subject is simply exposed to the same journey but has no navigational control (passive condition). Both subjects are directed to seek out an object (toy car) during the exploration, and differential memory performance between the two groups on spatial versus object memory of the environment is tested. In initial tests with normal populations, it was observed that the active groups showed better spatial memory for the route, whereas the passive group displayed superior object recall and recognition memory for the items viewed along the route. 46 Although one group replicated these results, 22 others have reported mixed findings. 13,47 Another study using this VE with clinical populations (MS and stroke) has produced results that suggest the value of this type of VE application to inform neuropsychological assessment. That is, performance was significantly poorer among stroke participants relative to performance of healthy controls. Interestingly, Although the typical spatial/content memory dissociation was found with unimpaired groups (i.e., active = better spatial memory; passive = better object memory) and the active stroke group displayed better spatial memory compared with the passive group, the stroke patients displayed no advantage on object memory while in the passive condition. 45 Similar findings using the UEL scenario were also reported by Pugnetti et al 19 in a study comparing individuals with MS to healthy controls. Another example of the use of VEs for spatial memory assessment was provided in a case report of a severely amnestic stroke patient who showed significant improvements in her ability to find her way around a rehabilitation unit after training within a VE modeled after the unit. 48,49 To date, investigations of memory functions using VEs have focussed far on spatial and object memory, and given the positive results future research using VR to address other aspects of memory functions may be fruitful. INSTRUMENTAL ACTIVITIES OF DAILY LIVING (IADLs) Some of the VE systems targeting component cognitive processes reported on previously could conceivably be reviewed here (i.e., virtual classroom, spatial navigation). The primary focus of this section is behaviors required for functional activities and/or IADLs. These functional approaches emphasize the design of ecologically valid VEs to test and train more integrated behavioral repertoires in populations who are limited in their capacity for real-world independent learning. Advances in this area could be of considerable value compared with the costly and timeconsuming efforts that are required for the in vivo, one-on-one approaches often required with certain clinical populations. In this way, VEs can be used to assess functional behavior and deliver consistent, hierarchical, and safe training that could free the therapist s time to be spent in other more necessary and intensive one-on-one client contact. Navigational skills VEs targeting functional skills have been developed in the areas of street crossing, wheelchair navigation, meal preparation, use of public transportation, driving, and obstacle avoidance. Strickland 50 reported initial results on the feasibility of using an HMD with autistic children with the long-term goal of targeting VE training for safe street crossing.

388 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 The two autistic children initially reported on adapted to the headset and were able to track moving automobile stimuli and select objects. Continued efforts with this population 51 and with children having motor disabilities 52 are underway to target the functional behaviors required for safe street crossing in a VE. Inman et al 51 have also reported success in training children with motor impairments to use motorized wheelchairs in a variety of simulated real-world conditions using an HMD VE system. Cooking behaviors Food preparation skills within a virtual kitchen scenario have also been investigated. Christiansen et al 53 have developed an HMD VE of a kitchen in which persons with TBI have been assessed in terms of their ability to perform 30 discrete steps required to prepare a can of soup. Contingent on success at each step, various auditory and visual cues can be presented to help prompt successful performance and learning. In the first reported pilot study with this system, 30 patients with closed-head injuries were able to successfully function using the HMD with minimal side effects. In addition, acceptable reliability coefficients were reported using a test/retest assessment. The stability of the VR assessment scores was analyzed using the intraclass correlation approach, coefficients of variation. The coefficient of variation for method error was computed for all intraclass correlations and ranged from 1.86% to 14.32%, indicating that the stability of most test items was good and that any low correlations were reflective of reduced variability and not a lack of consistency between test and retest. These researchers also report ongoing enhancements to the system regarding the delivery of more complex challenges and increased flexibility in the presentation of cueing stimuli. Learning and memory assessment in vocational environment One project, which is currently examining the use of VR for evaluating vocationally relevant cognitive functions, is being conducted with the use of a Virtual Office scenario. 54 Specifically, the VR Office is a relatively generic office scenario in which the subject could be seated at a desk or could be standing up. Currently on the subject s desk are a computer, telephone, and various desk accessories. There are other desks in the room, as well as filing cabinets and wall decorations. In addition, the office contains a large picture window and a clock that marks real time. A first study, examining learning and memory in individuals with TBI and MS, will use 16 target items (8 common and 8 uncommon) that will be placed throughout the VR Office. Measures on learning trials and verbal and visual recall will be collected. It is anticipated that learning and memory within the VR Office will allow for the use of verbal and visual memory compensatory strategies. As such, direct evaluation of performance may help to identify self-initiation of strategy use and allow evaluation of effectiveness of strategies. It is anticipated that such observations will allow for improvements in the identification and training of compensatory strategies. Driving assessment Virtual environments that target driving ability have been tested in TBI and elderly populations. Liu et al 55 reported that an HMD driving scenario successfully discriminated between the TBI group and an unimpaired group and that age effects were detected with this system. More recently, researchers have examined the use of VR Driving Assessment System for determining driving capacity after acquired brain injury (stroke, TBI). 56

Virtual Reality and Neuropsychology 389 The study compares a VR-based driving assessment (VRDS) protocol with the current gold standard of driving assessment, the behind-the-wheel evaluation. Specifically, the VRDS program consists of three driving environment scenarios. The first scenario was selected to allow direct comparison between driving performance on the behindthe-wheel (BTW) evaluation and a VR analog of this BTW route. The additional two driving scenarios were selected to allow examination of challenging factors on driving ability, including the addition of nighttime driving and stressful driving (i.e., congested traffic, pedestrian and emergency situations). Given the scarcity of studies with clinical populations in these environments, the effects of participation in a VRDS are also under investigation. That is, factors such as incidence of side effects, usability of the technology, and evaluation of factors contributing to performance (i.e., computer experience) are being examined. The pilot phase of the study is currently underway, and early observations suggest good usability of the technology with this clinical population. Finally, one particularly noteworthy and ambitious effort to address functional life skills is the work of the University of Nottingham group in the development of the Virtual City. 57 This flat-screen scenario addresses independent living skills in persons with learning and developmental disabilities, and the researchers have implemented good user-centered design principles by incorporating user group input in the selection of environments targeted in the application. These settings were designed to address the use of public transport, road safety, home safety, use of public facilities within a café, and shopping skills in a large supermarket. An initial study tested 20 subjects with mental retardation and addressed usability, enjoyment, skill learning, and transfer of skills to the real environment using subjects self-report and behavioral observation methods. 58 Subjects reported high enjoyment of the task and more ease of use with the system than anticipated in the expected usability ratings determined by support workers. Transfer from the VE to the real environment for some skills was noted in this initial feasibility study. Work with this evolving scenario continues, and results from more systematic measurement of learning and transfer are anticipated. Such results are important to support previous observations of transfer seen with these populations. For example, Cromby et al 39 reported evidence of positive learning transfer for another group of developmentally disabled students using a supermarket VE modeled after an actual market. Subjects were reported to be better able to navigate within and select specific items in the real supermarket after training in the VE, and these subjects outperformed those who practiced in nonspecific VEs. Positive results were also reported for a public transit VE training program called Train to Travel. 59 This HMD VE system was developed in the mid-1990s to assess and teach persons with developmental disabilities how to use key routes on the Miami Valley Regional Transit System. However, because the VE component was part of a comprehensive training package, including video and multimedia tutorials, the specific effect of the VE on successful independent transit use was not determinable. Future studies that methodologically target the specific effects of the VE ingredient are necessary to determine whether these types of programs are effective for efficiently and economically targeting functional abilities. The aforementioned investigations represent essential first steps in determining whether VE training can foster transfer of learning to activities of daily living. For persons whose learning abilities are challenged because of central nervous system dysfunction, this line of research is especially important.

390 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 CONCLUSIONS Although still an evolving technology, VR seems to offer many opportunities for the development of innovative, cost-effective methods of assessment and treatment of a variety of neuropsychological impairments. Although initial findings may be provocative, analysis of the cost-benefit issues for the application of VR technology is needed to better determine where VR may optimize neuropsychological research, assessment, and treatment. A detailed discussion of these topics can be found in Schultheis and Rizzo, 12 but some important considerations are highlighted in the following. While considering the advantages offered by VR, it is important to mention the minimal risk of side effects that is present when using VR technology. The most common risk of participation in a VE task is simulator sickness. Simulator sickness and motion sickness are similar in symptom constellation, but presentation is less severe in simulator sickness. Possible symptoms include vertigo, dizziness, headaches, and sweating. Although the rate of simulator sickness in clinical populations is not well known, prior studies reviewed herein have reported minimal (<10%) negative reactions in participants with MS and TBI. Additional research to further explain the impact of simulator sickness and/or other risks for VR participation of clinical populations will be crucial to the development and application of this technology. Another consideration involves questions frequently asked regarding the cost and availability of VR technology. 60 63 Continued advances in technology (i.e., hardware and software evolvement) and the steady decrease in prices of VR systems have served to minimize these concerns. A further concern has been raised regarding VR s ability to fit the needs of clinical populations. That is, although much of the work incorporating VR approaches has included healthy individuals and some psychiatric populations, studies that have included populations with neurological impairments (e.g., TBI, dementia) remain sparse. However, as previously noted, the few studies that have used clinical subject groups have reported a low rate of the most commonly seen adverse side effect of VR use (i.e., simulator sickness). Given the minimal risk level, the cost-benefit ratio for the use of VR as a research tool in clinical populations is favorable. Other considerations, including the relationship between disability (i.e., cognitive or physical impairment) and the complexity of VR systems, will require further research when considering future clinical applications. For example, the development of a VR environment for assessment purposes may be significantly altered by the selection of appropriate hardware (i.e., HMD versus flat screen). Although the enhanced 3D nature of the HMD and CAVE VR systems may provide more ecological validity than the flat-screen VR systems, the external real-world feedback that is provided when using flat-screen environments may minimize the adverse side effects possible with HMD and CAVE VR systems. However, the sense of immersion reported by participants in HMD, CAVE, and flat-screen VR environments has yet to be systematically compared. Research comparing the use of similar scenarios in HMD, CAVE, and flat-screen VR systems across healthy and clinical populations would also provide valuable information regarding the most appropriate environment for assessment purposes. Further research is also called for to test the notion that VEs are inherently ecologically valid. This may be accomplished by testing the same cognitive function across different VR scenarios. If VR scenarios do indeed tap the reported cognitive functions, then similar patterns of strengths and weaknesses should be revealed independent of the VR scenario used. The selection of which tasks (i.e., cognitive,

Virtual Reality and Neuropsychology 391 physical, functional) may be best addressed using VR with patient populations is an area that will require considerable future research. Although some applications may seem to be a natural match (i.e., driving), careful development of future applications should maintain a focus on the user and be driven by specific identified needs. Finally, ethical considerations will need to be clarified as new applications of this technology continue to evolve. For example, although VEs may be developed for clinically justifiable purposes (e.g., phobia treatment), it must be considered that the control of the VR experiences presented to the patient carries with it the capacity (and responsibility) to influence personal development of those patients exposed to this control. 64 Along these same lines, a number of authors have acknowledged the potential difficulties that could arise with the use of VEs by individuals with certain types of psychopathology. 65 67 An additional ethical concern raised by VR research is the mandate that research participants not be deceived. Given the sense of immersion and presence created in VR research relative to other computerized laboratory measures, participants may be more likely to believe that they are actually experiencing real-world events. Thus, events that transpire within VEs may be particularly salient for participants. This could be addressed by a debriefing session at the completion of the assessment session. Furthermore, the supposed enhanced ecological validity of VR may lead examiners to overestimate the relevance of test results to actual functional behaviors. A comprehensive discussion regarding the various ethical considerations in the application of VR technology and clinical treatment and research can be found in Rizzo et al. 40 In summary, like many emerging computer and information technologies, VR potentially offers numerous advantages and solutions to the field of neuropsychology. Clearly, an affordable tool that offers consistent yet modifiable ecologically valid testing and training environments tailored to the individual s needs could provide significant benefits to individuals with disabilities. The technology (for both hardware and software) continues to move forward, and work with clinical populations remains to be done. Although it is clear that a fair amount of work has been initiated in applying VR for the assessment of cognitive functions in clinical populations, much work is still needed to fully determine VR s contribution to neuropsychology. As such, the field is at a turning point, where much of the nascent efforts that hint at the potential of this technology need to be further explored and developed in depth. Effective application of VR will require the collaborative efforts of rehabilitation specialists, computer scientists, and engineers, and it is only through the integrated expertise of multiple fields that the potential benefits of VR can be fully realized. Although numerous challenges remain, the potential to augment traditional neuropsychological testing is impressive, and this application domain is at the early stages of development. As a larger body of VR-related literature emerges, we will be in a better position to formulate the standards for researchbased and clinically oriented applications. This could result in a new paradigm for using VR in research and clinical neuropsychology. REFERENCES 1. Lezak MD. Neuropsychological Assessment, 3rd ed. New York: Oxford University Press; 1995. 2. Rizzo AA, Buckwalter JG. Virtual reality and cognitive assessment and rehabilitation: the state of the

392 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 art. In: Riva G, ed. Psycho-neuro-physiological Assessment and Rehabilitation in Virtual Environments: Cognitive, Clinical and Human Factors in Advanced Human Computer Interactions. Amsterdam: IOS Press; 1997:123 146. 3. Cosman PH, Cregan PC, Martin CJ, Cartmill JA. Virtual reality simulators: current status in acquisition and assessment of surgical skills. ANZ J Surg. 2002; 72(1):30 34. 4. Satava RM. Medical applications of virtual reality. J Med Syst. 1995;19:275 280. 5. Burdea G, Deshpande S, Popescu V, et al. Computerized hand diagnostic/rehabilitation system using a force feedback glove. Stud Health Technol Inform. 1997;39:141 50. 6. Emmelkamp PM, Bruynzeel M, Drost L, van der Mast CA. Virtual reality treatment in acrophobia: a comparison with exposure in vivo. Cyberpsychol Behav. 2001;4(3):335 339. 7. Rothbaum BO, Hodges LF, Anderson PL, Price L, Smith S. Twelve month follow-up of virtual reality and standard exposure therapies for the fear of flying. J Consult Clin Psychol. 2002;70:428 432. 8. Rothbaum BO, Hodges LF, Alarcon R, et al. Virtual reality exposure therapy for PTSD Vietnam veterans: a case study. J Traumatic Stress. 1999;12:263 272. 9. Riva G. Virtual reality and body experience: a new approach to the treatment of eating disorders. Int J Virtual Reality. 1999;2(2):9 16. 10. Hoffman HG, Patterson DR, Carrougher GJ, Sharar SR. Effectiveness of virtual reality based pain control in multiple treatments. Clin J Pain. 2001;17:229 235. 11. Rizzo AA, Buckwalter JG, Neumann U. Virtual reality and cognitive rehabilitation: a brief review of the future. J Head Trauma Rehabil. 1997;12:1 15. 12. Schultheis MT, Rizzo AA. The application of virtual reality technology for rehabilitation. Rehabil Psychol. 2001;46:1 16. 13. Wilson BA. Ecological validity of neuropsychological assessment: do neuropsychological indexes predict performance in everyday activities? Appl Prev Psychol. 1993;2:209 215. 14. Luria A. The Working Brain: An Introduction to Neuropsychology (trans. B Haigh). New York: Basic Books; 1973. 15. Crawford J. Assessment of attention and executive functions. Neuropsychol Rehabil. 1998;8:209 211. 16. Elkind J. Uses of virtual reality to diagnose and habilitate people with neurological dysfunctions. CyberPsychol Behavior. 1998;1:263 274. 17. Pugnetti L, Mendozzi L, Attree E, et al. Probing memory and executive functions with virtual reality: Past and present studies. CyberPsychol Behavior. 1998;1:151 162. 18. Pugnetti L, Mendozzi L, Motta A, Cattaneo A, Barbieri E, Brancotti S. Evaluation and retraining of adults cognitive impairments: which role for virtual reality technology? Comput Biol Med. 1995;25:213 227. 19. Pugnetti L, Mendozzi L, Brooks BM, et al. Active versus passive exploration of virtual environments modulates spatial memory in MS patients: a yoked control study. Ital J Neurol Sci. 1998;19:S424 S430. 20. Mendozzi L, Motta A, Barbieri E, Alpini D, Pugnetti L. The application of virtual reality to document coping deficits after a stroke: report of a case. CyberPsychol Behavior. 1998;1:79 91. 21. Elkind JS, Rubin E, Rosenthal S, Skoff B, Prather P. A simulated reality scenario compared with the computerized Wisconsin Card Sorting test: an analysis of preliminary results. CyberPsychol Behavior. 2001;4:489 496. 22. Pugnetti L, Mendozzi L, Attree EA, et al. Probing memory and executive functions with virtual reality. CyberPsychol Behavior. 1998;1:151 161. 23. McGeorge P, Phillips LH, Crawford JR, et al. Using virtual environments in the assessment of executive dysfunction. Presence: Teleoperators Virtual Environ. 2001;10:375 383. 24. Sohlberg MM, Mateer CA. Introduction to Cognitive Rehabilitation. New York: Guilford Press; 1989. 25. Wann JP, Rushton SK, Smyth M, Jones D. Virtual environments for the rehabilitation of disorders of attention and movement. In: Riva G, ed. Virtual Reality in Neuro-Psycho-Physiology. Amsterdam: IOS Press; 1997:157 164. 26. Rizzo AA, Buckwalter JG, Humphrey L, et al. The virtual classroom: a virtual environment for the assessment and rehabilitation of attention deficits. CyberPsychol Behavior. 2000;3(2):483 501. 27. Rizzo A, Buckwalter J, Neumann U, et al. Virtual environments for targeting cognitive processes: an overview of projects at the University of Southern California Integrated Media Systems Center. Presence: Teleoperators Virtual Environ. 2001;10:384 400. 28. Kennedy RS, Lane NE, Berbaum KS, Lilienthal MG. Simulator sickness questionnaire: an enhanced method for quantifying simulator sickness. Int J Aviat Psychol. 1993;3:203 220. 29. Lengenfelder J, Schultheis MT, Al-Shihabi T, DeLuca J, Mourant R. Divided attention and driving: a pilot study using virtual reality technology. J Head Trauma Rehabil. 2002;17:26 37.

Virtual Reality and Neuropsychology 393 30. Carpenter JP, Just M. Spatial ability: an information processing approach to psychometrics. In: Sternberg RJ, ed. Advances in the Psychology of Human Intelligence, ed. 3. Hillsdale, NJ: Erlbaum; 1986. 31. Astur RS, Oriz ML, Sutherland RJ. The characterization of performance by men and women in a virtual morris water task: a large and reliable sex difference. Behav Brain Res. 1998;93:185 190. 32. Jacobs WJ, Laurance HE, Thomas K. Place learning in virtual space I: Acquisition, overshadowing and transfer. Learning Motivation. 1997;28:521 541. 33. Jacobs WJ, Laurance HE, Thomas K. Place learning in virtual space II: topographical relations as one measure of stimulus control. Learning Motivation. 1998;29:288 308. 34. Sandstrom NJ, Kaufman J, Heuttel SA. Males and females use different distal cues in a virtual environment navigation task. Cognitive Brain Res. 1995; 6:351 360. 35. Morris R. Spatial localization does not require the presence of local cues. Learning and motivation. Int J Neurosci. 1981;48:29 69. 36. Skelton RW, Bukach CM, Laurance HE, Thomas KG, Jacobs J. Humans with traumatic brain injuries show place learning deficits in computer generated virtual space. J Clin Exp Neuropscyhol. 2000;22:157 175. 37. Stanton D, Foreman N, Wilson P. Uses of virtual reality in clinical training: developing the spatial skills of children with mobility impairments. In: Riva G, et al, eds. Virtual Environments in Clinical Psychology and Neuroscience. Amsterdam: IOS Press; 1998:219 232. 38. McComas J, Pivik J, Laflamme M. Children s transfer of spatial learning from virtual reality to real environments. Cyberpsychol Behav. 1998;1:121 128. 39. Cromby J, Standen P, Newman J, Tasker H. Successful transfer to the real world of skills practiced in a virtual environment by student with severe learning disabilities. In: Sharkey PM, ed. Proceedings of the 1st European Conference on Disability, Virtual Reality and Associated technologies. Maidenhead, UK; 1996:305 313. 40. Rizzo AA, Schultheis MT, Rothbaum B. Ethical issues for the use of virtual reality in the psychological sciences. In: Bush S, Drexler M, eds. Ethical Issues in Clinical Neuropsychology. Lisse, NL: Swets & Zeitlinger Publishers; 2002. 41. Shepard RN, Metzler J. Mental rotation of threedimensional objects. Science. 1971;171:701 703. 42. Vandenbergm SG, Kluse AR. Mental rotations, a group test of three-dimensional spatial visualization. Percept Motor Skills. 1978;47:599 604. 43. Atkinson RC, Shiffrin RM. Human memory: a proposed system and its control processes. In: Spence KW, ed. The Psychology of Learning and Motivation: Advances in Research and Theory. Vol 2. New York: Academic Press; 1968. 44. Baddeley A. Working Memory. Oxford: Clarendon Press; 1986. 45. Rose FD, Attree EA, Johnson DA. Virtual reality: an assistive technology in neurological rehabilitation. Curr Opin Neurol. 1996;9:461 467. 46. Attree EA, Brooks BM, Rose FD, Andrews TK, Leadbetter AG, Clifford BR. Memory processes and virtual environments: I can t remember what was there, but I can remember how I got there. Implications for people with disabilities. In: Sharkey P, ed. Proceedings of the First European Conference on Disability, Virtual Reality, and Associated Technology. Maidenhead, UK: University of Reading; 117 121. 47. Peruch P, Verhcer J, Gauthier GM. Acquisition of spatial knowledge through visual exploration of simulated environments. Ecol Psychol. 1995;7:1 20. 48. Brooks BM, McNeil JE, Rose FD, Greenwood RJ, Attree EA, Leadbetter AG. Route learning in a case of amnesia: a preliminary investigation into the efficacy of training in a virtual environment. Neuropsychol Rehabil. 1999;9:63 76. 49. Brooks BM, Attree E, Rose FD, Clifford B, Leadbetter AG. The specificity of memory enhancement during interaction with a virtual environment. Memory. 1999;7:65 78. 50. Strickland D, Marcus LM, Mesibov GB, Hogan K. Brief report: two case studies using virtual reality as a learning tool for autistic children. J Autism Dev Disord. 1996;26:651 659. 51. Strickland D. Virtual reality for the treatment of autism. Stud Health Technol Inform. 1997;44:81 86. 52. Inman D, Peaks J, Loge K, Chen V. Teaching orthopedically impaired children to drive motorized wheelchairs in virtual reality. Proceedings from the 2nd Annual International Conference on Virtual Reality and Persons with Disabilities. San Francisco, CA; 1993. 53. Christiansen C, Abreu B, Ottenbacher K, Huffman K, Masel B, Culpepper R. Task performance in virtual environments used for cognitive rehabilitation after traumatic brain injury. Arch Phys Med Rehabil. 1998;79:888 892. 54. Schultheis MT, Rizzo AA. The virtual office: assessing & re-training vocationally relevant cognitive skills. Proceedings from the 9th Annual Medicine Meets Virtual Reality Conference, Newport Beach, CA; 2002.

394 JOURNAL OF HEAD TRAUMA REHABILITATION/OCTOBER 2002 55. Liu L, Miyazaki M, Watson B. Norms and validity of the DriVR: a virtual reality driving assessment for persons with head injuries. Cyberpsychol Behav. 1999;2:53 67. 56. Schultheis MT, Mourant RR. Virtual reality and driving: the road to better assessment of cognitively impaired populations. Presence: Teleoperators Virtual Environments. 2001;10:436 444. 57. Brown DJ, Kerr SJ, Bayon V. The development of the Virtual City: a user centered approach. In: Sharkey P, Rose D, Lindstrom J, eds. Proceedings of the 2nd European Conference on Disability, Virtual Reality and Associated Techniques. Skove, Sweden University of Reading; 1998: pp. 11 16. 58. Cobb SVG, Neale HR, Reynolds H. Evaluation of virtual learning environments. In: Sharkey P, Rose D, Lindstrom J, eds. Proceedings of the 2 nd European Conference on Disability, Virtual Reality and Associated Techniques. Sköve, Sweden: University of Reading; 1998:17 23. 59. Mowafty L, Pollack J. Train to travel. Ability. 1995;15: 18 20. 60. Jones LE. Does virtual reality have a place in the rehabilitation world? Disabil Rehabil. 1998;20:102 103. 61. Korpela R. Virtual reality: opening the way. Disabil Rehabil. 1998;20:106 107. 62. Wilson PN, Foreman N, Stanton D. Virtual reality, disability and rehabilitation. Disabil Rehabil. 1997; 19:213 220. 63. Foreman N, Orencas C, Nichols E, Morton P, Gell M. Spatial awarenes in 7- to 11-year old physically handicapped children in mainstream schools. Eur J Special Needs Educ. 1989;4:171 178. 64. Whalley LJ. 1996 Ethical consideration in the application of virtual reality to medicine. Comput Biol Med. 1989;25:107 114 65. Bloom RW. Psychiatric therapeutic applications of virtual reality technology (VRT): research prospectus and phenomenological critique. Stud Health Technol Inform. 1997;39:11 16. 66. Ring H. Is neurological rehabilitation ready for immersion in the world of virtual reality? Disabil Rehabil. 1998;20:98 101. 67. Rothbaum BO, Hodges LF, Kooper R, Opdyke D, Williford JS, North M. Effectiveness of computergenerated (virtual reality) graded exposure in the treatment of acrophobia. Am J Psychiatry. 1995; 152:626 628.